UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

The value of nutritional assessment in major abdominal surgery

Haverkort, E.B.

Link to publication

Citation for published version (APA):Haverkort, E. B. (2014). The value of nutritional assessment in major abdominal surgery.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 23 Jul 2020

The

valu

e o

f nu

trit

ion

al a

sses

smen

t in

maj

or

abd

om

inal

su

rger

y

Elizabeth Haverkort

The value of nutritional assessment in m

ajor abd

ominal surgery Elizabeth H

averkort

The value of nutritional assessment in major abdominal surgery

Elizabeth Haverkort

‘Let food be thy medicine’Hippocrates

Cover: Part of the fresco ‘The meeting at the Golden Gate’ (Domenico Beccafumi, 1486 -1551), Ospedale di Santa Maria della Scala, Siena, Italy. The Santa Maria della Scala, one of the first hospitals in Europe, was founded around 1090 and closed at the end of the 70s of the last century.

Financial support for the publication of this thesis was kindly received from Abbott B.V., ABN/AMRO kantoor AMC, Baxter B.V., Fresenius Kabi Netherlands B.V., Mediq Tefa, Nederlands Voedingsteam Overleg (NVO), Nestlé Health Science, Nutricia Advanced Medical Nutrition, Solgar Vitamins, Sorgente B.V., Universitaire Master Opleiding Evidence Based Practice (EBP) AMC-UvA, and Yakult Nederland B.V.

Layout and printing by: Buijten & Schipperheijn, AmsterdamCopyright 2014 ©E.B. Haverkort, Amsterdam, The Netherlands

The value of nutritional assessment in major abdominal surgery

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctoraan de Universiteit van Amsterdamop gezag van de Rector Magnificus

Prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde

commissie, in het openbaar te verdedigen in de Agnietenkapel op woensdag 28 mei 2014 te 14.00 uur

door

Elizabeth Barbara Haverkort

geboren te Beverwijk

Promotiecommissie

Promotores Prof. dr. D.J. Gouma Prof. dr. R.J. de Haan

Co-promotor Dr. J.M. Binnekade

Overige leden Prof. dr. J.J.G.H.M. Bergman Prof. dr. M.A. Cuesta Prof. dr. J.H.G. Klinkenbijl Prof. dr. H.W.M. van Laarhoven Prof. dr. E.M.H. Mathus – Vliegen Prof. dr. H. Obertop

Faculteit der Geneeskunde

5

Contents

Chapter 1 General introduction and outline of the thesis

Chapter 2 Self-reporting of height and weight: valid and reliable identification of malnutrition in preoperative patients.

Am J Surg. 2012 Jun;203(6):700-7.

Chapter 3 Handgrip strength by dynamometry does not identify malnutrition in individual preoperative outpatients.

Clin Nutr. 2012 Oct;31(5):647-51.

Letter to the editor. Clin Nutr. 2012 Oct;31(5):778 Response from the authors. Clin Nutr. 2012 Oct;31(5):779-80.

Chapter 4 Estimation of body composition depends on applied device in patients undergoing major abdominal surgery.

Submitted

Chapter 5 Bioelectrical impedance analysis to estimate body composition in surgical and oncological patients: a systematic review.

Submitted

Chapter 6 Presence and persistence of nutrition-related symptoms during the first year following esophagectomy with gastric tube reconstruction

in clinically disease-free patients. World J Surg. 2010 Dec;34(12):2844-52.

Chapter 7 Suboptimal intake of nutrients after esophagectomy with gastric tube reconstruction.

J Acad Nutr Diet. 2012 Jul;112(7):1080-7.

Chapter 8 General discussion

English summary

Nederlandse samenvatting

Addendum List of publications

Dankwoord

Curriculum Vitae

7

21

37

4950

53

67

93

107

121

135

139

145

148

151

Chapter 1

General introduction and outline of the thesis

8

The main goal of this thesis is to determine the value of nutritional assessment in pre-operative and postoperative patients, in particular those undergoing major abdominal surgery. This introduction addresses the consequences of malnutrition for these patients, discusses frequently used tools for assessing malnutrition and pays attention to bioelectri-cal impedance as a method to measure body composition. Finally, an outline of the thesis is given.

Nutrition Inadequate intake of fluids, macronutrients (proteins, carbohydrates and fat) and micro-nutrients (vitamins, minerals and trace elements) results in deterioration of physical condi-tion and may ultimately lead to death. The need for specific nutrients in relation to optimal nutritional status has been understood and described for centuries. Examples of break-throughs in nutrition research are the following discoveries: the association between the symptoms of scurvy and eating citrus fruit, published in 1753 by the British naval physician J. Lind (1716-1794), 1 the cause of beri beri (vitamin B1 deficiency) by the Dutch physician and later Nobel prize winner C. Eijkman, 2, 3 and the impact of fatty acids on the occurrence of cardiovascular disease, resulting from information gathered by H.O. Bang, J. Dyerberg and A. Brøndum Nielsen among the Inuit in Greenland.4

Nutrition and illness in ancient timesThousands of years ago, the Egyptians were aware that certain nutrients or food products had a positive influence on the course of an illness. They used herbs not only for preparing their food, but they also used indigenous and imported herbs for healing. Their papyrus scrolls described the effect of acacia in relation to cough, the use of pomegranate against tapeworms, henbane against colic, cumin and coriander against intestinal cramps and cel-ery and saffron in the treatment of rheumatic complaints.

In the Far East, the knowledge of medicine among the Chinese emperors was approxi-mately equal to that of the Egyptians; for an optimal balance between yin and yang they also used medicinal herbs.

Described and preserved on stone tablets, in sanatoriums such as Epidaurus (500 BC), the ancient Greeks treated their patients with a combination of relaxation therapies, physi-cal exercise, intellectual entertainment and dietary therapies. However, Hippocrates (460-377 BC) should be regarded as the founder of scientific, modern dietetics, as he definitively broke with the mythical and religious basis of Greek medicine. Next to hygiene, he was convinced that good eating and drinking habits were essential to optimal health.

In the Roman period, Aulus Cornelius Celsus (25 BC–50) and Claudius Galenius (129–199) were respected physicians who improved on various dietary requirements and thera-pies initiated by Hippocrates.

During the Middle Ages, monks had extensive medical knowledge but were not allowed by the clergy to apply their knowledge and skills. In these dark ages, surgeons and quacks carried out surgical procedures and bloodletting, but only doctors were concerned

9

Cha

pte

r 1G

ener

al in

trod

ucti

on a

nd o

utlin

e of

the

thes

is

with diets and lifestyle. It was not until the Renaissance that medicine and the associated knowledge about dietary therapies underwent a revival and improvement.

Nutrition in the present eraCurrent knowledge about nutrition has resulted in the formulation of requirements for energy and the recommended dietary allowances (RDAs) for macronutrients and micro-nutrients. The RDAs are the daily intake level of a nutrient considered to be sufficient to meet the requirements of 97- 98% of healthy people in every demographic segment of a country. It should be noted that RDAs differ between countries, and that some nutrients are specified by some countries while others are not. In healthy adults, the aim of opti-mal nutrition is to meet the RDAs to safeguard health, and for children and adolescents to ensure proper growth. 5 – 13

Nutrition during illness It is widely assumed that the nutritional requirements of ill people differ from those of healthy subjects. Depending on their physical condition, patients are often advised to increase or decrease the intake of certain nutrients. However, the actual needs for patients in terms of macronutrients and micronutrients are mostly unknown. 14-15

Nutritional goals and malnutrition in surgical patients Nutrition assessment and therapy in surgical patients during the preoperative, periop-erative and postoperative phase is of major clinical importance. The primary goal of nutri-tional therapy in these patients is the maintenance or restoration of the protein mass of the body, as it is assumed that substantial loss of fat-free mass negatively influences the course of disease and treatment. 16-18 In this context, the evaluation of body weight and body mass index is crucial. For example, substantial involuntary loss of body weight within a certain time frame and/or a body mass index below a defined cut-off point – which is defined as malnutrition – have been widely described as having a negative influence on adjuvant therapy and postoperative outcome. 18 – 24

Malnutrition in hospitals is a serious problem, with an international prevalence range between 5% and 55%, depending on type of patient, department, hospital and interven-tion. 25 - 33 Information about the presence of malnutrition in outpatient clinics is scarce, but recent studies have indicated a prevalence between 6% and 12%. 34 – 36 Risk factors for malnutrition are older age, co-morbidities and care dependency. 16, 25, 27 Diseases of the intestinal tract may also lead to malnutrition as a result of dysphagia, obstruction, vomit-ing, reduced digestion, impaired absorption or diarrhoea.16, 37 ,38 Malignant disease and the corresponding treatment also increase the risk of becoming malnourished. This is caused by or is in response to tumour-released inflammatory cytokines and hormones, catabo-lism (metabolic pathway to release energy) and cancer cachexia (loss of body weight, inflammation, and significant loss of appetite). 26, 39 - 42

Malnutrition reduces the fat-free mass resulting in a decline of vital physiological

10

functions: a perceptible loss of muscle strength and muscle function and reduced organ function, including the immune system. 43 - 48 In addition, malnutrition may reduce the response to chemotherapy and radiotherapy and increase the risk of postoperative mor-bidity, surgery related mortality, impaired quality of life and prolonged hospital stay. 25, 49 – 52

Major abdominal surgery Within the population of surgical patients, patients undergoing major abdominal sur-gery are a special subgroup, as the disease and the required surgical procedure have a direct impact on the functioning of the gastrointestinal tract. Patients in this subgroup often suffer from cancer of the esophagus, stomach and pancreas and are therefore at high risk of malnutrition during all phases – preoperative, perioperative and postop-erative. In the preoperative phase the malignant process in the gastro-intestinal tract often results in difficulties such as obstruction of swallowing, vomiting, nausea, pain, poor digestion, malabsorption and diarrhoea. In the perioperative phase patients in this subgroup are often not allowed, or are unable, to eat and drink for at least several days; artificial nutritional support is then often required. In the post-operative phase, all patients experience, to a greater or lesser extent, the physiological consequence of their major surgical procedure in terms of changes in the dietary pattern and inadequate food intake. 53 – 74

To increase the 5-year survival rate, the current treatment for cancer patients selected for major abdominal surgery often consists of the surgical procedure in combination with chemo and/or radiation therapy. It may be evident that side effects of these neo-adjuvant and adjuvant treatments like nausea, vomiting, pain and obstruction increases the risk of inadequate oral intake and can result in a deterioration of nutritional status in the preop-erative and postoperative phase. 75

Nutrition-related consequences of esophagectomy with gastric tube reconstruc-tion As described above, major abdominal surgery can strongly influence the functioning of the gastrointestinal tract and negatively affect the postoperative nutritional status in terms of body weight, nutrient intake and malnutrition. However, evidence-based nutri-tion related guidelines to support patients in the postoperative phase are scarce. The literature on this topic has indicated that patients who undergo an esophagec-tomy with gastric tube reconstruction (esophagectomy) are confronted with a range of nutrition-related difficulties and complaints such as dysphagia (difficulty in swallowing), reflux, early satiety, altered gastric emptying, dumping syndrome as well as deteriora-tion of nutritional status. 53 - 69, 76 - 78 Unfortunately, a systematic in-depth evaluation with regard to the nutrition related complaints lacks, and little is known about the short-term and long-term course of these symptoms, the nutrition-related adjustments needed, and the patient’s nutritional status in terms of body weight and intake of macronutrients and micronutrients in the first postoperative year.

11

Cha

pte

r 1G

ener

al in

trod

ucti

on a

nd o

utlin

e of

the

thes

is

Dietary advice for patients undergoing major abdominal surgery The protein requirement of healthy adults in the Netherlands is set at 0.8 grams/kilogram body weight/24 hours. However, in vivo neutron activation analysis has indicated that this level of intake does not maintain the protein body mass after major abdominal surgery. 79 A limited number of studies have demonstrated that the maximal protein synthesis for healthy persons and in case of sepsis is 1.5–1.7 gram grams of protein/kg body weight/24 hours and this advice was adopted by the National Dutch Guideline on Perioperative Nutrition. 16, 80, 81 This higher level of protein intake is prescribed during chemotherapy and radiotherapy, the preoperative phase and in the postoperative phase until six months after major abdominal surgery. 16, 43, 44, ,49, 82, 83 Thereafter, protein intake can be reduced to 1.2–1.3 grams/kg/24 hours, as patients are expected to be less catabolic by then. According to the National Guideline on Perioperative Nutrition for the Netherlands, the energy requirements for surgical patients are calculated with the Harris and Benedict equation (1984) plus 30% extra (20% for metabolic stress and 10% for activity). 16, 84, 85 A stable body weight is assumed to be important, as substantial weight loss indicates loss of fat-free mass, including muscle mass. 43, 44, 86-88

Screening tools to detect malnutrition Early detection and treatment of malnutrition is important, as the adverse consequences of malnutrition for both patient and society are considerable. 34 – 36, 89 – 91 However, systematically measuring height and weight to calculate the patient’s body mass index (BMI) is time consum-ing for caregivers. Our unpublished data shows that measuring height and weight requires 2-3 minutes extra time per patient. If healthcare professionals were to measure all preoperative outpatients in the Netherlands, this would result in over 32,000 additional working hours yearly.

To save time for caregivers, malnutrition screening tools have been introduced to obtain a rough estimate of patients’ nutritional status. In the Netherlands, the Short Nutritional Assessment Questionnaire (SNAQ) and the Malnutrition Universal Screening Tool (MUST) are frequently used for adults, and the Mini Nutritional Assessment (MNA) for the elderly.92 - 96 Not all of these screening tools have been validated for preoperative outpatients, and some of the tools must be used by trained personnel. In addition, ‘at risk’ patients need a complementary and more comprehensive assessment.16, 25, 33 - 36, 90, 91, 97

Self-reported anthropometric data as indicator for malnutritionSelf-reported anthropometric data on height and weight has been suggested as a substi-tute for screening tools. However, previous studies have indicated that healthy persons and patients suffering from eating disorders tend to under-report their body weight and over-report their height, which results in an underestimated body mass index. 98 – 102 In underweight patients this relationship seems to be reversed; patients over-report their body weight, which leads to an underestimation of malnutrition. 99 In preoperative surgi-cal outpatients, no studies have yet determined the adequacy of self-reported weight and height data to screen for malnutrition.

12

Handgrip strength measurement by dynamometry as indicator of malnutritionEarlier studies demonstrated significant associations between low handgrip strength and aspects such as malnutrition, postoperative complications, prolonged hospital stay, reduced ability to return home, reduced mobility, impaired quality of life and mortality. 44, 45, 103 - 107 A number of algorithms based on handgrip strength are available to screen for malnutrition 108, 109 or an increased risk for postoperative complications. 45, 110 Frequently used algorithms are those proposed by Mathiowetz et al. 111 Álvares-da-Silva et al., 109

Klidjian et al., 45 Matos et al., 108 and Webb et al. 110 Each of these algorithms uses its own cut-off points of normal values, but little is known about the screening abilities of these algorithms in individual patients. 45, 105 - 107

Measurement of body composition to assess nutritional status In recent years, it has been increasingly recommended to not only measure body weight and calculate body mass index, but also to measure the various body compartments in order to determine nutritional status more accurately. A number of reference methods are available to evaluate body compartments. For the measurement of total body water, deu-terium dilution and tritiated water dilution are used. Extracellular water is measured by bromide dilution. Intracellular water is measured by radioactive 40K total body potassium. Fat-free mass, fat mass and lean tissue mass can be estimated by hydrostatic weighing (hydro densitometry or underwater weighing) and by air-displacement plethysmogra-phy and dual energy X-ray absorptiometry. Magnetic resonance imaging and computed tomography are used for the measurement of fat mass, muscle, skin, viscera and bone tis-sue. Body cell mass can be estimated by radioactive total body potassium; and total body nitrogen by neutron activation. However, the applicability of these methods in clinical populations is limited, due to the fact that most of these methods are expensive, time con-suming and not always harmless 46, 47, 112 - 119

Bioelectrical impedance analysis of the entire body to evaluate nutritional status In the 1990s, estimations of body compartments by bioelectrical impedance analysis (BIA) became available. BIA is an easy, non-invasive method to estimate aspects such as total body water, extracellular water, fat-free mass and fat mass. The method is based on meas-uring the resistance and reactance of an alternating electrical current in the human body. Intracellular fluids, body fluids and electrolytes behave as electrical conductors (resist-ance) and cell membranes act as electric al condensers and are involved in capacitance (reactance). 46, 47 , 113, 114 , 118 - 134 To actually estimate the body compartments, the measured resistance and reactance are incorporated into to a statistical regression equation consid-ered most suitable for a certain target population. The equation usually consists of a set of person-related variables such as age, sex, height and body weight.

BIA estimates can be performed by various devices. In clinical practice, BIA estimates are routinely used in patients undergoing major abdominal surgery (often for cancer), but it is unclear whether BIA provide a valid estimation of body compartments.

13

Cha

pte

r 1G

ener

al in

trod

ucti

on a

nd o

utlin

e of

the

thes

is

Outline of this thesisThe overall aim of this thesis objective is to investigate the value of nutritional assessment in major abdominal surgery. The first four chapters present the clinimetric properties and the applicability of nutritional status related methods and techniques in both a general preoperative population and a group of patients undergoing major abdominal surgery. A detailed nutritional assessment in terms of nutrition related complaints, nutrition related adjustments, body weight and intake of nutrients after esophagectomy with gastric tube reconstruction is described in the subsequent chapters.

The first two chapters of the thesis report on a study of the applicability of two meth-ods for the detection of malnutrition in preoperative individual outclinic patients Chapter 2 focuses on the reliability and validity of self-reported anthropometric data on height and weight of preoperative surgical patients compared to anthropometric data assessed by healthcare professionals and three commonly used malnutrition screening tools. In Chapter 3 the diagnostic accuracy of handgrip strength by dynamometry in preoperative surgical patients is addressed.

Chapters 4 and 5 pay attention to the measurement of body composition to assess nutritional status. Chapter 4 reports on an evaluation of the estimates of two bioelectrical impedance analysis (BIA) devices with respect to the body compartments fat-free mass and fat mass among patients undergoing major abdominal surgery. Inter-observer agree-ment between the two devices was analysed in order to determine whether their estima-tions result in a similar classification of body composition. In Chapter 5 a systematic review is presented with the aim of determining the validity of BIA estimations in adult surgical and oncological patients.

Chapters 6 and 7 address the impact of major surgery on nutrition-related complaints, nutritional intake and risk of deterioration of the nutritional status in patients after esophagectomy with gastric tube reconstruction. Chapter 6 describes the presence of per-sistent nutrition-related complaints and the necessity of nutrition-related adjustments, and Chapter 7 reports on the intake of nutrients and potential risks of nutrient deficiencies one year after esophagectomy.

In Chapter 8 the main results of the studies are summarised, and unexpected study findings are discussed. Furthermore, the clinical implications and recommendations for future research are described. A summary in English and Dutch concludes the thesis.

References 1. J Lind. A treatise on the scurvy in three parts: containing an inquiry into the nature, causes and cure of that

disease, together with a critical and chronological view of what has been published on the subject (3rd edn.) Crowder, London (1772)

2. Fairfield KM, Fletcher RH. Vitamins for chronic disease prevention in adults: scientific review. JAMA. 2002 Jun 19;287(23):3116-26.

3. Fletcher RH, Fairfield KM. Vitamins for chronic disease prevention in adults: clinical applications. JAMA. 2002 Jun 19;287(23):3127-9.

14

4. Bang HO, Dyerberg J, Nielsen AB. Plasma lipid and lipoprotein pattern in Greenlandic West-coast Eskimos. Lancet. 1971 Jun 5;1(7710):1143-5.

5. FAO/WHO/UNU. Energy and protein requirements. Geneva: WHO; 1985. Technical Report Series 724 ISBN 92 4 120724 8.

6. Institute of Medicine. Dietary Reference Intakes for Thiamin, Riboflavin, Niacin, Vitamin B6, Folate, Vitamin B12, Pantothenic Acid, Biotin, and Choline. Washington, DC: National Academy Press; 2000.

7. Institute of Medicine. Dietary Reference Intakes for Vitamin C, Vitamin E, Selenium, and Carotenoids. Was-hington, DC: National Academy Press; 2000.

8. Institute of Medicine. Dietary Reference Intakes for Vitamin A, Vitamin K, Arsenic, Boron, Chromium, Copper, Iodine, Iron, Manganese, Molybdenum, Nickel, Silicon, Vanadium, and Zinc. Washington, DC: National Academy Press; 2001.

9. Institute of Medicine. Dietary Reference Intakes for Calcium, Phosphorous, Magnesium, Vitamin D, and Fluoride. Washington, DC: National Academy Press; 1999.

10. The Health Council of the Netherlands. Sufficient intake of vitamins and minerals. The Hague: Health Council of the Netherlands, 2009; publication number 2009/06.

11. The Health Council of the Netherlands. Dietary standards. The Hague: Health Council of the Netherlands, 2000; publication number 2000/12.

12. The Health Council of the Netherlands. Dietary standards: energy, proteins, fat and digestible carbohydrates. The Hague: Health Council of the Netherlands, 2001; publication number 2001/19.

13. The Health Council of the Netherlands. Optimal use of folic acid. The Hague: Health Council of the Netherlands, 2008; publication number 2008/02.

14. Sauerwein HP, Pesola GR, Brennan MF. Lowering of plasma glucose concentration in septic cancer-bearing patients: metabolic significance. Neth J Med. 1992 Dec;41(5-6):208-17.

15. Sauerwein HP, Romijn JA, Soeters PB. Kunstmatige voeding bij door ziekte veranderde stofwisseling. Elsevier/Bunge, Maarsen 1998.

16. CBO Guideline. National Dutch Guideline on Perioperative Nutrition, 2007. Dutch Institute for Healthcare Improvement CBO, Utrecht, the Netherlands. http://www.cbo.nl/product/richtlijnen/ folder 20021023121843/rl_periovoed_07.pdf. Accessed September, 2007.

17. Wolfe RR. Is the double-blind randomized trial the most valid experimental approach to evaluating treatment modalities in critical ill patiënts? Curr Opin Clin Nutr Metab Care. 1998 March; 20 (1): 185-186.

18. Stratton RJ, Green CJ, Elia M. Scientific Criteria for Defining Malnutrition. In: Disease-related Malnutrition: An Evidence-based Approach to Treatment, 1st ed. Wallingford: CABI Publishing; 2003:1-34.

19. Elia M, Ritz P, Stubbs RJ. Total energy expenditure in the elderly. Eur J Clin Nutr. 2000; 54 (suppl):S92-103. 20. van Bokhorst-de van der Schueren MA, van Leeuwen PA, Sauerwein HP, Kuik DJ, Snow GB, Quak JJ.

Assessment of malnutrition parameters in head and neck cancer and their relation to postoperative complications. Head Neck 1997 Aug;19(5):419-25.

21. Kelly IE, Tessier S, Cahill A, Morris SE, Crumley A, McLaughlin D, McKee RF, Lean ME. Still hungry in hospital: identifying malnutrition in acute hospital admissions. QJM 2000 Feb;93(2):93-8.

22. Perioperative total parenteral nutrition in surgical patients. The Veterans Affairs Total Parenteral Nutrition Cooperative Study Group. N Engl J Med 1991 Aug 22;325(8):525-32.

23. Detsky AS, Smalley PS, Chang J.The rational clinical examination. Is this patient malnourished? JAMA. 1994 Jan 5;271(1):54-8.

24. Soeters PB, Reijven PL, van Bokhorst-de van der Schueren MA, Schols JM, Halfens RJ, Meijers JM, van Gemert WG. A rational approach to nutritional assessment. Clin Nutr 2008 Oct;27(5):706-16.

25. Baldwin C, Weekes CE. Dietary advice with or without oral nutritional supplements for disease-related malnutrition in adults. Cochrane Database Syst Rev. 2011 Sep 7;(9):CD002008.

26. Bozzetti F, Gavazzi C, Miceli R, Rossi N, Mariani L, Cozzaglio L, Bonfanti G, Piacenza S., Perioperative total parenteral nutrition in malnourished, gastrointestinal cancer patients: a randomized, clinical trial. JPEN J Parenter Enteral Nutr. 2000 Jan-Feb;24(1):7-14.

27. Halfens RJ, Meijers JM, Du Moulin MF, van Nie NC, Neyens JC, Schols JM. Rapportage resultaten Landelijke Prevalentiemeting Zorgproblemen 2010. Retreived from http://www.nld.lpz-um.eu/media/text/home/LPZ_rapport_2010_.pdf

28. The Malnutrition Prevalence Group. Edington J, Boorman J, Durrant ER, Perkins A, Giffin CV, James R, Thomson JM, Oldroyd JC, Smith JC, Torrance AD, Blackshaw V, Green S, Hill CJ, Berry C, McKenzie C, Vicca N, Ward JE, Coles SJ. Prevalence of malnutrition on admission to four hospitals in England. Clin Nutr. 2000 Jun;19(3):191-5.

29. Corish CA, Kennedy NP. Undernutrition in hospitals. Br J Nutr. 2001 Apr;85(4):509-10.

15

Cha

pte

r 1G

ener

al in

trod

ucti

on a

nd o

utlin

e of

the

thes

is

30. Kelly IE, Tessier S, Cahill A, Morris SE, Crumley A, McLaughlin D, McKee RF, Lean ME. Still hungry in hospital: identifying malnutrition in acute hospital admissions. QJM. 2000 Feb;93(2):93-8.

31. Kyle UG, Pirlich M, Schuetz T, Luebke HJ, Lochs H, Pichard C. Prevalence of malnutrition in 1760 patients at hospital admission: a controlled population study of body composition. Clin Nutr. 2003 Oct;22(5):473-81.

32. Allison S P. Malnutrition, disease and outcome. Nutrition.2000 Jul-Aug;16(7-8):590-3.33. McWhirter JP, Pennington CR. Incidence and recognition of malnutrition in hospital. BMJ. 1994 Apr

9;308(6934):945-8.34. Neelemaat F, Kruizenga HM, de Vet HC, Seidell JC, Butterman M, van Bokhorst-de van der Schueren MA.

Screening malnutrition in hospital outpatients. Can the SNAQ malnutrition screening tool also be applied to this population? Clin Nutr. 2008 Jun;27(3):439-46.

35. Leistra E, Neelemaat F, Evers AM, van Zandvoort MH, Weijs PJ, van Bokhorst-de van der Schueren MA, Visser M, Kruizenga HM. Prevalence of undernutrition in Dutch hospital outpatients. Eur J Intern Med. 2009 Sep;20(5):509-13.

36. Wilson MM, Vaswani S, Liu D, Morley JE, Miller DK. Prevalence and causes of undernutrition in medical outpatients. Am J Med. 1998 Jan;104(1):56-63.

37. Ravasco P. Nutritional support in head and neck cancer: how and why? Anticancer Drugs. 2011 Aug;22(7):639-46.

38. Blum D, Omlin A, Baracos VE, Solheim TS, Tan BH, Stone P, Kaasa S, Fearon K, Strasser F; European Palliative Care Research Collaborative. Cancer cachexia: a systematic literature review of items and domains associated with involuntary weight loss in cancer. Crit Rev Oncol Hematol. 2011 Oct;80(1):114-44.

39. Fearon KC. Cancer cachexia and fat-muscle physiology. N Engl J Med. 2011 Aug 11;365(6):565-7. 40. Tisdale MJ. Cancer cachexia. Curr Opin Gastroenterol. 2010 Mar;26(2):146-51.41. Muscaritoli M, Molfino A, Gioia G, Laviano A, Rossi Fanelli F. The “parallel pathway”: a novel nutritional and

metabolic approach to cancer patients. Intern Emerg Med. 2011 Apr;6(2):105-12. 42. Arends J. Metabolism in cancer patients. Anticancer Res. 2010 May;30(5):1863-8.43. Windsor JA, Hill GL. Weight loss with physiologic impairment. A basic indicator of surgical risk. Ann Surg.

1988 Mar;207(3):290-6.44. Windsor JA, Hill GL. Grip strength: a measure of the proportion of protein loss in surgical patients. Br J Surg.

1988 Sep;75(9):880-2.45. Klidjian AM, Foster KJ, Kammerling RM, Cooper A, Karran SJ. Relation of anthropometric and dynamometric

variables to serious postoperative complications. Br Med J. 1980 Oct 4;281(6245):899-901.46. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Gómez JM, Heitmann BL, Kent-Smith L, Melchior

JC, Pirlich M, Scharfetter H, Schols AM, Pichard C; Composition of the ESPEN Working Group. Bioelectrical impedance analysis--part I: review of principles and methods. Clin Nutr. 2004 Oct;23(5):1226-43.

47. Ellis KJ. Human body composition: in vivo methods. Physiol Rev. 2000 Apr;80(2):649-80.48. Heitmann BL. Prediction of body water and fat in adult Danes from measurement of electrical impedance.

A validation study. Int J Obes. 1990 Sep;14(9):789-802.49. Sauerwein HP, Serlie MJ. Optimal nutrition and its potential effect on survival in critically ill patients. Neth J

Med. 2010 Mar;68(3):119-22.50. Braga M, Ljungqvist O, Soeters P, Fearon K, Weimann A, Bozzetti F. ESPEN Guidelines on Parenteral Nutrition:

surgery. Clin Nutr. 2009 Aug;28(4):378-86. 51. van den Berg MG, Rasmussen-Conrad EL, Wei KH, Lintz-Luidens H, Kaanders JH, Merkx MA. Comparison of

the effect of individual dietary counselling and of standard nutritional care on weight loss in patients with head and neck cancer undergoing radiotherapy. Br J Nutr. 2010 Sep;104(6):872-7.

52. Isenring EA, Capra S, Bauer JD. Nutrition intervention is beneficial in oncology outpatients receiving radiotherapy to the gastrointestinal or head and neck area. Br J Cancer. 2004 Aug 2;91(3):447-52.

53. Baba M, Aikou T, Natsugoe S, Kusano C, Shimada M, Kimura S, Fukumoto T. Appraisal of ten-year survival following esophagectomy for carcinoma of the esophagus with emphasis on quality of life. World J Surg. 1997 Mar-Apr;21(3):282-5.

54. de Boer AG, van Lanschot JJ, van Sandick JW, Hulscher JB, Stalmeier PF, de Haes JC, Tilanus HW, Obertop H, Sprangers MA. Quality of life after transhiatal compared with extended transthoracic resection for adenocarcinoma of the esophagus. J Clin Oncol. 2004 Oct 15;22(20):4202-8.

55. Martin L, Jia C, Rouvelas I, Lagergren P. Risk factors for malnutrition after oesophageal and cardia cancer surgery. Br J Surg. 2008 Nov;95(11):1362-8.

56. Martin L, Lagergren P. Long-term weight change after oesophageal cancer surgery. Br J Surg. 2009 Nov;96(11):1308-14.

57. Poghosyan T, Gaujoux S, Chirica M, Munoz-Bongrand N, Sarfati E, Cattan P. Functional disorders and quality

16

of life after esophagectomy and gastric tube reconstruction for cancer. J Visc Surg. 2011 Oct;148(5):e327-35. 58. Baker A, Wooten LA, Malloy M. Nutritional considerations after gastrectomy and esophagectomy for

malignancy. Curr Treat Options Oncol. 2011 Mar;12(1):85-95. 59. McLarty AJ, Deschamps C, Trastek VF, Allen MS, Pairolero PC, Harmsen WS. Esophageal resection for cancer

of the esophagus: long-term function and quality of life. Ann Thorac Surg. 1997 Jun;63(6):1568-72.60. Ludwig DJ, Thirlby RC, Low DE. A prospective evaluation of dietary status and symptoms after near-total

esophagectomy without gastric emptying procedure. Am J Surg. 2001 May;181(5):454-8.61. Verschuur EM, Steyerberg EW, Kuipers EJ, Essink-Bot ML, Tran KT, Van Der Gaast A, Tilanus HW, Siersema

PD. Experiences and expectations of patients after oesophageal cancer surgery: an explorative study. Eur J Cancer Care (Engl). 2006 Sep;15(4):324-32.

62. Wainwright D, Donovan JL, Kavadas V, Cramer H, Blazeby JM. Remapping the body: learning to eat again after surgery for esophageal cancer. Qual Health Res. 2007 Jul;17(6):759-71.

63. Finley FJ, Lamy A, Clifton J, Evans KG, Fradet G, Nelems B. Gastrointestinal function following esophagectomy for malignancy. Am J Surg. 1995 May;169(5):471-5.

64. Blazeby JM, Williams MH, Brookes ST, Alderson D, Farndon JR. Quality of life measurement in patients with oesophageal cancer. Gut. 1995 Oct;37(4):505-8.

65. Moraca RJ, Low DE. Outcomes and health-related quality of life after esophagectomy for high-grade dysplasia and intramucosal cancer. Arch Surg. 2006 Jun;141(6):545-549.

66. Conroy T, Marchal F, Blazeby JM. Quality of life in patients with oesophageal and gastric cancer: an overview. Oncology. 2006;70(6):391-402.

67. De Leyn P, Coosemans W, Lerut T. Early and late functional results in patients with intrathoracic gastric replacement after oesophagectomy for carcinoma. Eur J Cardiothorac Surg. 1992;6(2):79-84.

68. van Knippenberg FC, Out JJ, Tilanus HW, Mud HJ, Hop WC, Verhage F. Quality of Life in patients with resected oesophageal cancer. Soc Sci Med. 1992 Jul;35(2):139-45.

69. Kight CE. Nutrition considerations in esophagectomy patients. Nutr Clin Pract. 2008 Oct-Nov;23(5):521-8.70. Radiagan AE. Post-gastrectomy: managing the nutrition fall-out. Practical Gastroenterology 2004 June; 28

(6):63 – 74. 71. Qu H, Sun GR, Zhou SQ, He QS. Clinical risk factors of delayed gastric emptying in patients after pancreati-

coduodenectomy: a systematic review and meta-analysis. Eur J Surg Oncol. 2013 Mar;39(3):213-23. 72. Lermite E, Sommacale D, Piardi T, Arnaud JP, Sauvanet A, Dejong CH, Pessaux P. Complications after

pancreatic resection: Diagnosis, prevention and management. Clin Res Hepatol Gastroenterol. 2013 Jun;37(3):230-9.

73. Lim H, Cho G, Kim S. Evaluation of nutrient intake and diet quality of gastric cancer patients in Korea. Nutr Res Pract. 2012 Jun;6(3):213-20.

74. Rogers C. Postgastrectomy nutrition. Nutr Clin Pract. 2011 Apr;26(2):126-36. 75. Awad S, Tan BH, Cui H, Bhalla A, Fearon KC, Parsons SL, Catton JA, Lobo DN. Marked changes in body

composition following neoadjuvant chemotherapy for oesophagogastric cancer. Clin Nutr. 2012 Feb;31(1):74-7.

76. Ryan AM, Rowley SP, Healy LA, Flood PM, Ravi N, Reynolds JV. Post-oesophagectomy early enteral nutrition via a needle catheter jejunostomy: 8-year experience at a specialist unit. Clin Nutr. 2006 Jun;25(3):386-93.

77. Gawad KA, Hosch SB, Bumann D, Lübeck M, Moneke LC, Bloechle C, Knoefel WT, Busch C, Küchler T, Izbicki JR. How important is the route of reconstruction after esophagectomy: a prospective randomized study. Am J Gastroenterol. 1999 Jun;94(6):1490-6.

78. Page RD, Oo AY, Russell GN, Pennefather SH. Intravenous hydration versus naso-jejunal enteral feeding after esophagectomy: a randomised study. Eur J Cardiothorac Surg. 2002 Nov;22(5):666-72.

79. Sevette A, Smith RC, Aslani A, Kee AJ, Hansen R, Barratt SM, Baxter RC. Does growth hormone allow more efficient nitrogen sparing in postoperative patients requiring parenteral nutrition? A double-blind, placebo-controlled randomised trial. Clin Nutr. 2005 Dec;24(6):943-55.

80. Jeevanandam M, Lowry SF, Horowitz GD, Legaspi A, Brennan MF. Influence of increasing dietary intake on whole body protein kinetics in normal man. Clin Nutr. 1986 Feb;5(1):41-8.

81. Shaw JH, Wildbore M, Wolfe RR. Whole body protein kinetics in severely septic patients. The response to glucose infusion and total parenteral nutrition. Ann Surg. 1987 Mar;205(3):288-94.

82. Steering Committee Malnutrition. Screening and treatment of malnutrition. In: Guideline Malnutrition. Amsterdam, the Netherlands, 2009.

83. Symons TB, Sheffield-Moore M, Wolfe RR, Paddon-Jones D. A moderate serving of high-quality protein maximally stimulates skeletal muscle protein synthesis in young and elderly subjects. J Am Diet Assoc. 2009 Sep;109(9):1582-6.

17

Cha

pte

r 1G

ener

al in

trod

ucti

on a

nd o

utlin

e of

the

thes

is

84. Weijs PJ, Kruizenga HM, van Dijk AE, van der Meij BS, Langius JA, Knol DL, Strack van Schijndel RJ, van Bokhorst-de van der Schueren MA. Validation of predictive equations for resting energy expenditure in adult outpatients and inpatients. Clin Nutr. 2008 Feb;27(1):150-7.

85. Sauerwein HP, Strack van Schijndel RJ. Perspective: How to evaluate studies on peri-operative nutrition? Considerations about the definition of optimal nutrition for patients and its key role in the comparison of the results of studies on nutritional intervention. Clin Nutr. 2007 Feb;26(1):154-8.

86. Heymsfield SB, McManus C, Stevens V, Smith J. Muscle mass: reliable indicator of protein-energy malnutrition severity and outcome. Am J Clin Nutr. 1982 May;35(5 Suppl):1192-9.

87. Bourdel-Marchasson I, Joseph PA, Dehail P, Biran M, Faux P, Rainfray M, Emeriau JP, Canioni P, Thiaudière E. Functional and metabolic early changes in calf muscle occurring during nutritional repletion in malnourished elderly patients. Am J Clin Nutr. 2001 Apr;73(4):832-8.

88. Lopes J, Russell DM, Whitwell J, Jeejeebhoy KN. Skeletal muscle function in malnutrition. Am J Clin Nutr. 1982 Oct;36(4):602-10.

89. Baldwin C, Parsons TJ. Dietary advice and nutritional supplements in the management of illness-related malnutrition: systematic review. Clin Nutr. 2004 Dec;23(6):1267-79.

90. Naber TH, Schermer T, de Bree A, Nusteling K, Eggink L, Kruimel JW, Bakkeren J, van Heereveld H, Katan MB. Prevalence of malnutrition in nonsurgical hospitalized patients and its association with disease complications. Am J Clin Nutr. 1997 Nov;66(5):1232-9.

91. Smedley F, Bowling T, James M, Stokes E, Goodger C, O’Connor O, Oldale C, Jones P, Silk D. Randomized clinical trial of the effects of preoperative and postoperative oral nutritional supplements on clinical course and cost of care. Br J Surg. 2004 Aug;91(8):983-90.

92. Kruizenga HM, Seidell JC, de Vet HC, Wierdsma NJ, van Bokhorst-de van der Schueren MA. Development and validation of a hospital screening tool for malnutrition: the short nutritional assessment questionnaire (SNAQ). Clin Nutr. 2005 Feb;24(1):75-82.

93. Vellas B, Villars H, Abellan G, Soto ME, Rolland Y, Guigoz Y, Morley JE, Chumlea W, Salva A, Rubenstein LZ, Garry P. Overview of the MNA--Its history and challenges. J Nutr Health Aging. 2006 Nov-Dec;10(6):456-63.

94. Guigoz Y. The Mini Nutritional Assessment (MNA) review of the literature--What does it tell us? J Nutr Health Aging. 2006 Nov-Dec;10(6):466-85.

95. Elia M. (editor) The ‘MUST’ report. A report by the Malnutrition Advisory Group of the British Association for Parenteral and Enteral Nutrition (BAPEN). Nutritional screening for adults: a multidisciplinary responsibility. Development and use of the‘ Malnutrition Universal Screening Tool’ (‘MUST’) for adults 2003.

96. Kondrup J, Allison SP, Elia M, Vellas B, Plauth M; Educational and Clinical Practice Committee, European Society of Parenteral and Enteral Nutrition (ESPEN). ESPEN guidelines for nutrition screening 2002. Clin Nutr. 2003 Aug;22(4):415-21.

97. Davies M. Nutritional screening and assessment in cancer-associated malnutrition. Eur J Oncol Nurs. 2005;9 Suppl 2:S64-73.

98. Engstrom JL, Paterson SA, Doherty A, Trabulsi M, Speer KL. Accuracy of self-reported height and weight in women: an integrative review of the literature. J Midwifery Womens Health. 2003 Sep-Oct;48(5):338-45.

99. Rowland ML. Self-reported weight and height. Am J Clin Nutr. 1990 Dec;52(6):1125-33.100. Gunnell D, Berney L, Holland P, Maynard M, Blane D, Frankel S, Smith GD. How accurately are height, weight

and leg length reported by the elderly, and how closely are they related to measurements recorded in childhood? Int J Epidemiol. 2000 Jun;29(3):456-64.

101. Taylor AW, Dal Grande E, Gill TK, Chittleborough CR, Wilson DH, Adams RJ, Grant JF, Phillips P, Appleton S, Ruffin RE. How valid are self-reported height and weight? A comparison between CATI self-report and clinic measurements using a large cohort study. Aust N Z J Public Health. 2006 Jun;30(3):238-46.

102. Spencer EA, Appleby PN, Davey GK, Key TJ. Validity of self-reported height and weight in 4808 EPIC-Oxford participants. Public Health Nutr. 2002 Aug;5(4):561-5.

103. Watters DA, Haffejee AA, Angorn IB, Duffy KJ.Nutritional assessment by hand grip dynamometry. S Afr Med J. 1985 Oct 12;68(8):585-7.

104. Norman K, Schütz T, Kemps M, Josef Lübke H, Lochs H, Pirlich M. The Subjective Global Assessment reliably identifies malnutrition-related muscle dysfunction. Clin Nutr. 2005 Feb;24(1):143-50.

105. Guo CB, Zhang W, Ma DQ, Zhang KH, Huang JQ. Hand grip strength: an indicator of nutritional state and the mix of postoperative complications in patients with oral and maxillofacial cancers. Br J Oral Maxillofac Surg. 1996 Aug;34(4):325-7.

106. Corish CA. Pre-operative nutritional assessment. Proc Nutr Soc. 1999 Nov;58(4):821-9.107. Jakobsen LH, Rask IK, Kondrup J. Validation of handgrip strength and endurance as a measure of physical

function and quality of life in healthy subjects and patients. Nutrition. 2010 May;26(5):542-50.

18

108. Matos LC, Tavares MM, Amaral TF. Handgrip strength as a hospital admission nutritional risk screening method. Eur J Clin Nutr. 2007 Sep;61(9):1128-35.

109. Álvares-da-Silva MR, Reverbel da Silveira T. Comparison between handgrip strength, subjective global assessment, and prognostic nutritional index in assessing malnutrition and predicting clinical outcome in cirrhotic outpatients. Nutrition. 2005 Feb;21(2):113-7.

110. Webb AR, Newman LA, Taylor M, Keogh JB. Hand grip dynamometry as a predictor of postoperative complications reappraisal using age standardized grip strengths. JPEN J Parenter Enteral Nutr. 1989 Jan-Feb;13(1):30-3.

111. Mathiowetz V, Kashman N, Volland G, Weber K, Dowe M, Rogers S. Grip and pinch strength: normative data for adults. Arch Phys Med Rehabil. 1985 Feb;66(2):69-74.

112. Lukaski H C. Methods for the assessment of human body composition: traditional and new. Am J Clin Nutr. 1987 Oct;46(4):537-56.

113. Cox-Reijven PLM. The validation of bio-electrical impedance spectroscopy (BIS) for measuring body composition in patients. Thesis. Maastricht: ISBN 90-5681-17-9, 2002.

114. Dittmar M. Reliability and variability of bioimpedance measures in normal adults: effects of age, gender, and body mass. Am J Phys Anthropol. 2003 Dec;122(4):361-70.

115. Jaffrin M Y, Morel H. Body fluid volumes measurements by impedance: A review of bioimpedance spectroscopy (BIS) and bioimpedance analysis (BIA) methods. Med Eng Phys. 2008 Dec;30(10):1257-69.

116. Heymsfield SB, Wang Z, Visser M, Gallagher D, Pierson RN Jr. Techniques used in the measurement of body composition: an overview with emphasis on bioelectrical impedance analysis. Am J Clin Nutr. 1996 Sep;64(3 Suppl):478S-484S.

117. Ellis K J, Shypailo R J, Pratt J A, Pond W G. Accuracy of dual-energy x-ray absorptiometry for body-composition measurements in children. Am J Clin Nutr. 1994 Nov;60(5):660-5.

118. Mattsson S, Thomas BJ. Development of methods for body composition studies. Phys Med Biol. 2006 Jul 7;51(13):R203-28.

119. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Manuel Gómez J, Lilienthal Heitmann B, Kent-Smith L, Melchior JC, Pirlich M, Scharfetter H, M W J Schols A, Pichard C; ESPEN. Bioelectrical impedance analysis-part II: utilization in clinical practice. Clin Nutr. 2004 Dec;23(6):1430-53.

120. De Lorenzo A, Andreoli A, Matthie J, Withers P. Predicting body cell mass with bioimpedance by using theoretical methods: a technological review. J Appl Physiol. 1997 May;82(5):1542-58.

121. Selberg O, Selberg D. Norms and correlates of bioimpedance phase angle in healthy human subjects, hospitalized patients, and patients with liver cirrhosis. Eur J Appl Physiol. 2002 Apr;86(6):509-16.

122. Barbosa-Silva MC, Barros AJ. Bioelectric impedance and individual characteristics as prognostic factors for post-operative complications. Clin Nutr. 2005 Oct;24(5):830-8.

123. Zarowitz BJ, Matthie JR. Bioimpedance and the estimation of net fluid balance in critical care patients: problems and possibilities. Crit Care Med. 1999 Aug;27(8):1655-7.

124. Toso S, Piccoli A, Gusella M, Menon D, Crepaldi G, Bononi A, Ferrazzi E. Bioimpedance vector pattern in cancer patients without disease versus locally advanced or disseminated disease. Nutrition. 2003 Jun;19(6):510-4.

125. Aghdassi E, Arendt B, Salit IE, Allard JP. Estimation of body fat mass using dual-energy x-ray absorptiometry, bioelectric impedance analysis, and anthropometry in HIV-positive male subjects receiving highly active antiretroviral therapy. JPEN J Parenter Enteral Nutr. 2007 Mar-Apr;31(2):135-41.

126. van Marken Lichtenbelt WD, Westerterp KR, Wouters L, Luijendijk SC. Validation of bioelectrical-impedance measurements as a method to estimate body-water compartments. Am J Clin Nutr. 1994 Aug;60(2):159-66.

127. Lukaski HC. Regional bioelectrical impedance analysis: applications in health and medicine. Acta Diabetol. 2003 Oct;40 Suppl 1:S196-9.

128. Foster KR, Lukaski HC. Whole-body impedance--what does it measure? Am J Clin Nutr. 1996 Sep;64(3 Suppl):388S-396S.

129. Foley K, Keegan M, Campbell I, Murby B, Hancox D, Pollard B. Use of single-frequency bioimpedance at 50 kHz to estimate total body water in patients with multiple organ failure and fluid overload. Crit Care Med. 1999 Aug;27(8):1472-7.

130. Pencharz PB, Azcue M. Use of bioelectrical impedance analysis measurements in the clinical management of malnutrition. Am J Clin Nutr. 1996 Sep;64(3 Suppl):485S-488S.

131. Lukaski HC, Johnson PE, Bolonchuk WW, Lykken GI. Assessment of fat-free mass using bioelectrical impedance measurements of the human body. Am J Clin Nutr. 1985 Apr;41(4):810-7.

132. Barbosa-Silva MC, Barros AJ, Post CL, Waitzberg DL, Heymsfield SB. Can bioelectrical impedance analysis identify malnutrition in preoperative nutrition assessment? Nutrition. 2003 May;19(5):422-6.

133. Barbosa-Silva MC, Barros AJ. Bioelectrical impedance analysis in clinical practice: a new perspective on its

19

Cha

pte

r 1G

ener

al in

trod

ucti

on a

nd o

utlin

e of

the

thes

is

use beyond body composition equations. Curr Opin Clin Nutr Metab Care. 2005 May;8(3):311-7.134. Lukaski HC, Bolonchuk WW, Hall CB, Siders WA. Validation of tetrapolar bioelectrical impedance method to

assess human body composition. Appl Physiol. 1986 Apr;60(4):1327-32.

Chapter 2

Self-reporting of height and weight: valid and reliable identification of malnutrition in preoperative patients

E.B. HaverkortR. J. de Haan

J.M. BinnekadeM.A.E. van Bokhorst – de van der Schueren

Am J Surg. 2012 Jun; 203(6):700-7

22

Abstract

BackgroundPreoperative screening for malnutrition has become mandatory in The Netherlands. A sensitive method to diagnose malnutrition would save time and improve effectiveness.

MethodsA prospective cross-sectional study of 488 adult elective preoperative outpatients was performed. The accuracy of self-reported height and weight was compared with meas-ured data and three commonly used malnutrition screening tools. Interobserver agree-ment was calculated by the intraclass correlation coefficient, studied in Bland and Altman plots, and analyzed by using Cohen’s Ƙ statistic. Accuracy was expressed in sensitivity, specificity and false-negative rates.

ResultsDifferences between self-reported and measured data were significant, but clinically irrelevant, as only one patient was falsely identified well nourished. Intraclass correlation coefficient for height, weight and body mass index was high (.97-.99). Bland-Altman plots showed that the mean ± standard deviation differences 95% limits of agreement between both methods were as follows: height, .0096 m(± .0262, -.0417 to +0.0609 m); weight, -1.28 kg (± 2.29, -5.76 to +3.20 kg); body mass index -.72 kg/m² (± 1.11, -2.92 to +1.46 kg/m²). The Ƙ coefficient was .84 (95% confidence interval, .75-.94). Sensitivity was .97 and specificity was .98. Sensitivity and false negative rate of self-reported data were better overall com-pared to the screening tools.

ConclusionsSelf-reported data provide highly sensitive information to diagnose malnutrition in pre-operative outpatients.

Key words

Preoperative; Malnutrition; Surgery; Elective; Outpatient; Self-reporting; Height; Weight; Diagnostic accuracy.

Cha

pte

r 2Se

lf-re

por

ting

of h

eig

ht a

nd w

eig

ht: v

alid

and

relia

ble

iden

tific

atio

n of

mal

nutr

itio

n in

pre

oper

ativ

e p

atie

nts

23

Introduction

Every year, approximately 1.3 million patients in The Netherlands undergo a surgical pro-cedure, and approximately 75% of these procedures are elective. Although these surgeries vary in duration and intensity, and patients differ in age and health, all outpatients are required to undergo preoperative screening for malnutrition as specified by the National Dutch Guideline on Perioperative Nutrition. 1

In clinical practice, malnutrition is often defined as involuntary weight loss within a cer-tain time frame and/or a Body Mass Index (BMI) below a defined cut-off point. 2-8Although the prevalence rates of malnutrition for adults in outpatient clinics are low, ranging between 7% and 12%, early preoperative identification and treatment of malnutrition is considered essential because of the negative influence of malnutrition on postoperative outcome (eg, increased postoperative complications, higher mortality, increased length of hospital stay). 1, 9-16 Because of a change in policy, the (para) medical staffs in The Netherlands have been made responsible for screening all pre-operative patients for malnutrition in compliance with the National Dutch Guideline on Perioperative Nutrition. 1 Although they are convinced of the importance of such screening, this has created extra work for them, such as mak-ing systematic measurements of height and weight and calculating BMI and percentage of recent involuntary weight loss. Our observed, unpublished data show that these meas-urements require 2 to 3 minutes of additional time per patient. If healthcare professionals screened all preoperative outpatients, this would result nationwide in more than 32 thou-sand additional hours yearly - time that could be spent more effectively by these caregivers. To save time, malnutrition screening tools have been introduced to obtain a rough esti-mate of patients’ nutritional status. In The Netherlands, the Short Nutritional Assessment Questionnaire (SNAQ), 10, 17 the Mini Nutritional Assessment (MNA) for the elderly subpop-ulation, 18, 19 and the Malnutrition Universal Screening Tool (MUST) 20 frequently are used for adults. 21 However, these tools have not all been validated for outpatients, and at least some of the tools need trained personnel to use them. In addition, patients at risk for mal-nutrition need a complementary and more comprehensive assessment. 1, 22 It is thus very unlikely that these instruments will be applied for the purpose of preoperative screening. If we can shorten the malnutrition screening time required, improve the sensitivity of a malnutrition screening tool and thus increase the effectiveness, then the (para) medi-cal staff can focus on their core work of preparing preoperative patients while still com-plying with the National Dutch Guideline on Perioperative Nutrition. 1 Self-reported data on height and weight are a possibility for reducing this screening time. According to our knowledge, no studies have yet determined the adequacy of self-reported data to screen for malnutrition in preoperative surgical outpatients. In general, the literature indicates that healthy persons and patients suffering from eating disorders tend to under-report their body weight and to over-report their height, which results in an underestimated BMI. 23-27 In underweight patients this relationship

24

seems to be inverted; patients over-report their body weight, which leads to an underesti-mation of malnutrition based on weight loss. 24

In this study we addressed the following research questions: (1) what is the sensitivity of self-reported height and weight by adult preoperative patients compared with objective anthropometric data assessed by healthcare professionals, and how many malnourished patients are missed by using self-reported data? (2) How does self-reported anthropomet-ric data perform relative to three malnutrition screening instruments that frequently are used in the Netherlands: the SNAQ, MNA and the MUST? The hypothesis of our study was that self-reporting of height and weight is an effective method to diagnose malnutrition in preoperative elective patients with a better predictive value than the commonly used screenings tools.

Methods

Patients The study was designed as a prospective cross-sectional study involving adult preopera-tive patients scheduled for elective surgery who were visiting the Pre-Operative Screening Department of the VU University Medical Center (Amsterdam, The Netherlands) from March till June 2008. This study was approved by the institutional Medical Ethics Committee. Patients were invited to participate and received verbal and written information about the study before visiting the anesthesiologist.

Study population The study population initially consisted of all consecutive patients age 18 years and older (no age limitation) who visited the Pre-Operative Screening Department. From this group we collected patient-, clinical- and surgery-related characteristics and height and weight. Exclusion criteria comprised: patients suffering from cardiac failure, kidney disease, or liver failure. Because of a decrease in osmotic pressure these conditions can result in edema and fluid disturbances and affect a reliable measurement of body weight. Pregnant women and patients who were unable to undergo the physical examinations also were excluded.

Patient self-reported dataFor this study a questionnaire was developed, consisting of four components: (1) socio-demographic characteristics (age, sex, living, and social situation); (2) clinical character-istics (comorbidity, indication for surgery, referring department); (3) anthropometric data (height, usual and present body weight, voluntary and involuntary changes in body weight in the last month and over the last 6 months); and (4) the questions from the three most frequently used malnutrition screening tools in The Netherlands: SNAQ, MNA (for patients ≥ 65y), and MUST. 17-21 Patients filled out the questionnaire while in the waiting room at the outpatients clinic.

Cha

pte

r 2Se

lf-re

por

ting

of h

eig

ht a

nd w

eig

ht: v

alid

and

relia

ble

iden

tific

atio

n of

mal

nutr

itio

n in

pre

oper

ativ

e p

atie

nts

25

Objective clinical assessment Height in meters and body weight in kilograms, without shoes and in light indoor cloth-ing, were measured by a healthcare professional (E.B.H., research dietician) after the con-sultation of the patient with the anesthesiologist. The Seca stadiometer 222 was used for measuring height; weight was measured with the Seca 888 (Seca GMBH, Hamburg, Germany). The BMI was calculated by dividing body weight by the square of body height (kg/m²). Patients were asked to recall their weight at one and six months before to the study. In case of weight loss, patients were asked whether they lost weight voluntary or involuntary. Only in cases of involuntary weight loss was the percentage of weight loss is calculated. Objective definition of malnutrition For patients younger than age 65, malnutrition was defined in accordance with the National Dutch Guideline on Perioperative Nutrition: (1) involuntary weight loss of 5% or greater within 1 month; and/or (2) involuntary weight loss of 10% or greater within 6 months; and/or (3) a BMI less than 18.5. In accordance with hospital clinical guidelines, a BMI limit of 20.0 was used for patients age 65 and older. 1-3, 8

Aspects influencing malnutrition Information about age, sex, living situation, ethnic origin (Dutch or non-Dutch), family or friends to support and help, ability to prepare food, household composition, comorbidity, indication for the surgical procedure, and referring department were collected. Such fac-tors could influence the development or explain the presence of malnutrition.

Presence of malnutrition If a patient was identified as malnourished after clinical assessment, a protein-rich and energy-rich diet was advised (at least until the surgical procedure). If necessary, sip feed-ing (a complete enteral formula to be taken as an oral beverage rich in energy, proteins and micronutrients) was prescribed according to the hospital standard guidelines.

Statistical analysis Patient and clinical characteristics, including anthropometric data, were summarized using descriptive statistics. Differences between mean scores were analyzed using the 2-group dependent t test, when appropriate. The Kolmogorov-Smirnov test was used to determine whether the anthropometric data were distributed normally. Interobserver agreement between self-reported data and clinical assessment in rela-tion to the anthropometric indicators (height, body weight, BMI) was assessed by calculat-ing the intraclass correlation coefficient. In addition, the interobserver agreements were studied in scatter plots as described by Bland and Altman. 28 These plots show the differ-ences of an anthropometric indicator between both methods for each patient on the ver-tical y-axis against their mean on the horizontal x-axis. To describe the possible range of

26

differences, we also calculated the 95% limits of agreement (mean difference ± 1.96 times the standard deviation of the difference scores). Interobserver agreement regarding patients’ nutritional status was analyzed using Cohen’s Kappa Ƙ statistic. Finally, we determined the criterion validity of the self-reported data on mal-nutrition and the screening instruments SNAQ, MNA (in patients ≥ 65 y), and MUST relative to the objective definition of malnutrition, in terms of sensitivity, specificity, positive predictive value, negative predictive value, false-positive rates and false-negative rates. Statistical uncertainty was expressed in 95% confidence intervals (95% CIs). A P value less than .05 was considered statistically significant. All analyses were performed in SPSS 16.0 (SPSS Corp. Chicago, IL).

Figure 1 Study flow chart

70 patients declined to participate

22 patients had an incomplete questionnaire

115 patients did not undergo a clinical assessment due to lack

of time

563 patients

678 patients

700 patients were included

770 patients visited the Preoperative Screening (POS)

outpatient clinic between March and June 2008

75 patients had an incomplete clinical assessment

488 patients had a complete data set (both self-reported

and clinical assessment)

Figure 1 Study flow chart

Cha

pte

r 2Se

lf-re

por

ting

of h

eig

ht a

nd w

eig

ht: v

alid

and

relia

ble

iden

tific

atio

n of

mal

nutr

itio

n in

pre

oper

ativ

e p

atie

nts

27

Results

We identified 770 eligible patients. Seventy of them declined to participate, and the data from 212 patients (212 of 700; 30%) could not be analyzed for a number of reasons (Figure 1). In total, 488 patients were included. Except for household composition, the characteristics of participants and patients excluded owing to lack of time due to other medical appointments; excluded patients lived alone less often (P = .01). Patients incapable of undergoing clinical assessment dif-fered from participants with regard to living situation, household composition, sex, refer-ring department, and ability to prepare food (P ≤ .007); excluded patients lived indepen-dently more often, lived with a partner (and children) more frequently, were pregnant women referred from the Obstetrics Department, and were patients referred from the Traumatology Department. The excluded trauma patients less often were capable of pre-paring their own food. Table 1 describes in detail the sociodemographic and clinical characteristics of the included patients. The mean age (± standard deviation) of the patients was 51 years (± 17), the age range 18 to 91 years, and 48% of the patients were men. A neoplasm was the main indication for the surgical procedure. The Otorhinolaryngology and Surgery Departments referred the most patients (30% of all patients). As shown in Table 2, differences between self-reported data and clinically assessed height (.01 m), weight (1.3 kg) and BMI (.7 kg/m²) were small, albeit statistically significant (P < .001). The intraclass correlation for height, weight and BMI ranged from .97 to .99. Figure 2 shows the Bland-Altman plots of the differences between height, weight and BMI scores based on self-reported data and clinical assessment. Mean (± SD) differences and the 95% limits of agreement between the self-reported method and the clinical assessment were as follows: height, .0096 m (± .0262), 95% limits of agreement -.0417 to +.0609 m; weight, -1.28 kg (± 2.29), 95% limits of agreement -5.76 to +3.20 kg; and BMI .72 kg/m² (± 1.11), 95% limits of agreement -2.92 to +1.46 kg/m². By using our objective definition of malnutrition, 38 patients (8%) were identified as mal-nourished based on self-reported data compared with 30 patients (6%) based on clini-cal assessment (Table 3). The difference was mainly caused by the calculated BMI below the normal range; self-reporting identified 18 patients with a low BMI compared with 10 patients identified by clinical assessment. One patient was identified falsely as well-nourished by self-reported data, whereas in fact the patient was malnourished (false-negative), and nine patients (<2%) were wrongly identified as malnourished (false-positive). The interobserver agreement (Ƙ) was .84 (95% CI; .75 - .94). Compared with the objective definition, self-reports had a sensitivity of .97 (95% CI; .83 - .99) and a specificity of .98 (95% CI; .96 - .99). The positive predictive value was .76 (95% CI; .61 - .87) and the negative pre-dictive value .99 (95% CI; .99 - 1.0).

28

Table 1 Socio-demographic and clinical characteristics of the study population (N = 488)

Age, y * 51 (17)

Age range, y 18 - 91

Male sex 234 (48)

Top 4 living situation

Independent 444 (91)

Living with parents 28 (6)

Partially independent 9 (2)

Institution 4 (1)

Not of Dutch ethnic origin 65 (13)

Presence of family/friends for support and help 459 (94)

Ability to prepare food 450 (92)

Top 3 household compositions

Single 132 (27)

With a partner 181 (37)

Partner and child(ren) and/or others 165 (34)

Presence of co-morbidity 298 (61)

Top 5 indication for surgical procedure

ICD 140-239 neoplasm 105 (25)

ICD 710-739 diseases of the musculoskeletal system and connective tissue

47(11)

ICD 800-999 injuries and poisoning 46 (11)

ICD 360-389 diseases of the sense organs 41 (10)

ICD 580-629 diseases of the genitourinary system 36 (8)

Top 5 referring departments

Otorhinolaryngology 75 (15)

Surgery 67 (14)

Gynecology 51 (10)

Orthopedics 45 (9)

Plastic surgery 45 (9)

All values expressed in n (%) unless otherwise indicatedICD = International Statistical Classification of Diseases and Related Health Problems* Mean (± SD)

Cha

pte

r 2Se

lf-re

por

ting

of h

eig

ht a

nd w

eig

ht: v

alid

and

relia

ble

iden

tific

atio

n of

mal

nutr

itio

n in

pre

oper

ativ

e p

atie

nts

29

Table 2 Anthropometric characteristics of self-reports and clinical assessments of the study population (N = 488)

Self-reported Clinical assessment

ICC(95% confidence interval)

Height, m * 1.74 (0.10) 1.73 (0.10) ‡ 0.97 (0.96 – 0.97)

Body weight (kg), usually self-reported * 77.0 (16.8) Not applicable

Body weight (kg), present * 77.6 (16.1) 78.9 (16.1) ‡ 0.99 (0.98 – 0.99)

Self-reported weight loss ≥5% in 1 mo † 9 (2) Not applicable

Self-reported weight loss ≥10% in 6 mo † 16 (3) Not applicable

BMI calculated, kg/m² * 25.6 (4.7) 26.3 (4.8) ‡ 0.97 (0.97 – 0.98)

ICC = Intraclass Correlation Coefficient* Mean (± SD)† n (%)‡ Two-group dependent t test, P < .001

A B

C

Figure 2 Bland-Altman plots (N = 488). The difference between height, weight and BMI by self-report and clinical assessment for each person (y-axis) is plotted against the mean height, weight and BMI averaged from the 2 methods (x-axis). The hori-zontal solid line (y = 0) represents ideal agreement; the upper and lower dotted lines show the 95% limits of agreement.

30

There were significant differences between malnourished and well-nourished patients with regard to living situation, ability to prepare food, household composition, indication for surgical procedure and adjuvant therapy (P ≤ .01) (Table 4). When we compared the self-reported data with the results of the screening instru-ments SNAQ, MNA (for patients ≥ 65 y), and MUST, the diagnostic accuracy of self-reported

Table 3 Malnutrition characteristics of malnourished preoperative surgical outpatients based on self-reports (n=38) and clinical assessments (n =30)

Self-reported

n = 38

Clinical assessment

n =30

Malnutrition based on 1 characteristic

Loss of ≥5% body weight in 1 mo 4 4

Loss of ≥10% body weight in 6 mo 9 9

BMI below normal range * 18 10

Malnutrition based on 2 characteristics

Loss of ≥5% body weight in 1 mo and loss of ≥10% body weight in 6 mo 3 4

Loss of ≥10% body weight in 6 mo and BMI below the normal range 2 2

Malnutrition based on all 3 characteristics

Loss of ≥5% body weight in 1 mo and loss of ≥10% body weight in 6 mo and BMI below the normal range 2 1

Normal range BMI patients < 65 years, BMI >18.5; Patients ≥ 65 years, BMI > 20.0

Table 5 Diagnostic accuracy of self-reported anthropometric data (N = 488), SNAQ (N = 488), MNA in the elderly subpopulation (n = 113), and MUST (N = 488)

Self-report * SNAQ * MNA * † MUST *

Kappa .84 (.75-.94) .63 (.50 – .77) .26 (.07 – .46) .53 (.40 – .67)

Sensitivity .97 (.83- .99) .63 (.49 – .78) .89 (.62 – .98) .67 (.53 – .81)

Specificity .98 (.96-.99) .98 (.97 – .99) .73 (.65 – .80) .95 (.94 – .97)

Positive predictive value .76 (.61- .87) .68 (.53 – .82) .22 (.13 – .35) .49 (.36 – .62)

Negative predictive value .99 (.99 -1.0) .98 (.96 – .99) .99 (.94 – .99) .98 (.97 – .99)

False-positive rate .02 (.007 – .03) .02 (.007 – .03) .27 (.18 - .35) .05 (.03 - .06)

False-negative rate .03 (-.03 - .09) .37 (.20-.54) .11 (-.09 – .32) .33 (.16 - .50)

SNAQ = Short Nutritional Assessment QuestionnaireMNA = Mini Nutritional AssessmentMUST = Malnutrition Universal Screening Tool* Value (95% CI)† Patients ≥ 65 years of age (n = 113)

Cha

pte

r 2Se

lf-re

por

ting

of h

eig

ht a

nd w

eig

ht: v

alid

and

relia

ble

iden

tific

atio

n of

mal

nutr

itio

n in

pre

oper

ativ

e p

atie

nts

31

Table 4 Socio-demographic and clinical characteristics differences between well-nourished preoperative surgical outpatients (n = 458) and malnourished preoperative surgical outpatients (n=30) *

Well-nourished patients (n = 458)

Malnourished patients (n = 30)

P value

Age, y * 50 (17) 56 (16) .08

Male sex 221 (48) 13 (43) .71

Top 4 living situations .01 †

Independent 418 (91) 26 (87)

Living with parents 26 (6) 2 (7)

Partially independent 9 (2) -

Institution 2 (0.4) 2 (7)

Not of Dutch ethnic origin 60 (13) 5 (17) .98

Presence of family / friends to support and help 430 (94) 29 (97) .81

Ability to prepare food 425 (93) 25 (83) .01 †

Top 3 household composition .01 †

Single 121 (26) 11 (37)

With a partner 171 (37) 10 (33)

Partner and child(ren) and/or others 160 (35) 5 (17)

Presence of co-morbidity 280 (61) 18 (60) 1.0

Top 5 indication for surgical procedure .01 †

ICD 140-239 neoplasm 92 (20) 13 (43)

ICD 710-739 diseases of bone, muscle and connective tissue

47 (10) -

ICD 800-999 injuries and poisoning 44 (10) 2 (7)

ICD 360-389 diseases of the sense organs 40 (9) 1 (3)

ICD 580-629 diseases of the genitals 35 (8) 1 (3)

Top 5 referring departments .25

Otorhinolaryngology 66 (14) 9 (30)

Surgery 64 (14) 3 (10)

Gynecology 47 (10) 4 (13)

Orthopedics 41 (9) 4 (13)

Plastic surgery 44 (10) 1 (3)

Adjuvant therapy 16 (4) 7 (23) <.001 †

All values are expressed as n (%)ICD = International Statistical Classification of Diseases and Related Health ProblemsContinuous variables were tested with the independent t test. Nominal variables were tested by the chi-square test* Mean (± SD)† A P value of less than .05 was considered statistically significant

32

data overall proved to be better than the diagnostic accuracy of the screening instruments (Table 5).

Comments

This prospective cross-sectional study showed that self-reporting of height and weight is a highly sensitive alternative for malnutrition screening in adult preoperative patients. We observed a high level of agreement between self-reported anthropometric data and clinical assessment by a professional for height, weight, calculated BMI, and classification of nutritional status. In addition, self-reporting was found to be more sensitive than the generally applied malnutrition screening tools (SNAQ, MNA and MUST). Of our 488 patients, only one malnourished patient had a false-negative test result and was therefore at risk of not being treated in the preoperative stage. Nine patients (<2%) had a false-positive result. Based on the test results they would run the risk of being treated with a protein- and energy-enriched diet until the surgical procedure. However, this only results in a treatment that is excessive but considered harmless. Our results confirm the tendency of patients to overestimate their height, to underesti-mate their body weight and consequently to underestimate their BMI. 23-27 We found significant differences between self-reported height and weight and meas-ured data. Nevertheless, we consider these differences of no clinical importance because they only overestimated the number of malnourished patients (estimated body weight and estimated BMI were too low), and did not underestimate the prevalence of malnutri-tion. This study also highlights groups at special risk that need further attention: patients unaware of their height and/or weight, patients physically unable to undergo an examina-tion, and patients with an unreliably measured body weight. Our results therefore imply that a more thorough nutritional assessment is still necessary for patients with incomplete data: approximately 3% of the population. This subgroup was unable to provide adequate information about their present height and/or weight. Additional professional assess-ments therefore still will be required even if a systematic screening program is introduced based on self-reported data. Moreover, the nutritional status of about 11% of the patients could be estimated based only on self-reported data because their physical condition pre-vented a reliable examination (eg, patients who were unable to stand up, patients with a heavy plaster, patients with edema or fluid disturbances, and pregnant women). This is a clear advantage of self-reporting over objective assessment. Because malnutrition prevalence rates are low in outpatients, the additional task of performing systematic malnutrition screening of all elective surgical outpatients will increase the workload considerably, but without yielding a corresponding benefit in terms of detection of malnourished patients. In our study we found a prevalence of about 6% based on clinical assessment, which is comparable with previous studies. 10, 11 Use of self-reported height and weight for screening malnutrition could result in greater efficiency.

Cha

pte

r 2Se

lf-re

por

ting

of h

eig

ht a

nd w

eig

ht: v

alid

and

relia

ble

iden

tific

atio

n of

mal

nutr

itio

n in

pre

oper

ativ

e p

atie

nts

33

In our study, 30% of the eligible patients had to be excluded: 16% owing to lack of time, and 14% owing to incomplete data (3%) or an unreliable examination (11%). If self-report-ing is implemented in the future, this remaining 14% will need special attention. Still, we believe that the efficiency gain for the majority of the population is worthwhile, given the large number of patients undergoing preparation for surgery every year. There were significant differences between included and excluded patients regarding a number of sociodemographic and clinical characteristics. We initially assumed that only the inability to prepare food could have influenced the nutritional status. However, this aspect was reported by only five of the excluded patients, so we have no reason to assume that the prevalence of malnutrition would have been different in the excluded group com-pared with the participating patients. Although nutritional support was provided to patients at risk of malnutrition, this study did not evaluate the effects of nutritional intervention on postoperative outcome param-eters such as complications and mortality. It would be of interest to perform a follow-up study on whether self-reporting of height and weight has a better predictive value than other screening methods when it comes to outcome (reduced morbidity and decreased length of hospital stay). More research also is needed to validate the results in clinical and nonsurgical patients and in nontertiary hospitals. In conclusion, self-reporting of anthropometric data is a highly sensitive method to diagnose malnutrition among preoperative elective surgical outpatients and performs better than three of the most frequently used screening instruments SNAQ, MNA and MUST. Self-reporting of height and weight perhaps can be implemented directly at the Preoperative Outpatients Department, thereby reducing the workload and at the same time eliminating the problem of noncompliance with the National Dutch Guideline on Perioperative Nutrition.

Acknowledgements

The authors gratefully acknowledge the students Marloes Bakker, Marieke Berkhout, Marcella Martin and Leonie Roeleveld from the Department of Nutrition and Dietetics at the School of Sports and Nutrition in Amsterdam, The Netherlands, for their help and sup-port during the study.

34

References 1. Dutch Institute for Healthcare Improvement CBO, Utrecht, The Netherlands. National Dutch Guideline on

Perioperative Nutrition 2007. Available at: http://www.cbo.nl/Downloads/114/rl_periovoed_07.pdf. 2. Stratton RJ, Green CJ, Elia M. Scientific Criteria for Defining Malnutrition. In: Disease-related Malnutrition: An

Evidence-based Approach to Treatment, 1st ed. Wallingford: CABI Publishing; 2003:1-34. 3. Elia M, Ritz P, Stubbs RJ. Total energy expenditure in the elderly. Eur J Clin Nutr. 2000; 54 (suppl):S92-103. 4. van Bokhorst-de van der Schueren MA, van Leeuwen PA, Sauerwein HP, Kuik DJ, Snow GB, Quak JJ. Assessment

of malnutrition parameters in head and neck cancer and their relation to postoperative complications. Head Neck 1997;19:419-425.

5. Kelly IE, Tessier S, Cahill A, Morris SE, Crumley A, McLaughlin D, McKee RF, Lean ME. Still hungry in hospital: identifying malnutrition in acute hospital admissions. QJM 2000;93:93-98.

6. Perioperative total parenteral nutrition in surgical patients. The Veterans Affairs Total Parenteral Nutrition Cooperative Study Group. N Engl J Med 1991;325:525-532.

7. Detsky AS, Smalley PS, Chang J. The rational clinical examination. Is this patient malnourished? JAMA 1994;271:54-58.

8. Soeters PB, Reijven PL, van Bokhorst-de van der Schueren MA, Schols JM, Halfens RJ, Meijers JM, van Gemert WG. A rational approach to nutritional assessment. Clin Nutr 2008;27:706-716.

9. Fry DE, Pine M, Jones BL, Meimban RJ. Patient characteristics and the occurrence of never events. Arch Surg 2010;145:148-151.

10. Neelemaat F, Kruizenga HM, de Vet HC, Seidell JC, Butterman M, van Bokhorst-de van der Schueren MAE. Screening malnutrition in hospital outpatients. Can the SNAQ malnutrition screening tool also be applied to this population? Clin Nutr 2008;27:439-446.

11. Leistra E, Neelemaat F, Evers AM, Zandvoort van HWM, Weijs PJM, van Bokhorst-de van der Schueren MAE, Visser M, Kruizenga HM. Prevalence of undernutrition in Dutch hospital outpatients. Eur J Intern Med 2009; 20:509-513.

12. Wilson MMG, Vaswani S, Liu D, Morley JE, Miller DK. Prevalence and causes of undernutrition in medical outpatients. Am J med 1998;104:56-63.

13. Baldwin C, Weekes CE. Dietary advice for illness-related malnutrition in adults. Cochrane Database of Systematic Reviews 2008, Issue 1. Art. No.: CD002008. DOI: 10.1002/14651858.CD002008.pub3.

14. Naber THJ, Schermer T, Bree de A, Nusteling K, Eggink L, Kruimel JW, Bakkeren J, Heereveld van H, Katan MB. Prevalence of malnutrition in nonsurgical hospitalized patients and its associations with disease complications. Am J Clin Nutr 1997;66:1232-1239.

15. McWhirter JP, Pennington CR. Incidence and recognition of malnutrition in hospital. BMJ 1994;308:945-948. 16. Smedley F, Bowling T, James M, Stokes E, Goodger C, O’Connor O, Oldale C, Jones P, Silk D. Randomized

clinical trial of the effects of preoperative and postoperative oral nutritional supplements on clinical course and cost of care. Br J Surg 2004;91:983-990.

17. Kruizenga HM, Seidell JC, de Vet HC, Wierdsma NJ, van Bokhorst-de van der Schueren MA. Development and validation of a hospital screening tool for malnutrition: the short nutritional assessment questionnaire (SNAQ). Clin Nutr 2005;24:75-82.

18. Vellas B, Villars H, Abellan G, Soto ME, Rolland Y, Guigoz Y, Morley JE, Chumlea W, Salva A, Rubenstein LZ, Garry P. Overview of MNA® - Its History and Challenges. J Nut Health Aging 2006;10:456-465.

19. Guigoz Y. The Mini-Nutritional Assessment (MNA®) Review of the literature – What does it tell us? J Nutr Health Aging 2006;10:466-487.

20. Elia M. (editor) The ‘MUST’ report. A report by the Malnutrition Advisory Group of the British Association for Parenteral and Enteral Nutrition (BAPEN). Nutritional screening for adults: a multidisciplinary responsibility. Development and use of the‘ Malnutrition Universal Screening Tool’ (‘MUST’) for adults 2003.

21. Kondrup J, Allison SP, Elia M, Vellas B, Plauth M. Educational and Clinical Practice Committee, European Society of Parenteral and Enteral Nutrition (ESPEN). ESPEN guidelines for nutrition screening 2002. Clin Nutr 2003;22:415-421.

22. Davies M. Nutritional screening and assessment in cancer-associated malnutrition. Eur J Oncol Nurs 2005;9(suppl):S64-73.

23. Engstrom JL, Paterson SA, Doherty A, Trabulsi M, Speer KL. Accuracy of Self-Reported Height and Weight in Women: An Integrative Review of the Literature. J Midwifery Womens Health 2003;48:338-345.

24. Rowland ML. Self-reported weight and height. Am J Clin Nutr. 1990;52:1125-1133.25. Gunnell D, Berney L, Holland P, Maynard M, Blanc D, Frankel S, Davey Smith G. How accurately are height,

Cha

pte

r 2Se

lf-re

por

ting

of h

eig

ht a

nd w

eig

ht: v

alid

and

relia

ble

iden

tific

atio

n of

mal

nutr

itio

n in

pre

oper

ativ

e p

atie

nts

35

weight and leg length reports by the elderly, and how closely are they related to measurements recorded in childhood? International Journal of Epidemiology 2000;29:456-464.

26. Taylor AW, Dal Grande E, Gill TK, Chittleborough CR. How valid are self-reported height and weight? A comparison between CATI self-report and clinic measurements using a large cohort study. Aust N Z J Public Health 2006;30:238-46.

27. Spencer EA, Appleby PN, Davey GK, Key TJ. Validity of self-reported height and weight in 4808 EPIC-Oxford participants. Public Health Nutr 2002;5:561-565.

28. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307-310.

Chapter 3

Handgrip strength by dynamometry does not identify malnutrition in individual preoperative outpatients

E. B. Haverkort J.M. Binnekade

R. J. de HaanM.A.E. van Bokhorst – de van der Schueren

Clin Nutr. 2012 Oct; 31(5):647-51

Letter to the editor. Clin Nutr. 2012 Oct;31(5):778

Response from the authors. Clin Nutr. 2012 Oct;31(5):779-80.

38

Abstract

Background & aimsLow handgrip strength by dynamometry is associated with increased postoperative mor-bidity, higher mortality and reduced quality of life. The aim of this study was to evaluate the accuracy of four algorithms in diagnosing malnutrition by measuring handgrip strength.

MethodsWe included 504 consecutive preoperative outpatients. Reference standard for malnutri-tion was defined based on percentage involuntary weight loss and BMI. Diagnostic char-acteristics of the handgrip strength algorithms (Álvares-da-Silva, Klidjian, Matos, Webb) were expressed by sensitivity, specificity, positive and negative predictive value, false posi-tive and negative rate.

ResultsThe prevalence of malnutrition was 5.8%. Although Klidjian showed the highest sensitiv-ity (79%, 95%CI 62% - 90%), six out of 29 malnourished patients were falsely identified as wellnourished (false positive rate 21%, 95%CI 9% -38%). In contrast, this algorithm showed the lowest positive predictive value (8%, 95%CI 5% -13%). Matos presented the highest positive predictive value; the post-test probability increased to 13% (95%CI 8% – 20%). The 1-minus negative predictive value ranged between 3% and 5% for all algorithms.

ConclusionsNone of the algorithms derived from handgrip strength measurements was found to have a diagnostic accuracy good enough to introduce handgrip strength as a systematic insti-tutional screening tool to detect malnutrition in individual adult preoperative elective outpatients.

Key words

Surgery; Malnutrition; Handgrip strength dynamometry; Diagnostic accuracy.

39

Cha

pte

r 3H

and

grip

str

eng

th b

y d

ynam

omet

ry d

oes

not i

den

tify

mal

nutr

itio

n in

ind

ivid

ual p

reop

erat

ive

outp

atie

nts

Introduction

The prevalence of malnutrition in outpatients in The Netherlands ranges from 6% to 12%.1-3 Factors such as old age, diseases of the intestine, surgery for malignant disease and co-morbidity predispose to malnutrition.4, 5

As preoperative malnutrition is associated with postoperative morbidity, it is impor-tant to diagnose and treat surgical patients with malnutrition as early as possible. 6-10

Nevertheless, a simple and sensitive method for screening preoperative outpatients at risk of malnutrition is lacking.

Therefore we are looking for easy markers of nutritional status that may correlate well with poor nutritional status in adult preoperative surgical outpatients.

Earlier studies that have investigated handgrip strength (HGS) at group level found significant associations between low HGS and malnutrition, postoperative complications, prolonged hospital stay, reduced ability to return home, reduced mobility, impaired qual-ity of life and mortality. 11-18

A number of algorithms based on HGS are available to diagnose malnutrition 13, 19 or an increased risk for postoperative complications. 15, 20 Frequently used algorithms are the ones proposed by Álvares-da-Silva et al., 19 Klidjian et al., 15 Matos et al., 13 and Webb et al.. 20 Each of these algorithms uses its own cut-off points of normal values and some also correct for age and sex.20 Nonetheless, little is known about the screening abilities of these algorithms.

The objective of this study was to investigate the accuracy of the different HGS-based tests to diagnose malnutrition at the individual level.

Patients and Methods

Design and settingThis was a cross-sectional study among patients visiting the PreOperative Screening (POS) department of the VU University Medical Center, a tertiary care university-affiliated hospi-tal, Amsterdam, The Netherlands.

The study was approved by the local Medical Ethical Committee and qualified to improve patients‘ care and carried no extra risk of harm to the patients, making written informed consent unnecessary. Patients received verbal and written information about the purpose of the study and verbal consent was given before the start of the study.

Patients All consecutive patients between March and June 2008 ≥ 18 years of age who visited the POS department in order to be prepared for elective surgery were included in the study.

Exclusion criteria: patients who were pregnant and who had a disturbed fluid balance (e.g. edema). In addition, patients with cognitive impairment, neuromuscular diseases, hemiplegia, joint diseases and/or arthritis were also excluded since these diseases may influence HGS measurements.13, 14, 18

40

Baseline assessmentsPatient data (sex, age) and medical diagnosis (presence of malignancy, indication for the surgical procedure, presence of co-morbidity) were collected from medical records.

Definition and reference standard of malnutrition Worldwide, there is a lack of agreement regarding the definition of malnutrition.21 Malnutrition is frequently defined as a state resulting from lack of uptake or intake of nutri-tion leading to altered body composition (decreased fat free mass but specifically body cell mass) and diminished function. 22, 23 More recently, an adjusted definition has been proposed; a sub-acute or chronic state of nutrition in which a combination of varying degrees of over- or under-nutrition and inflammatory activity has led to a change in body composition and diminished function. 24, 25

We operationalized malnutrition following the ‘National Dutch Guideline Perioperative Nutrition’ by: (a) involuntary weight loss of ≥5% within 1 month; and/or (b) involuntary weight loss of ≥10% within 6 months; and/or (c) a BMI <18.5. 26 Patients were asked by a healthcare professional (EH) to recall their weight of six months and one month prior to the study assessments in order to calculate involuntary weight loss.

Present body weight in kilograms and height in centimeters, without shoes and in light indoor clothing, were measured by a healthcare professional (EH). Weight was measured with the Seca 888© (Seca GMBH, Hamburg, Germany); a Seca stadiometer 222© was used for measuring height. In case of involuntary weight loss, the percentage of weight loss was calculated.

In addition patients were requested to answer a questionnaire, consisting of four com-ponents: (a) socio-demographic characteristics; (b) clinical characteristics; (c) the ques-tions from the three most frequently used malnutrition screening tools in the Netherlands: SNAQ, MNA (for patients ≥ 65 years of age), and MUST; and (d) self-reported anthropomet-ric data (height, usual body weight, present body weight, body weight one and six months prior to the study).

Handgrip strength measurement by dynamometryAfter demonstrating the technique to the patient, the HGS measurements were carried out by the Jamar hydraulic dynamometer© (Sammons Preston Inc. Illinois USA) as it dem-onstrates the highest calibration accuracy.28 If necessary, the instrument was adjusted to the size of the patients’ hand. The patient was sitting in a chair, his upper arm by his side of the body and the forearm stretched to an angle of 90º, with the elbow unsupported. 27-30

The HGS by dynamometry was expressed in kilograms (round down) and carried out three times by both hands.

Brief pauses were taken between each measurement. The patient was encouraged to squeeze the dynamometer as hard as possible. For all four algorithms the highest of three measurements was recorded for both the right and the left hand.

41

Cha

pte

r 3H

and

grip

str

eng

th b

y d

ynam

omet

ry d

oes

not i

den

tify

mal

nutr

itio

n in

ind

ivid

ual p

reop

erat

ive

outp

atie

nts

Algorithms based on HGS to diagnose malnutritionFour different algorithms were applied to determine the absence or presence of malnu-trition by hand grip strength: the algorithms by Álvares-da-Silva et al., 19 Klidjian et al., 15 Matos et al., 13 and Webb et al. 20 (Appendix 1).

Reference values and control groupFor interpretation of the algorithms of Álvares-da-Silva et al., 19 Klidjian et al.15 we used the reference values of a sample of 62 healthy adult volunteers visiting the hospital, with absence of co-morbidity or impairment affecting the upper limb function. HGS dyna-mometry procedures in the control groups were identical as described for patients. For the algorithm of Matos the reference value was defined within the study population (HGS within the lowest quartile of the population). 13 In the algorithm of Webb we used the reference values as described in his article 20 (Appendix 1).

Appendix 1 Definition of malnutrition as defined by the four handgrip strength (HGS) algorithms

Algo-rithm ref

Definition of malnutrition

Hand to measure

Measurement(s) to report

Type of control group

Characteristics of control group

Álvares-da-Silva 19

HGS > 2 SD below the mean of the control group

Non-dominant hand

The best of 3 measurements

Control group healthy volun-teers

N = 62 volunteers (31 men and 31 women). Amsterdam, The Netherlands. Mean age 45 years (SD 13), range 20 - 73 years. HGS mean 44.6 kg (SD 11.6). Two SD below the mean was 21.4 kg.

Klidjian 15 HGS below 85% of the mean of the control group

Non-dominant hand

The best of 3 measurements

Control group healthy volun-teers

N = 62 volunteers (31 men and 31 women). Amsterdam, The Netherlands. Mean age 45 years (SD 13), range 20 - 73 years. HGS mean 44.6 kg (SD 11.6). 85% of the HGS below the mean was 37.9 kg.

Matos 13 HGS within the lowest quartile of the study population

Non-dominant hand

The best of 3 measurements

Patients are their own controls

Webb 20 HGS below 85% of the reference values for age and sex of the control group

The non-dominant or the dominant hand

The best of 3 measurements

Control group Webb

Non-dominant hand: N = 247 healthy volunteers (108 Men and 139 women). London, UK. Age range 16-95 years. HGS men mean 47.5 kg (SD 9.6), HGS women mean 29.6 kg (SD 9.8). Dominant hand of subgroup of these healthy volunteers: N = 119 (53 men and 66 women). HGS men mean 48.0 kg (SD 8.9) and HGS women mean 24.7 kg (SD 10.2).

42

Diagnostic accuracy The diagnostic accuracy was expressed in terms of sensitivity (the true-positives) and specificity (the true-negatives) rates. It was especially important to minimize the false negative rate (FNR), reflecting malnourished patients with HGS values within the normal range, as this implies missing and not treating a malnourished patient.

Additionally we calculated the predictive values of the algorithms in terms of positive predictive value (PPV) and 1-minus the negative predictive value (1-NPV). The PPV refers to the probability of being malnourished following a low HGS value, whereas the 1-NPV reflects the probability of being malnourished following an HGS value within the normal range.

Statistical analysisPatient characteristics were summarized using descriptive statistics. Analyses on diag-nostic accuracy were made on the basis of 2x2 tables, with malnutrition according to the reference standard and HGS below the normal value of the algorithm defined as present or absent.

For each HGS-algorithm the prevalence of malnutrition as well as the diagnostic accu-racy parameters were calculated. When appropriate, statistical uncertainty was expressed by the 95% confidence interval (95% CI). A P value of < 0.05 was considered significant. Data were analyzed with SPSS (version 16.0), STAT (version 10) and CIA (Confidence Interval Analysis).

Results

Study populationDuring the study period 655 eligible patients visited the POS department. Seventy-three patients declined to participate, 48 patients were excluded because of pregnancy and or extensive edema and 30 patients were excluded because of missing or inconsistent patient history data. In total, 504 patients were included. Forty-eight percent of the study population was male, the mean BMI of the population was 26.3 kg/m2 (SD 4.8), and for 17% of the patients a malignancy was the indication for the surgical procedure (Table 1).

The Dutch control group consisted of 31 men and 31 women, mean age 45 years (SD 13), range 20 - 73 years. The mean HGS of the non-dominant hand of the volunteers was 44.6 kg (SD 11.6). To define the cut-off values for the algorithms of Álvares-da-Silva we calculated two standard deviations below the mean (21.4 kg); for the algorithm of Klidjian cut-off values were determined by calculating 85% below the mean (37.9 kg) (Appendix 1).

We identified 29/504 patients (5.8%) as being malnourished according to the definition of the National Dutch Guideline Perioperative Nutrition. When applying the four different algorithms, between 8.5% (Álvares-da-Silva) and 60.1% (Klidjian) of the study population was classified as malnourished (Table 1).

43

Cha

pte

r 3H

and

grip

str

eng

th b

y d

ynam

omet

ry d

oes

not i

den

tify

mal

nutr

itio

n in

ind

ivid

ual p

reop

erat

ive

outp

atie

nts

The diagnostic accuracy of the four HGS algorithms Table 2 shows the diagnostic accuracy of the four algorithms compared to our pre-set definition of malnutrition. The sensitivity of Klidjians’ algorithm was the highest at 79% (95% CI 62% - 90%). Thus, its FNR (1-sensitivity) turned out to be the lowest: six out of 29 malnourished patients (21%, 95% CI 9% - 38%) were falsely identified ‘well-nourished’ while they were identified malnourished according to our reference standard (Table 3).

The algorithms of Álvares-da-Silva and Webb showed the lowest sensitivities 14% (95% CI 6% - 31%) and 35% (95% CI 20% - 53%) respectively. Hence, the false negative test results in these algorithms were substantial. The sensitivity rate of Matos’ algorithm was moderate (52%, 95% CI 34% - 69%).

As shown in Table 1 the prevalence or pre-test probability for being malnourished was 5.8% (29/504 patients). As expressed by the PPVs column in Table 2, the algorithm of Matos modestly increased the post-test probability of being malnourished following a poor HGS

Table 1 Patient characteristics and prevalence of malnutrition in relation to handgrip strength (HGS) algo-rithms (N=504)

n % SD

Male 241 48

Age in years 51 17

Age in years, range 20 - 91

Weight, kg 78.8 16

Height, m 1.73 0.10

BMI, calculated on weight/height 26.3 4.8

Indication for surgical procedure – top 5

Neoplasm 85 17

Diseases of bone, muscle and connective tissue 43 9

Injuries and poisoning 39 8

Diseases of the sense organs 32 6

Diseases of the genitals 33 7

Presence of co-morbidity 295 59

Malnutrition according to Dutch reference standard 29 5.8

Malnutrition according to four different algorithms

Álvares-da-Silva 43 8.5

Klidjian 303 60.1

Matos 118 23.4

Webb 84 16.7

44

value to a point estimate of 13% (95% CI 8% – 20%), directly followed by the PPV of Webb; 12% (95% CI 7% – 21%).

The 1-minus NPVs, expressing the probability of malnourished patients with an adequate HGS value, hardly decreased compared to the pre-test probability and ranged between 3% and 5% for all algorithms.

Discussion

This cross-sectional study among elective adult preoperative patients visiting the preop-erative screening department showed that none of the algorithms that we studied per-formed well enough to diagnose malnutrition by HGS sufficiently at the individual patient level. The approach of Klidjian resulted in the highest sensitivity rate compared to the

Table 2 Diagnostic accuracy of the four handgrip strength (HGS) algorithms

Algorithm Sensitivity (TPR) a

Specificity(TNR) b

PPV c 1-NPVd

Álvares-da-Silva 14 (6 - 31) 92 (89 - 94) 9 (4 – 22) 5 (4 – 8)

Klidjian 79 (62 - 90) 41 (37 - 46) 8 (5 – 11) 3 (1 – 6)

Matos 52 (34 - 69) 78 (74 - 82) 13 (8 – 20) 4 (2 – 6)

Webb 35 (20 - 53) 84 (81 - 87) 12 (7 – 21) 4 (3 – 7)

The prevalence of malnutrition based on the reference standard was 5.8%.

a Sensitivity = TPR (true positive rate) expressed in % (95% CI)b Specificity = TNR (true negative rate) expressed in % (95% CI)c PPV = positive predictive value expressed in % (95% CI)d 1-NPV = 1 - negative predictive value expressed in % (95% CI)

Table 3 Number of patients with positive and negative test results in relation to the presence of malnutrition, categorized per handgrip strength (HGS) algorithm

Algorithm Positive test result(Malnourished according to test)

Negative test result(Well-nourished according to test)

MalnourishedTP a

Well-nourishedFP b

MalnourishedFN c

Well-nourishedTN d

Álvares-da-Silva 4 39 25 436

Klidjian 23 280 6 195

Matos 15 103 14 372

Webb 10 74 19 401

a TP = number of true-positivesb FP = number of false-positives c FN = number of false-negativesd TN = number of true-negatives

45

Cha

pte

r 3H

and

grip

str

eng

th b

y d

ynam

omet

ry d

oes

not i

den

tify

mal

nutr

itio

n in

ind

ivid

ual p

reop

erat

ive

outp

atie

nts

other screening tools. Nevertheless, one-fifth of the malnourished patients was falsely identified as ‘well-nourished’ using this algorithm. In addition, compared to the pre-test probability of malnutrition, the post-test probability to be malnourished following a poor HGS value did not increase substantially with any of the algorithms.

Elderly and cancer patients are at increased risk to develop malnutrition.4, 5 To study whether there were differences between the total population and these high risk groups, we performed post-hoc analyses in these groups. The algorithm of Klidjian, in contrast to the other algorithms, performed very well in the subgroup of patients ≥65 years of age (n=118) with a sensitivity of 100% (95% CI 68% – 100%). We suggest a more thorough eval-uation of the added value of HGS in older preoperative patients’ in a future study. In cancer patients (n=85), the post-hoc analyses showed that the algorithms performed compara-ble to the total population.

We have tried to explain the poor diagnostic accuracy of the four algorithms. First of all, the lack of accuracy may be caused by unreliable measurements. We used the Jamar hydraulic dynamometer©, the instrument with the highest calibration accuracy, and the procedure was performed by one single clinical experienced researcher (EBH), according to a standardized protocol, as described by Mathiowetz et al.28 Therefore, we do not believe that the instrument or the caregiver significantly influenced the result.

A more plausible explanation for the poor accuracy may be found in the existing refer-ence values and methods. Can HGS values measured in relatively small groups of (healthy) adults living in the UK, USA, Portugal or Brazil be extrapolated to Dutch patients being worked-up for surgery? 13, 15, 19, 20, 27 In addition to other factors, HGS depends on the stature or constitution of a population. Autochthonic Dutch inhabitants are relatively tall and it is assumed that this may result in a basic increase of HGS. 18 Probably, HGS ratios (dividing the measured HGS in kg by the square of body height in meters) would be more valid, although reference values are not yet available.

As HGS is also age related, 13, 15, 19, 20, 27, 29 we evaluated whether age may have caused the lack of diagnostic accuracy to identify malnutrition in our population of patients. The age of our patients was comparable to the control groups provided by Webb et al., 20 thus not explaining the poor diagnostic accuracy by that algorithm. However, our healthy controls were, on average, 6 years younger than our patients. To study whether this may have influenced the results, we, post-hoc, applied the sex and age specific nor-mative data for handgrip strength described in a meta-analysis by Bohannon et al., 29 to the algorithms of Álvares-da-Silva et al. 19 and Klidjian et al. 15 instead of the reference values of our healthy volunteers. Results revealed even lower sensitivities and higher FNR for both algorithms and a PPV of 8% for Álvares-da-Silva and 11% for Klidjian.

HGS is often used as a predictor of postoperative complications, 15, 16, 20 but this was not the aim of the present study. However, in future we will study in greater detail the predic-tive value of HGS as well as the value of repeated HGS measurements on one person in relation to postoperative outcome.

46

This study has some limitations. Only 29 out of 504 patients were found to be mal-nourished according to the reference standards that we applied (involuntary weight loss and/ or a low BMI). 26 Therefore, conclusions drawn from a dataset with low prevalence rate must be interpreted with caution.

For the algorithms of Álvares-da-Silva et al. 19 and Klidjian et al. 15 we used our own con-trol group that consisted of 62 healthy volunteers; this small number of controls may be regarded as a limitation of the study.

The normative HGS data of the research group of Mathiowetz could not be used in this study. 27 Our written communication with the researcher makes clear that he never focused on identifying malnourished patients, but on helping occupational therapists and physiotherapists to identify patients who are in need of hand strengthening. For that purpose it has been suggested that a patient scoring more than 3 SD below the mean for their age and gender are likely to need hand strengthening; those scoring 2 SD below the mean might need hand strengthening. Since the algorithm of Mathiowetz et al. has not been designed to diagnose malnutrition, a cut-off value for malnutrition could not be recommended and the algorithm, in its present form, is inappropriate for identification of malnourished patients.

Finally, there seems to be a critical extent of body protein loss that must occur before decomposition of vital physiological functions (e.g. a perceptible reduction of muscle strength and muscle function) will appear. 10, 11, 15 Only 15/29 (52%) malnourished patients in our population had lost >10% of their body weight within 6 months before the study. This may explain why HGS was not always reduced and remained within the normal range for age and gender, although we identified those patients as malnourished.

If true prevalence rates are as low as in our study, one should question the effective-ness and efficacy of performing (time-consuming) measures such as HGS in this target population.

In a previous study we evaluated the diagnostic accuracy of self-reported anthro-pometry data (height and weight), and three well-known screenings tools (SNAQ, MNA (in the elderly subpopulation) and MUST) to diagnose malnutrition in the same popula-tion as the current study. The self-reported data showed a higher sensitivity and speci-ficity than any of the three applied screening tools. Results of this study, suggest that self-reported anthropometry data may be considered as a more accurate and easier alternative. 31

Conclusion

We showed that the algorithms of Álvares-da-Silva, Klidjian, Matos and Webb to diagnose malnutrition by measurement of HGS lack sufficient diagnostic accuracy. Our study results do not confirm HGS as an accurate alternative to systematically screen for malnutrition in elective preoperative outpatients.

47

Cha

pte

r 3H

and

grip

str

eng

th b

y d

ynam

omet

ry d

oes

not i

den

tify

mal

nutr

itio

n in

ind

ivid

ual p

reop

erat

ive

outp

atie

nts

Acknowledgements

The authors have no disclosure of interest regarding the article. The authors’ contributions; study conception and design: EH, MvB. Acquisition of data: EH. Analysis and interpretation of data: EH, RdH, JB. Drafting of manuscript: EH, RdH, JB, MvB. Critical revision: EH, RdH. All authors have actively contributed, read and approved the final manuscript.

We acknowledge Marloes Bakker, Marieke Berkhout, Marcella Martin and Leonie Roeleveld for their assistance during the study (at the time of the study students from the School of Sports and Nutrition (Department of Nutrition and Dietetics), Amsterdam, The Netherlands).

References 1. Neelemaat F, Kruizenga HM, de Vet HC, Seidell JC, Butterman M, van Bokhorst-de van der Schueren MA.

Screening malnutrition in hospital outpatients. Can the SNAQ malnutrition screening tool also be applied to this population? Clin Nutr 2008;27:439-46.

2. Leistra E, Neelemaat F, Evers AM, van Zandvoort MH, Weijs PJ, van Bokhorst-de van der Schueren MA, et al. Prevalence of undernutrition in Dutch hospital outpatients. Eur J Intern Med 2009;20:509-13.

3. Wilson MM, Vaswani S, Liu D, Morley JE, Miller DK. Prevalence and causes of undernutrition in medical outpatients. Am J Med 1998;104:56-63.

4. Bozzetti F, Gavazzi C, Miceli R, Rossi N, Mariani L, Cozzaglio L, et al. Perioperative total parenteral nutrition in malnourished, gastrointestinal cancer patients: a randomized, clinical trial. JPEN J Parenter Enteral Nutr 2000;24:7-14.

5. Halfens RJG, Meijers JMM, Du Moulin MFMT, Nie NC van, Neyens JCL, Schols JMGA. Landelijke prevalentie-meting zorgproblemen Universiteit Maastricht 2010. ISBN: 978 94 90411 02 2. Dutch. (National prevalence measurement of healthcare problems University of Maastricht 2010).

6. Baldwin C, Weekes CE. Dietary advice with or without oral nutritional supplements for disease-related malnutrition in adults (Review). Cochrane Database of Systematic Reviews 2011, Issue 9. Art. No.: CD002008.

7. Naber TH, Schermer T, de Bree A, Nusteling K, Eggink L, Kruimel JW, et al. Prevalence of malnutrition in non-surgical hospitalized patients and its association with disease complications. Am J Clin Nutr 1997;66:1232-9.

8. McWhirter JP, Pennington CR. Incidence and recognition of malnutrition in hospital. BMJ 1994;308:945-8.9. Smedley F, Bowling T, James M, Stokes E, Goodger C, O’Connor O, et al. Randomized clinical trial of the effects

of preoperative and postoperative oral nutritional supplements on clinical course and cost of care. Br J Surg 2004;91:983-90.

10. Windsor JA, Hill GL. Weight loss with physiologic impairment. A basic indicator of surgical risk. Ann Surg 1988;207:290-6.

11. Windsor JA, Hill GL. Grip strength: a measure of the proportion of protein loss in surgical patients. Br J Surg 1988;75:880-2.

12. Watters DA, Haffejee AA, Angorn IB, Duffy KJ. Nutritional assessment by hand grip dynamometry. S Afr Med J 1985;68:585-7.

13. Matos LC, Tavares MM, Amaral TF. Handgrip strength as a hospital admission nutritional risk screening method. Eur J Clin Nutr 2007;61:1128-35.

14. Norman K, Schütz T, Kemps M, Josef Lübke H, Lochs H, Pirlich M. The Subjective Global Assessment reliably identifies malnutrition-related muscle dysfunction. Clin Nutr 2005;24:143-50.

15. Klidjian AM, Foster KJ, Kammerling RM, Cooper A, Karran SJ. Relation of anthropometric and dynamometric variables to serious postoperative complications. Br Med J 1980;281:899-901.

16. Guo CB, Zhang W, Ma DQ, Zhang KH, Huang JQ. Hand grip strength: an indicator of nutritional state and the mix of postoperative complications in patients with oral and maxillofacial cancers. Br J Oral Maxillofac Surg 1996;34:325-7.

17. Corish CA. Pre-operative nutritional assessment. Proc Nutr Soc 1999;58:821-9.18. Jakobsen LH, Rask IK, Kondrup J. Validation of handgrip strength and endurance as a measure of physical

48

function and quality of life in healthy subjects and patients. Nutrition 2010;26:542-50.19. Álvares-da-Silva MR, Reverbel da Silveira T. Comparison between handgrip strength, subjective global

assessment, and prognostic nutritional index in assessing malnutrition and predicting clinical outcome in cirrhotic outpatients. Nutrition 2005;21:113-7.

20. Webb AR, Newman LA, Taylor M, Keogh JB. Hand grip dynamometry as a predictor of postoperative complications reappraisal using age standardized grip strengths. JPEN J Parenter Enteral Nutr 1989;13:30-3.

21. Meijers JM, van Bokhorst-de van der Schueren MA, Schols JM, Soeters PB, Halfens RJ. Defining malnutrition. Mission or mission impossible? Nutrition 2010;26:432-40.

22. Lochs H, Allison SP, Meier R, Pirlich M, Kondrup J, Schneider S, et al. Introductory to the ESPEN guidelines on enteral nutrition: terminology, definitions and general topics. Clin Nutr 2006;25:180-6.

23. Stratton RJEM, Green CJ. Scientific criteria for defining malnutrition. In: Disease- related malnutrition, 1st edn. Wallingford, UK: CABI Publishing, 2003: 1-22.

24. Soeters PB, Reijven PL, van Bokhorst-de van der Schueren MA, Schols JM, Halfens RJ, Meijers JM, et al. A rational approach to nutritional assessment. Clin Nutr 2008;27:706-16.

25. Norman K, Pichard C, Lochs H, Pirlich M. Prognostic impact of disease-related malnutrition. Clin Nutr 2008;27:5-15.

26. CBO Guideline Perioperative Nutrition. Dutch Institute for Healthcare Improvement. Utrecht, The Netherlands, 2007 (Accessed 2007, at www.cbo.nl/Downloads/114/rl_periovoed_07.pdf).

27. Mathiowetz V, Kashman N, Volland G, Weber K, Dowe M, Rogers S. Grip and pinch strength: normative data for adults. Arch Phys Med Rehabil 1985;66:69-74.

28. Mathiowetz V, Weber K, Volland G, Kashman N. Reliability and validity of grip and pinch strength evaluations. J Hand Surg Am 1984;9:222-6.

29. Bohannon RW, Peolsson A, Massy-Westropp N, Desrosiers J, Bear-Lehman JB. Reference values for adult grip strength measured with a Jamar dynamometer: a descriptive meta-analysis. Physiotherapy 2006;92:11–5.

30. Vaz M, Thangam S, Prabhu A, Shetty PS. Maximal voluntary contraction as a functional indicator of adult chronic undernutrition. Br J Nutr 1996;76:9-15.

31. Haverkort EB, de Haan RJ, Binnekade JM, van Bokhorst-de van der Schueren MA. Self-reporting of height and weight: valid and reliable identification of malnutrition in preoperative patients. Am J Surg. 2012 Jun;203(6):700-7.

49

Cha

pte

r 3H

and

grip

str

eng

th b

y d

ynam

omet

ry d

oes

not i

den

tify

mal

nutr

itio

n in

ind

ivid

ual p

reop

erat

ive

outp

atie

nts

Letter to the Editor

Sir, Haverkort et al. 1 investigated the accuracy of handgrip strength (HGS) in diagnosing undernutrition. They indicated that “algorithms based on HGS are available to diagnose malnutrition” and they studied their diagnostic accuracy using previously published cut-offs for estimates of sensitivity and specificity. One of these used cut-offs corresponds to the 25th percentile of HGS distribution values of a hospitalized sample evaluated by our research group.2 The used cut-off was only a way to overcome the absence of reference HGS values to screen hospital undernutrition 2 and not a systematic instruction to diag-nose this condition or an “algorithm”, according to Haverkort et al. terminology. 1

Indeed, the screening ability of these cut-offs should a priori have a limited applica-tion to their sample of preoperative outpatients, whose HGS values are not shown, but are presumably higher.

The reference method used by Matos et al. 2 to test HGS was the “Nutritional Risk Screening 2002” which is composed by nutritional status dimensions that have previ-ously shown to have high validity in predicting nutritional risk. 3,4 Contrary, the tool used by Haverkort et al. 1 is based only on changes in body mass or BMI, which is expected to have a different association with HGS. Also, the sample used by Haverkort et al. 1 has a low frequency of undernourished patients, limiting its ability to evaluate the diagnostic value of HGS. Finally, the dynamometers used in these two studies were different, limiting the comparability of the HGS values. 5

For these reasons we consider that this design is not appropriate to answer the rel-evant question whether HGS identifies undernutrition in preoperative outpatients and we strongly encourage the authors to test it against their own distribution values.

T.F. AmaralJ. MendesFaculdade De Ciências Da Nutrição E Alimentação Da, Universidade Do Porto, Rua Dr Roberto Frias, 4200-465 Porto, Portugal

References1. Haverkort EB, Binnekade JM, de Haan RJ, van Bokhorst-de van der Schueren MA. Handgrip strength by

dynamometry does not identify malnutrition in individual preoperative outpatients. Clin Nutr. 2012; 31(5):647-51

2. Matos LC, Tavares MM, Amaral TF. Handgrip strength as a hospital admission nutritional risk screening method. Eur J Clin Nutr 2007;61:1128–35.

3. Kondrup J, Allison S, Elia M, Vellas B, Plauth M. ESPEN guidelines for nutrition screening 2002. Clin Nutr 2003;22:415–21.

4. Rasmussen HH, Holst M, Kondrup J. Measuring nutritional risk in hospitals. Clin Epidemiol 2010;21:209–16.5. Guerra RS, Amaral TF. Comparison of hand dynamometers in elderly people. J Nutr Health Aging 2009;13:907–

12.

50

Response from the authors

Handgrip strength reconsidered: Continuous poor accuracy to diagnose malnutri-tion in preoperative outpatients

Sir, We thank Dr. Mendes and Professor Dr. Amaral for their comments regarding our recent paper on the accuracy of handgrip strength (HGS) to diagnose malnutrition in preopera-tive outpatients.1 We hope that they do not disagree with the central premise of our study: the necessity to diagnose and treat preoperative malnourished patients as early as pos-sible to avoid negative effects in the postoperative phase.

A simple and sensitive method for screening on malnutrition is lacking for (preoper-ative) outpatients. We therefore studied not only the accuracy of HGS, 1 but also that of self-reported anthropometric data and the screenings tools SNAQ, MNA (for patients ≥ 65 years of age) and MUST to diagnose malnutrition in a large heterogeneous group of pre-operative outpatients. 2

The remarks of Dr. Mendes and Professor Dr. Amaral relate to two aspects: 1. Failure to use the NRS -2002 as the gold standard to screen for malnutrition 2. The use of the HGS in the lowest quartile of the study population as a cut-off point to determine malnutrition.

We deliberately did not choose the NRS-2002 as the gold standard to define malnu-trition. The reason is that NRS-2002 has not been validated for the outpatient setting. Especially the item ‘severity of disease’ focuses typically on clinical patients, which hinders the applicability of the tool for use in outpatients. Moreover, in an extensive literature search we have not been able to identify any papers describing the validity of NRS-2002 for outpatients. Matos et al. described that the accuracy of HGS compared to NRS-2002 was the highest for the lowest quartile. Therefore, we decided to use the lower quartile as a cut-off point in our study.

In answer to the remarks of Dr. Mendes and Professor Dr. Amaral we performed a post-hoc analysis. In 419 patients we were able to collect reliable data with regard to our defi-nition of malnutrition (involuntary weight loss / low BMI), the four HGS algorithms, self-reported anthropometric data, the earlier mentioned screening tools and the NRS-2002.

The prevalence of malnutrition according to our reference method was 5.3% (22/419) in this group. According to the NRS-2002 this was 15.3% (64/419). When we applied the NRS-2002 instead of our reference method as the gold standard, this - remarkably - resulted in equal (poor) accuracy for the Matos equation: Sensitivity for both references 45%, specificity 80% and 83% respectively (Table 1). These new results show that the NRS-2002 does not positively affect the diagnostic accuracy of HGS in preoperative out-patients. As prevalence rates varied widely we also calculated the positive likelihood ratio (LR+). This also resulted in almost identical values (2.2 and 2.6 respectively).

We agree with the remark that the dynamometers used in these two studies were different. We used the Jamar hydraulic dynamometer© as this is the instrument with the

51

Cha

pte

r 3H

and

grip

str

eng

th b

y d

ynam

omet

ry d

oes

not i

den

tify

mal

nutr

itio

n in

ind

ivid

ual p

reop

erat

ive

outp

atie

nts

highest calibration accuracy according to Mathiowetz et al. 3

In conclusion: we endorse the view that there is no global definition of malnutrition and that there is uncertainty about the instrument that can determine malnutrition the best. We maintain our view that, regardless the reference method, HGS is no accurate method to determine preoperative malnutrition in a heterogeneous group of preopera-tive outpatients.

E.B. HaverkortDepartment of Nutrition and Dietetics, Internal Medicine, VU University Medical Center,Amsterdam, The Netherlands. Department of Dietetics, Academic Medical Center,University of Amsterdam, Amsterdam, The Netherlands.J.M. BinnekadeDepartment of Intensive Care, Academic Medical Center, University of Amsterdam,Amsterdam, The Netherlands.Marian A.E. van Bokhorst - de van der SchuerenDepartment of Nutrition and Dietetics, Internal Medicine, VU University Medical Center,Amsterdam, The Netherlands.

52

References1. Haverkort EB, Binnekade JM, de Haan RJ, van Bokhorst-de van der Schueren MA. Handgrip strength by dynamometry does not identify malnutrition in individual preoperative outpatients.

Clin Nutr. 2012 Feb 26. 2. Haverkort EB, de Haan RJ, Binnekade JM, van Bokhorst-de van der Schueren MA. Self-reporting of height and weight: valid and reliable identification of malnutrition in preoperative patients.

Am J Surg. 2012 Jun;203(6):700-707. 3. Mathiowetz V, Weber K, Volland G, Kashman N. Reliability and validity of grip and pinch strength evaluations.

J Hand Surg Am 1984;9:222-226.

Table 1 Diagnostic accuracy of the four handgrip strength (HGS) algorithms and self-reported anthropomet-ric data compared to two different reference methods (1) involuntary weight loss/ low BMI and (2) NRS-2002 (N=419)

Reference method

Sensitivity (TPR) a

in % (95% CI)Specificity (TNR) b

in % (95% CI)Positive likelihood

ratio (LR+) c

Álvares-da-Silva

WL/low BMI d 9(3 – 28)

92(89-94)

1.1

NRS-2002 14(8 – 25)

93(90 – 95)

2.0

Klidjian WL/low BMI d 77(57 – 90)

43(38-48)

1.4

NRS-2002 72(60 – 81)

45(39 – 50)

1.3

Matos WL/low BMI d 45(27 – 65)

80(75 – 83)

2.2

NRS-2002 45 (34 – 57)

83(78 – 86)

2.6

Webb WL/low BMI d 32(16 – 53)

85(81 – 88)

2.1

NRS-2002 27 (17 – 39)

86(82 – 89)

1.9

Self-report WL/low BMI d 95(78 – 99)

98(96 – 99)

42.1

NRS-2002 25(16-37)

96 (94 – 98)

6.3

Prevalence of malnutrition: 5.3% according to WL/low BMI (involuntary weight loss and of low BMI) and 15.3% according to NRS-2002.

a Sensitivity = TPR (true positive rate) expressed in % (95% CI)b Specificity = TNR (true negative rate) expressed in % (95% CI)c Positive likelihood ratio (LR+) = (a / [a + c]) / (b / [b + d])d Malnutrition was defined following the ‘National Dutch Guideline Perioperative Nutrition’ by: (a) involuntary weight loss of ≥5% within 1 month; and/or (b) involuntary weight loss of ≥10% within 6 months; and/or (c) a low BMI (BMI <18.5 for patients younger than 65 years and BMI < 20.0 for patients 65 years and older)

Chapter 4

Estimation of body composition depends on applied device in patients undergoing major abdominal surgery

E.B. HaverkortJ.M. Binnekade

M.A.E. van Bokhorst – de van der SchuerenD.J. Gouma

R.J. de Haan

Provisionally accepted by Nutrition in Clinical Practice

54

Abstract

Background/Objectives: Bioelectrical impedance analysis (BIA) is a method used to esti-mate body compartments such as fat free mass (FFM) and fat mass (FM).Two bioelectri-cal impedance analysis (BIA) devices, a single-frequency bioelectrical impedance analysis device (SF-BIA) and a bioimpedance spectroscopy device (BIS) were compared in order to evaluate their reliability and to study whether their estimations resulted in similar classifi-cations of body composition.

Subjects/Methods: In a prospective observational study, body composition was esti-mated by SF-BIA and BIS in 123 adult patients scheduled for major abdominal surgery. Measurement agreement for the continuous variables fat free mass (FFM) and fat mass (FM) were analyzed using intraclass correlation coefficient (ICC), the mean differences and their limits of agreement. Measurement differences were also visualized by Bland Altman plots. For the dichotomized fat free mass index (FFMI) and fat mass index (FMI), inter-observer agreement was calculated using Cohen’s Kappa (K) statistics; the McNemar test was performed to compare the paired proportions.

Results: Agreement for the continuous variables was ‘almost perfect’ for FM (.86, 95%CI .80 to 90), ‘substantial’ for FFM (.78, 95%CI .70 to .84). For the dichotomous variables, the agreement was ‘substantial’ for FMI (.67, 95%CI .51 to .83), ‘slight’ for FFMI (.19, 95%CI .01 to .37). BIS classified a larger proportion having a low FFMI and a high FMI.

Conclusions: Good ICCs between SF-BIA and BIS for FFM and FM. However, the mean dif-ferences were substantial, whereas the classification of body composition based on FFMI and FMI was influenced by the device. Therefore, BIA devices are not interchangeable.

Key words

Bio-electrical impedance analysis; Fat free mass; Fat free mass index; Fat mass; Fat mass index; Surgery.

55

Cha

pte

r 4Es

tim

atio

n of

bod

y co

mp

osit

ion

dep

end

s on

ap

plie

d d

evic

e in

pat

ient

s un

der

goi

ng m

ajor

ab

dom

inal

sur

ger

y

Introduction

Body composition plays an important role in the occurrence of postoperative complica-tions. 1 – 7 Substantial preoperative weight loss; 6, 8, 9 both high and low body mass index; 8 - 11 high fat mass - especially the presence of increased visceral fat -; 12, 13 low fat free mass; 7, 14 –

17 and sarcopenic obesity 7, 14 – 20 are potential risk factors for postoperative complications in terms of infections, leakage of the anastomoses, abscesses, re-operation, increased length of hospital stay, re-admission, and mortality.

However, measuring an altered body composition in terms of fat free mass (FFM) and fat mass (FM) is invasive, expensive and time consuming when using the recommended reference methods such as hydrostatic weighing, dual energy X-ray absorptiometry or magnetic resonance imaging. 21 – 26

Bioelectrical impedance analysis (BIA) has been described as a simple, easy and non-invasive method that claims to give a good estimate of body compartments such as FFM and FM during illness, recovery and treatment. 4, 21, 27 – 31Two types of devices frequently used in clinical practice are the single-frequency bioelectrical impedance analysis device (SF-BIA) and the bioimpedance spectroscopy device (BIS). As the devices are based on dif-ferent theories and techniques, it is unclear to what extent these devices demonstrate the same measurement outcomes.

In this reliability study, we investigated whether the used BIA devices influence the estimation of FFM and FM and the derived calculations fat free mass index (FFMI) and fat mass index (FMI) in patients undergoing major abdominal surgery. And if so, to what extent this affects the classification of the body composition within or outside the normal value range.

Methods

Design and settingThis study was designed as a prospective observational study among preoperative out-patients scheduled for major abdominal surgery at the Department of Surgery of the Academic Medical Center (Amsterdam, the Netherlands), a tertiary care hospital with 1000 beds specialized in the treatment of gastro-intestinal oncological diseases.

Participation Nutritional assessment, including bio-electrical impedance, is part of routine patient care in our hospital and carries no risk to harm patients. However, patients were asked per-mission for use of their data and permission of the local Medical Ethical Committee was obtained.

Patients Included were consecutive patients, aged 18 and older, who were scheduled for a curative

56

surgical procedure (esophagectomy with gastric tube reconstruction, pylorus preserving pancreatico duodenectomy, partial or total gastrectomy, hepatico-jejunostomy after iat-rogenic bile duct injury) or a non-curative procedure (gastro-enterostomy and or hepa-tico-jejunostomy, gastro-jejunostomy, partial or total resection of the pancreas) between June 2007 and June 2013.

Patients with a stent, metal material or metal prosthesis (eg, hart, hip, knee), an implanted pacemaker or implantable defibrillator-converter were excluded. Also excluded were patients with clinical presence of edema or fluid disturbances as this may result in overesti-mation or underestimation of the body compartments by bio-electrical impedance estima-tions. 27

Data collection Baseline characteristics (age, gender), physical status-related characteristics (presence of co-morbidity, presence of malignancy, neo-adjuvant chemo-radiation treatment, height, usual and present body weight, presence and severity of involuntary body weight loss) and surgery-related characteristics (American Society of Anesthesiologists [ASA] classi-fication defining the preoperative fitness, performed surgical procedure) were collected from medical and dietetic records.

Body mass index (BMI) and malnutrition BMI was calculated by dividing the present measured body weight by squared body height (m2). BMI classification according to the World Health Organization; BMI < 18.5 kg/m2 underweight; BMI 18.5 kg/m2 - 25 kg/m2 normal weight; BMI ≥ 25 kg/m2 overweight, BMI 25 - 30 kg/m2 pre-obese and BMI ≥ 30 kg/m2 obesity. Malnutrition was defined as BMI < 18.5 and /or body weight loss ≥ 10%.

BIA estimationsBIA is a method to estimate FFM (eg, including body cell mass, extracellular solids, extra-cel-lular water, and intra-cellular water) and FM, and is based on measuring the resistance and reactance of an alternating electrical current in the human body. Intracellular fluids, body fluids and electrolytes behave as electrical conductors (resistance) and cell membranes act as electrical condensers and are involved in capacitance (reactance). 4,5, 21 , 27 – 29, 31 - 33 To actu-ally estimate the body compartments, the measured resistance and /or reactance is /are incorporated into a statistical equation with other patient-related variables such as height, weight, gender and age. As we used the equations incorporated in the software of the SF-BIA and BIS device, details about these patient-related variables cannot be described as the manufacturers are not entirely transparent about the variables included.

BIA devices and measurement procedureTwo BIA devices estimating the entire body were used: a single-frequency BIA analysis (SF-BIA, BF-906, measurement at one frequency, 50 kHz, Maltron International Ltd. Essex,

57

Cha

pte

r 4Es

tim

atio

n of

bod

y co

mp

osit

ion

dep

end

s on

ap

plie

d d

evic

e in

pat

ient

s un

der

goi

ng m

ajor

ab

dom

inal

sur

ger

y

United Kingdom) and a bioimpedance spectroscopy (BIS, Body Scout, measurement of various frequencies between 5 kHz and 1 MHz, 5-800 µA; Fresenius Kabi AG, Bad Homburg, Germany).

The BIA measurements were performed following the standardized procedures by Lukaski. 33 To avoid electrode polarization and to minimize the effects of the impedance of skin beneath the electrodes, the four-electrode technique was used. 30 Both impedance measurements were performed with the same four electrodes. The first measurement was always performed with the SF-BIA device, the second, directly afterwards, with the BIS device. The complete output was directly and automatically transferred and saved on a chip card and was read out at a later time point. The room temperature, the movements, and (supine) position of the patients were identical during the two measurements. In case of doubt about the correctness of (one of ) the measurements, the measurement was repeated.

The two devices are based on different methods and techniques. The SF-BIA measures at one frequency, 50 kHz, and is therefore assumed to measure extracellular water and a part of the intracellular water. In contrast, BIS measures over a range of different frequen-cies, and is assumed, although not proven, to estimate intracellular water more accurate.

Total body water is calculated as the weighted sum of extracellular water plus intracel-lular water. The FFM compartment is derived from total body water as a constant hydra-tion of 73% of this compartment is assumed in healthy adults. According to this assump-tion, the FFM may not be estimated properly under conditions of a significantly altered hydration. The FM compartment can be calculated indirectly as the difference between body weight and FFM. 21, 22, 27 Normal values of FFM and FM according to Kyle et. al. are described in Appendix 1.1

Appendix 1 Normal values for fat free mass (index) and fat mass (index)

Normal value range

Item mean reference value ± SD within outside

Fat free mass, men 59.1 kg ± 5.6

Fat free mass, women 42.4 kg ± 4.3

Fat mass, men 15.0 kg ± 5.5

Fat mass, women 17.6 kg ± 6.1

Fat free mass index, men 19.1 kg/m2 ± 1.4 16.7 to 19.8 kg/m2 < 16.7 kg/m2

Fat free mass index, women 15.9 kg/m2 ± 1.3 14.6 to 16.8 kg/m2 < 14.6 kg/m2

Fat mass index, men 4.9 kg/m2 ± 1.8 1.8 to 5.2 kg/m2 > 5.2 kg/m2

Fat mass index, women 6.6 kg/m2 ± 2.4 3.9 to 8.2 kg /m2 > 8.2 kg /m2

BIA = Bioelectrical impedance analyses measured by 50-kHz SF-BIA in healthy white adultsReference: Kyle et. al. 1

58

Fat free mass index (FFMI) and fat mass index (FMI) In addition to FFM and FM, the fat free mass index (FFMI) and fat mass index (FMI) were calculated by dividing FFM (kg) and FM (kg), measured by bioelectrical impedance, by squared height (m2). These indexes were calculated in order to adjust for differences in height and also as a method to classify patients within or outside the normal range with regard to their FFMI and FMI (Appendix 1). 1

Statistical analysesContinuous patient characteristics were described by their mean and standard deviation. Categorical variables were expressed as n (%). The measurement agreements between the continuous variables FFM, FFMI, FM and FMI, were analyzed using the intraclass correla-tion coefficient (ICC). The ICC can be interpreted as follows: < 0 indicating no agreement, 0 – .20 slight agreement, .21 – .40 fair agreement, .41 – .60 moderate agreement, .61 – .80 substantial agreement, and .81 – 1.0 almost perfect agreement.34

Additionally, we calculated the mean differences between SF-BIA and BIS meas-urements, and described the likely range of differences in terms of the 95% limits of agreement (that is, the interval of 1.96 standard deviation of the measurement differences either side of the mean difference). The measurement differences were also visualized by Bland Altman (BA) plots.35 BA plots graph the difference between SF-BIA and BIS against the average values measured by SF-BIA and BIS. To further enhance the interpretation of the BA plots, histograms of the differences between SF-BIA and BIS were added.36

Finally, the continuous FFMI and FMI were dichotomized based on the predefined cutoff normal values to classify patients within or outside the normal range with regard to FFMI and FMI (Appendix 1). 1 The classification agreement was expressed in Kappa coefficients, which can be interpreted in the same way as the ICC. 37 To compare the paired proportions with regard to the dichotomized FFMI and FMI, the McNemar test was used.

Statistical uncertainty was expressed in 95% confidence intervals (95% CIs). A P value < .05 was considered statistical significant. All analyses were performed in SPSS 21.0 (SPSS corp. Chicago Illinois USA). BA plots were produced in R. 38

Results

A total of 123 consecutive preoperative patients were included into the study. Baseline characteristics are shown in Table 1. Co-morbidity was present in 76% of the patients, with cardiovascular disease accounting for the largest proportion (22%). The presence of a malignant process was the most common indication for surgery (89%). Eighty-five percent of the patients underwent a (potentially) curative surgical procedure.

Involuntary loss of body weight before diagnosis had occurred in 68% of the patients (mean loss 7.5 kg ± 6.7 kg); still the mean body mass index of the study group was 25.2 kg/m2 ± 3.8 kg/m2 indicating weight of the majority of patients within the normal range or

59

Cha

pte

r 4Es

tim

atio

n of

bod

y co

mp

osit

ion

dep

end

s on

ap

plie

d d

evic

e in

pat

ient

s un

der

goi

ng m

ajor

ab

dom

inal

sur

ger

y

higher. Fourteen patients could be defined obese (BMI ≥ 30 kg/m2) and 31 patients (25%) were malnourished.

Measurement agreement between SF-BIA and BIS - continuous data (FFM, FFMI, FM, FMI)The ICC between SF-BIA and BIS could be defined as ‘almost perfect’ for both FM and FMI (.86, 95% CI .80 to .90); ‘substantial’ for FFM (.78, 95% CI .70 to .84); and ‘moderate’ for FFMI (.55, 95% CI .41 to .66) (Table 2).

Table 1 Patient baseline characteristics (N = 123)

Age, y † 61 (11)

Age range, y 21- 84

male sex 77 (63)

Presence of co-morbidity 94 (76)

Presence of malignancy 110 (89)

ASA classification ≥ 3 17 (14)

Top 3 surgical procedures:

Potentially curative esophagectomy with gastric tube reconstruction 49 (40)

Potentially curative pylorus preserving pancreatico-duodenectomy 35 (29)

Non-curative gastro-enterostomy and/or hepaticjejunostomy (single or double bypass) 17 (14)

Potentially curative surgical procedure 104 (85)

Usual bodyweight kg † 83.1 (15.8)

Present body weight kg † 75.6 (13.9)

Involuntary weight loss 83 (68)

Mean weight loss kg † 7.5 (6.7)

Weight loss ≥ 10% 31 (25)

BMI † 25.2 (3.8)

BMI < 18.5 2 (2)

BMI ≥ 30 14 (11)

Identified malnourished 31 (25)

All values are expressed as n (%) unless other indicated† Mean (± SD)ASA = American Society of Anaesthesiologists [ASA] classification defining preoperative fitnessBMI = body mass index was calculated by dividing the present body weight by squared body height (m2)A BMI < 18.5 was defined as malnourished and a BMI ≥ 30.0 as obeseMalnutrition = was defined as >10% involuntary body weight loss and / or a BMI < 18.5

60

The mean differences between the two measurements are shown in Table 3. The mean difference was the largest for FFM: 4.93 kg (± 6.22 kg, 95% limits of agreement -7.26 kg to 17.12 kg) and the smallest for FMI: -1.46 kg/m2 (± 1.65 kg/m2, 95% limits of agreement -4.69 kg/m2 to 1.78 kg/m2). The BA plots are depicted in the Figures 1a to 1d (left side).

Classification agreement between SF-BIA and BIS – dichotomized data (FFMI, FMI)Classification agreement between the devices could be classified as ‘slight’ for FFMI (.19, 95% CI .01 to .37) and ‘substantial’ for FMI (.67, 95% CI .51 to .83) (Tables 4a and 4b). Compared to the SF-BIA device, the BIS device classified a larger proportion of the patients as having a body composition outside the normal range in terms of low FFMI and high FMI.

Table 2 Measurement agreement between SF-BIA and BIS to estimate body composition – continuous data (N=123)

SF-BIA BIS ICC (95% CI)

Fat free mass kg 53.3 (9.7) 48.4 (8.8) .78 (.70 – . 84)

Fat free mass index kg/m2 17.7 (2.5) 16.1 (2.2) .55 (.41 – .66)

Fat mass kg 23.8 (8.7) 28.2 (10.0) .86 (.80 – . 90)

Fat mass index kg/m2 8.0 (2.9) 9.5 (3.3) .86 (.80 – . 90)

All values are expressed as mean (± SD)SF-BIA = single-frequency bioelectrical impedance analysisBIS = bioimpedance spectroscopyICC = intraclass correlation coefficient and 95% confidence interval (95% CI)Fat free mass index kg/m2 = calculated by dividing fat free mass (kg) estimated by bioelectrical impedance analysis by squared height m2

Fat mass index kg/m2 = calculated by dividing fat mass (kg) estimated by bioelectrical impedance analysis by squared height m2

Table 3 Mean differences and 95% limits of agreement between SF-BIA and BIS (N=123)

Mean difference (± SD)

95% limits of agreement

Fat free mass kg 4.93 kg (6.22) -7.26 to 17.12

Fat free mass index kg/m2 1.66 (2.25) -2.75 to 6.06

Fat mass kg -4.40 (5.01) -14.23 to 5.42

Fat mass index kg/m2 -1.46 (1.65) -4.69 to 1.78

SF-BIA = single-frequency bioelectrical impedance analysisBIS = bio-impedance spectroscopy95% limits of agreement = mean difference ± 1.96 standard deviation of the difference between SF-BIA and BIS estimationFat free mass index kg/m2 = calculated by dividing fat free mass (kg) estimated by bioelectrical impedance analysis by squared height m2

Fat mass index kg/m2 = calculated by dividing fat mass (kg) estimated by bioelectrical impedance analysis by squared height m2

61

Cha

pte

r 4Es

tim

atio

n of

bod

y co

mp

osit

ion

dep

end

s on

ap

plie

d d

evic

e in

pat

ient

s un

der

goi

ng m

ajor

ab

dom

inal

sur

ger

y

Figure 1 Bland-Altman plots and absolute differences (N = 123)

Figure 1a Fat free mass (FFM) in kg

Figure 1b Fat free mass index (FFMI) in kg/m2

Figure 1c Fat mass (FM) in kg

62

Figure 1d Fat mass index (FMI) in kg/m2

Table 4a Classification agreement between SF-BIA and BIS – fat free mass index (FFMI)

SF-BIA

Outside normal range Within normal range

BISOutside normal range 15 43 58

Within normal range 5 60 65

20 103 123

Kappa coefficient (95% confidence interval): .19 (.01 to .37)

Table 4b Classification agreement between SF-BIA and BIS – fat mass index (FMI)

SF-BIA

Outside normal range Within normal range

BIS Outside normal range 86 14 100

Within normal range 1 22 23

87 36 123

Kappa coefficient (95% confidence interval): .67 (.51 to .83)

The Bland-Altman plots on the left side of the figures show the difference between the estimated FFM, FFMI, FM, and FMI by SF-BIA and BIS for each patient (y-axis) against the average value estimated by SF-BIA and BIS for each patient (x-axis). The horizontal solid line (y=0) represents perfect agreement; the upper and lower dotted lines show the 95% limits of agreement.

The right side of the figures show the absolute differ-ences for FFM, FFMI, FM, and FMI between SF-BIA and BIS.

FFMI = fat free mass index kg/m2. FFMI is calculated by dividing fat free mass (kg) estimated by bioelectrical impedance analysis by squared height m2

FMI = fat mass index kg/m2. FMI is calculated by di-viding fat mass (kg) estimated by bioelectrical imped-ance analysis by squared height m2

SF-BIA = single-frequency bioelectrical impedance analysisBIS = bioimpedance spectroscopy

63

Cha

pte

r 4Es

tim

atio

n of

bod

y co

mp

osit

ion

dep

end

s on

ap

plie

d d

evic

e in

pat

ient

s un

der

goi

ng m

ajor

ab

dom

inal

sur

ger

y

Sixteen percent (20 /123) of the patients demonstrated a low FFMI range if measured with a SF-BIA vs. 47% (58/123) by BIS (P < .001). A high FMI was demonstrated in 71% (87/123) of the patients according to SF-BIA vs. 81% (100/123) by BIS (P = .001).

Discussion

This prospective observational study among preoperative patients scheduled for major abdominal surgery, evaluates the reliability between the estimates of SF-BIA and BIS, two different BIA devices to determine body composition.

The ICCs between the measures showed substantial to almost perfect agreement for FFM and FM, respectively. However, in-depth analyses showed that the SF-BIA measure-ments exceeded those of the BIS with an average of about 5 kg for FFM. Consequently, the FM measurements by SF-BIA were on average about 4 kg lower than the BIS estimates. The intervals of the limits of agreement of FFM and FM were also relatively wide, indicating that large differences in measurements were observed.

Moreover, the classification of body composition based on the dichotomized FFMI and FMI, was influenced by the used device. Compared to the SF-BIA device, the BIS device classified a larger proportion of the population as having a body composition outside the normal range in terms of low FFMI and high FMI.

In recent years, it has increasingly been recommended not only to evaluate body weight and to calculate BMI, but also to measure the various body compartments in order to determine the nutritional status. BIA estimates and their derived calculations are often routinely used for hospital patients to evaluate (changes in) nutritional status and follow-ing therapeutic (surgical) interventions.39

This reliability study shows that the results derived from two different BIA devices are not simply interchangeable; the SF-BIA device and BIS device classify substantially other numbers of patients as having a body composition outside the normal range.

Given the fact that clinical decisions are made based on these dichotomized FFMI and FMI values, this can influence the choices of the caregiver. Therefore, an inadequate estimation of the FFMI is of great significance as underestimation may wrongly result in the start of physical therapy, and/or dietary therapy or even in the postponement of the scheduled surgical procedure. 7, 14, 19

The differences between SF-BIA and BIS with regard to the estimations of body compo-sition cannot be explained by circumstances in the examination room or patients’ related aspects (eg, movements, position) as the two measured were performed under the same circumstances within a very short time frame.

A minor weakness in this study was the fact that we did not use the SF-BIA device and BIS device in a random order. However, we do not assume this has distorted our results because the output generated by the devices was automatically transferred and could not be manipulated by the researcher who performed the measurements.

In general, it is assumed that an estimation performed by BIS is more reliable than

64

an estimation performed by SF-BIA. Theoretically, BIS estimates intracellular water more accurately, which subsequently should contribute to a better estimate of the FFM. 40 With regard to optimal care and efficiency it is necessary to have certainty which BIA device contributes to the most valid estimate of a specific body compartment in a population of preoperative patients. Based on the present data, we cannot conclude which of the two devices is superior to the other, as no validation was carried out comparing the devices to a reference method regarded the gold standard (hydrostatic weighing, air-displacement plethysmography, dual energy X-ray absorptiometry, magnetic resonance imaging, and computed tomography). 21, 22 – 26 According to our knowledge, no studies have been per-formed in patients undergoing major abdominal surgery measuring a certain body com-partment using a reference method assumed suitable and various BIA devices at the same point in time to evaluate which device contributes to the most valid estimation. In future we plan to perform BIA estimations and CT scans at the same time point in patients sched-uled for major surgery. These CT scans, primarily performed to establish a diagnosis and to follow disease progression, can also be used as reference method to evaluate FFM and FM. Based on these data, we will be able to conclude whether the SF-BIA device or the BIS device generates the most valid estimations in patients undergoing major abdominal surgery, or whether they are possibly both inaccurate.

ConclusionThis prospective observational study among preoperative patients scheduled for major abdominal surgery shows good ICCs between SF-BIA and BIS for FFM and FM. However, the mean differences between SF-BIA and BIS measurements were substantial, indicating that the two devices are not interchangeable. In addition, compared to SF-BIA, BIS clas-sified a larger proportion of patients as having a body composition outside the normal range. In order to achieve optimal care, more research is needed to determine which device obtains the most valid estimate for measuring body composition.

References 1. Kyle UG, Schutz Y, Dupertuis YM, Pichard C. Body composition interpretation. Contributions of the fat-free

mass index and the body fat mass index. Nutrition. 2003;19:597-604.2. Kyle UG, Genton L, Slosman DO, Pichard C. Fat-free and fat mass percentiles in 5225 healthy subjects aged 15

to 98 years. Nutrition. 2001;17:534-41.3. Schutz Y, Kyle UU, Pichard C. Fat-free mass index and fat mass index percentiles in Caucasians aged 18-98 y.

Int J Obes Relat Metab Disord. 2002;26: 953-60.4. Selberg O, Selberg D. Norms and correlates of bioimpedance phase angle in healthy human subjects,

hospitalized patients, and patients with liver cirrhosis. Eur J Appl Physiol. 2002; 86:509-16. 5. Barbosa-Silva MC, Barros AJ. Bioelectric impedance and individual characteristics as prognostic factors for

post-operative complications. Clin Nutr. 2005;24:830-8.6. Gupta R, Ihmaidat H. Nutritional effects of oesophageal, gastric and pancreatic carcinoma. Eur J Surg Oncol.

2003;29:634-43. 7. CBO Guideline Perioperative Nutrition. Dutch Institute for Healthcare Improvement. Utrecht, The Netherlands.

http://www.cbo.nl/Downloads/114/rl_periovoed_07.pdf [Accessed 2007].

65

Cha

pte

r 4Es

tim

atio

n of

bod

y co

mp

osit

ion

dep

end

s on

ap

plie

d d

evic

e in

pat

ient

s un

der

goi

ng m

ajor

ab

dom

inal

sur

ger

y

8. Antoun S, Rey A, Béal J et. al. Nutritional risk factors in planned oncologic surgery: what clinical and biological parameters should be routinely used? World J Surg. 2009;33:1633-40.

9. van Venrooij LM, de Vos R, Borgmeijer-Hoelen MM et. al. Preoperative unintended weight loss and low body mass index in relation to complications and length of stay after cardiac surgery. Am J Clin Nutr. 2008;87:1656-61.

10. Kanda M, Fujii T, Kodera Y et.al. Nutritional predictors of postoperative outcome in pancreatic cancer. Br J Surg. 2011;98:268-74.

11. Pausch T, Hartwig W, Hinz U et.al. Cachexia but not obesity worsens the postoperative outcome after pancreatoduodenectomy in pancreatic cancer. Surgery. 2012;152:S81-8.

12. Waisbren E, Rosen H, Bader AM et.al. Percent body fat and prediction of surgical site infection. J Am Coll Surg. 2010;210:381-9.

13. Shimizu A, Tani M, Kawai M et.al. Influence of visceral obesity for postoperative pulmonary complications after pancreaticoduodenectomy. J Gastrointest Surg. 2011;15:1401-10.

14. Wolfe RR. Is the double-blind randomized trial the most valid experimental approach to evaluating treatment modalities in critical ill patiënts? Curr Opin Clin Nutr Metab Care. 1998;1:185-187.

15. Pichard C, Kyle UG, Morabia A et.al. Nutritional assessment: lean body mass depletion at hospital admission is associated with an increased length of stay. Am J Clin Nutr. 2004;79:613-8.

16. Prado CM, Lieffers JR, McCargar LJ et.al. Prevalence and clinical implications of sarcopenic obesity in patients with solid tumours of the respiratory and gastrointestinal tracts: a population-based study. Lancet Oncol. 2008;9:629-35.

17. van Venrooij LM, de Vos R, Zijlstra E et.al. The impact of low preoperative fat-free body mass on infections and length of stay after cardiac surgery: a prospective cohort study. J Thorac Cardiovasc Surg. 2011;142:1263-9.

18. Visser M, van Venrooij LM, Vulperhorst L et.al. Sarcopenic obesity is associated with adverse clinical outcome after cardiac surgery. Nutr Metab Cardiovasc Dis. 2013;23:511-8.

19. Windsor JA, Hill GL. Weight loss with physiologic impairment. A basic indicator of surgical risk. Ann Surg 1988; 207:290-6.

20. Windsor JA, Hill GL. Grip strength: a measure of the proportion of protein loss in surgical patients. Br J Surg 1988;75:880-2.

21. Ellis K J. Human body composition: in vivo methods. Physiol Rev. 2000;80:649-80. 22. Jaffrin MY, Morel H. Body fluid volumes measurements by impedance: A review of bioimpedance spectroscopy

(BIS) and bioimpedance analysis (BIA) methods. Med Eng Phys. 2008;30:1257-69. 23. Mattsson S, Thomas BJ. Development of methods for body composition studies. Phys Med Biol. 2006;51:R203-

28. 24. Brodie D, Moscrip V, Hutcheon R. Body composition measurement: a review of hydrodensitometry, anthro-

pometry, and impedance methods. Nutrition. 1998;14:296-310. 25. Heymsfield SB, Wang Z, Visser M et. al. Techniques used in the measurement of body composition: an

overview with emphasis on bioelectrical impedance analysis. Am J Clin Nutr. 1996;64:478S-484S.26. Ellis KJ, Shypailo RJ, Pratt JA, Pond WG. Accuracy of dual-energy x-ray absorptiometry for body-composition

measurements in children. Am J Clin Nutr. 1994;60:660-5.27. Kyle UG, Bosaeus I, De Lorenzo AD et.al. Bioelectrical impedance analysis-part I: review of principles and

methods. Clin Nutr. 2004; 23:1226-43. 28. Kyle UG, Bosaeus I, De Lorenzo AD et.al. Bioelectrical impedance analysis-part II: utilization in clinical practice.

Clin Nutr. 2004;23:1430-53.29. De Lorenzo A, Andreoli A, Matthie J, Withers P. Predicting body cell mass with bioimpedance by using

theoretical methods: a technological review. J Appl Physiol. 1997;82:1542-58. 30. Foster KR, Lukaski HC. Whole-body impedance--what does it measure? Am J Clin Nutr. 1996;64:388S-396S. 31. Pencharz PB, Azcue M. Use of bioelectrical impedance analysis measurements in the clinical management of

malnutrition. Am J Clin Nutr. 1996;64:485S-488S.32. Heitmann B L. Prediction of body water and fat in adult Danes from measurement of electrical impedance. A

validation study. Int J Obes. 1990;14:789-802.33. Lukaski HC, Johnson PE, Bolonchuk WW, Lykken GI. Assessment of fat-free mass using bioelectrical impedance

measurements of the human body. Am J Clin Nutr. 1985;41:810-7.34. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33:159-74.35. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical

measurement. Lancet 1986;1:307-310. 36. Smith MW, Ma J, Stafford RS. Bar charts enhance Bland-Altman plots when value ranges are limited. J Clin

Epidemiol. 2010;63:180-4. 37. Altman DG. Practical statistics for medical research, London and New York: Chapman & Hall, 1991.

66

38. The R Development Core Team. The R Foundation for Statistical Computing, a language and environment for statistical computing. Vienna, Austria. http://www.R-project.org.

39. Dutch Malnutrition Steering Group, Amsterdam, The Netherlands. http:// www.stuurgroepondervoeding.nl [Accessed July 2008].

40. Buchholz AC, Bartok C, Schoeller DA. The validity of bioelectrical impedance models in clinical populations. Nutr Clin Pract. 2004;19:433-46.

Chapter 5

Bioelectrical impedance analysis to estimate body composition in surgical and oncological patients: a systematic review

E.B. HaverkortP.L.M. Reijven

J.M. Binnekade M.A.E. van Bokhorst – de van der Schueren

C.P. Earthman D.J. Gouma

R.J. de Haan

Submitted

68

Abstract

Bioelectrical impedance analysis (BIA) is a commonly used method for the evaluation of body composition. However, BIA estimations are subject to uncertainties. The aim of this systematic review was to explore the variability of regression equations used in the BIA estimations and to evaluate the validity of BIA estimations in adult surgical and oncological patients.

Included were studies developing new equations and studies evaluating the validity of BIA estimations compared with a reference method. Only studies using BIA devices meas-uring the entire body were included. Excluded were studies including patients with and altered body composition or a disturbed fluid balance, and studies written in languages other than English.

To illustrate variability between equations, fixed normal reference values were entered into the equations and the results plotted in figures. The validity was expressed by the difference in means between the BIA estimates and reference method, and relative differ-ence in %.

Substantial variability between equations was found for both total body water (newly developed equations up to 5 litres, existing equations up to 20 litres or kilograms), and fat free mass (over 25 kg). BIA mainly underestimated total body water (range relative difference -18.8% to +7.2%) and fat free mass (range relative differences -15.2% to +3.8%). Estimates of the fat mass demonstrated a large variability (range relative difference -15.7 % to +43.1%).

The absence of measurement precision precludes a valid estimate of a body compart-ment. We suggest that BIA estimations can only be useful when performed longitudinally and under strict conditions.

Key words

Surgery; Oncology; Bioelectrical impedance analysis; Variability; Validity; Systematic review.

69

Cha

pte

r 5Bi

oele

ctri

cal i

mp

edan

ce a

naly

sis

to e

stim

ate

bod

y co

mp

osit

ion

in s

urgi

cal a

nd o

ncol

ogic

al p

atie

nts:

a s

yste

mat

ic re

view

Introduction

In clinical practice bioelectrical impedance analysis (BIA) is routinely used in The Nether-lands to surgical and oncological patients where quick measurement of body compart-ments are needed, in terms of total body water, fat free mass, fat mass or other body compartments. BIA is considered to be an easy, non-invasive, less expensive and less time consuming method compared to other methods such as deuterium dilution; tritiated water dilution; bromide dilution; and dual energy X-ray absorptiometry (Appendix 1 at page 70-71). 1 -18

Reaching an estimate of a certain body compartment based on BIA estimation is achieved in two steps. In the first step a single or multiple frequencies BIA device meas-ures the value of resistance and reactance of the body at a certain frequency (Appendix 1). In the second step the measured resistance and/or reactance is incorporated into a statistically-derived regression equation considered most suitable to estimate the body compartments of interest. The equation usually consists of a set of population-related vari-ables such as height, body weight, age and gender and was originally derived from refer-ence data obtained in a specific population (e.g. specific disease state, ethnicity).

Although BIA is a commonly used method, 19 BIA estimations are subject to uncertain-ties. Firstly, a large number of equations with a variety of included variables are used to estimate a certain body compartment; secondly, different types of devices are based on different mathematical methods and techniques, and finally, the validity of the BIA estima-tion itself has not been unequivocally demonstrated.

The purpose of this systematic review was therefore to explore the variability of the equations used, and to investigate the validity of BIA estimations compared to a sound reference method in surgical and oncological patients.

Subjects and methods

For systematic reporting, the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement were followed. 20

Search strategy A systematic literature search was conducted in several electronic databases up to August 2012: Medline; the Cochrane Central Register of Controlled Trials (CENTRAL); EMBASE; the meta-search engine Sumsearch and CINAHL/Ebsco.

The following terms were used: electric impedance; body composition; surgery; oper-ation; oncology; and cancer. Details of the search strategy are described in Appendix 2 (page 72).

70

Appendix 2 Literature search - search terms and number of studies

Data base: PubMed

Impedance and surgery AND …. reference (136) - body composition (215) - reference AND body composition (20)

Impedance AND operation AND …. reference (105) - body composition (141) - reference AND body composition (12)

Impedance AND oncology AND …. reference (9) - body composition (23) - reference AND body composition (3)

Impedance AND cancer AND ….. reference (50) - body composition (149) - reference AND body composition (10)

(“electric impedance” [Mesh] OR electrical impedance OR *impedance) AND (“reproducibility of results” [MeSH] OR reference values“ [MeSH] OR “reference standards” [MeSH]) …. NOT (adipose OR fat) (723) – AND body (441). Limits: Humans, English, All Adult: 19+ years

Data base: Cochrane

Impedance AND surgery AND …. reference (6) - body composition (6) - reference AND body composition (0)

Impedance AND operation AND …. reference (1) - body composition (0) - reference AND body composition (0)

Impedance AND oncology AND …. reference (0) - body composition (0) - reference AND body composition (0)

Impedance AND cancer AND ….. reference (0) - body composition (7) - reference AND body composition (0)

(electric Impedance OR electrical impedance OR *impedance) AND (reproducibility of results OR reference values OR reference standards) AND (body) NOT (child OR infant OR adolescent) (30)

Data base: Embase – Ovid

Impendance AND surgery AND …. reference (58) - body composition (96) - reference AND body composition (17)

Impendance AND operation AND ….. reference (19) - body composition (13) - reference AND body composition (1)

Impendance AND oncology AND …. reference (4) - body composition (7) - reference AND body composition (1)

Impendance AND cancer AND …. reference (26) - body composition (113) - reference AND body composition (6)

(electric impedance OR electrical impedance OR *impedance) NOT (child OR infant OR adolescent) (397) …. AND (reproducibility of results OR reference values OR reference standards AND (body) (11). Limits: Human, English language

Data base: SUMsearch

Impedance AND surgery AND …. reference (124) - body composition (195) - reference AND body composition (18)

Impedance AND operation AND …. reference (82) - body composition (97) - reference AND body composition (11)

Impedance AND oncology AND …. reference (9) - body composition (20) - reference AND body composition (3)

Impedance AND cancer AND …. reference (48) - body composition (139) – reference AND body composition (9)

(electric impedance OR electrical impedance OR *impedance) AND (reproducibility of results OR reference values OR reference standards ) AND (body) (511). Limits: Humans only, English only – Age adult.

Data base: CINAHL

(“Electric Impedance” MH) AND surgery AND …. reference (5) - body composition (25) - reference AND body composition (2)

(“Electric Impedance” MH) AND operation AND …. reference (1) - body composition (1) - reference AND body composition (1)

(“Electric Impedance” MH) AND oncology AND …. reference (0) - body composition (0) - reference AND body composition (0)

(“Electric Impedance” MH) AND cancer AND …. reference (5) - body composition (28) - reference AND body composition (1)

(“MH body composition”) AND (“MH electric impedance”) AND adult NOT child NOT adolescent NOT infant (445)

71

Cha

pte

r 5Bi

oele

ctri

cal i

mp

edan

ce a

naly

sis

to e

stim

ate

bod

y co

mp

osit

ion

in s

urgi

cal a

nd o

ncol

ogic

al p

atie

nts:

a s

yste

mat

ic re

view

Study eligibility criteria Types of studiesIncluded were two different types of studies: (1) studies focusing on the development of new equations suitable for a specific target population, (2) studies describing the validity of BIA estimations compared to a reference method.

All study designs were allowed, with the exception of case reports, case series, meet-ing abstracts and editorial letters. Excluded were studies written in languages other than English.

Types of participants Included were studies in adult human surgical and oncological patients. The term ‘surgical’ was defined as the period before the surgical procedure up to one year postoperatively. Studies with regard to oncological patients were included up to one year after finishing therapy (e.g. chemotherapy or radiotherapy).

Excluded were studies in patients with altered body composition and/or a disturbed fluid balance as it can result in overestimation and underestimation of body compart-ments measured by BIA, i.e. patients with a BMI ≥ 35 (extreme obesity); patients with endocrine diseases that influence body composition (e.g. Cushing syndrome); patients treated with (growth) hormone; acutely ill intensive care patients; patients with clinical signs of oedema ; and patients before or after organ transplantation. 1, 5, 13

Types of BIA devicesIncluded were studies using bioimpedance devices measuring the entire body: single-frequency bioelectrical impedance analysis (SF-BIA); multiple-frequency bioelectrical impedance analysis (MF-BIA); and bioimpedance spectroscopy (BIS). 1 - 3, 7, 12, 13

Excluded were studies using foot-to-foot SF-BIA devices and segmental impedance techniques as these methods do not measure the entire body.

Study selection and data extraction One review author (EBH) collected the potential studies from the various databases and screened the articles on title and abstracts. From the full texts of the selected studies, three review authors (EBH, PLMR, MAEvB) independently included the studies into this review. Disagreement about inclusion was resolved by consensus. Data were extracted by one review author (EBH) with the use of an extraction form containing: aim of study (develop-ment of equation; measurement validity); body compartment of study (total body water (TBW); body cell mass (BCM); extra cellular water (ECW); intra cellular water (ICW); fat free mass (FFM); fat mass (FM); and lean body mass (LBM) which is also referred to as lean tissue mass (LTM); target population (surgery; oncology); number of patients studied; reference method; type and manufacturer of BIA device; and characteristics of the equations (new; existing; set of variables included in equation).

72

Types of reference methods used in the validity studiesFor the measurement of TBW deuterium dilution and tritiated water dilution are used. BCM can be estimated by radioactive total body potassium (TBK) by whole body counting and total body nitrogen (TBN) by neutron activation. ECW is measured by bromide dilu-tion and ICW can be measured by radioactive TBK. FFM, FM and LTM can be measured by hydrostatic weighing (hydro densitometry or underwater weighing), air-displacement plethysmography and dual energy X-ray absorptiometry (DXA). Magnetic resonance imaging (MRI) and computed tomography (CT) can both be used for the measurement of FM, muscle, skin, viscera and bone tissue. 3, 7, 12, 17, 18, 21, 22

Risk of Bias Altered body composition and/or fluid disturbances can result in overestimation and underestimation of body compartments measured by BIA. Therefore, patients suffering from oedema; significant fluid disturbances; extreme obesity; illnesses and treatments that could influence body composition and/or fluid balance should be excluded from analyses.

In order to evaluate whether the included studies are well designed, performed and described general information with regard to the reference method and the target popu-lation (compartment of study; main characteristics of the study population; surgical or oncological patients; number of participants; inclusion and exclusion criteria of the study population; recruitment procedures) should be clearly described.

In studies that focus on the development of new equations, besides a detailed descrip-tion of the used device (manufacturer; model), the variables included in the new equa-tion (resistance, gender, height, weight, age, others) must clearly be recorded. In addition, studies on the validity of BIA estimations must give detailed information on the type of device, existing equation(s), and reference methods.

Best-evidence synthesis This systematic review is a qualitative synthesis of the available evidence. In view of the heterogeneity of the target population, the variability of study objectives and differences in methodological quality, a meta-analysis could not be performed.

Statistical Analysis Study characteristics were summarized using descriptive statistics. To explore the varia-tions of newly developed and existing equations, we used the frequently used normal ref-erence values published by Kyle as input for the equations to simulate measures of body compartments (TBW and FFM). 23 Kyle’s reference values are available for eight different age groups, each with a bandwidth of 9 years ranging from 15 to over 85 years. Hence, the clinical measures normally obtained from patients (gender, height, weight and age) and the direct estimations of the BIA device (resistance and reactance) at 50 kHz were replaced by these reference values. So, in our simulation approach each body compartment meas-ure was the result of a series of fixed reference values for patient and BIA, but calculated

73

Cha

pte

r 5Bi

oele

ctri

cal i

mp

edan

ce a

naly

sis

to e

stim

ate

bod

y co

mp

osit

ion

in s

urgi

cal a

nd o

ncol

ogic

al p

atie

nts:

a s

yste

mat

ic re

view

with different equations. Consequently, variability found in body compartments is the reflection of the differences between equations. Estimation results were grouped by body compartment and plotted per age group.

With regard to the validity of BIA estimations, we assessed the discrepancy between BIA estimations and reference methods, using the difference in means and the relative difference in %. The relative difference in % was calculated by: [(compartment estimated by BIA – compartment measured by reference method) / compartment measured by ref-erence method] x 100%. Statistical uncertainty was expressed in 95% confidence inter-vals (CI).

Results

Literature search and study selectionThe literature search revealed a total of 4369 studies. After exclusion of the duplicates (n = 838), and screening of titles and abstracts, the full text of 63 studies were indepen-dently examined by three reviewers. Fifty-two studies were excluded based on disease (e.g. obesity; nephrology; intensive care); uncertainty about the actual absence of fluid disturbances; age (adolescent; child); performed surgical procedure (e.g. gastric bypass due to obesity; organ transplantation); time point (more than one year postoperative); or lack of an actual reference method (e.g. comparison of two different BIA devices). A total of 11 studies (six studies with surgical patients and five studies with oncological patients) met the inclusion criteria, and were considered suitable for the systematic review (Figure 1). 24 – 34

Study characteristics Table 1 shows detailed information about the aim of the study (development of equations 24 - 29, 31 and/or the validity of the BIA estimations 25, 29, 30, 32 – 34) and the measured body com-partments (TBW; BCM; ECW; FFM; and (%)FM). The main characteristics of the study popu-lation were well described in all studies except for one. 31

Surgery was the field of study in six of the included studies. 24, 26, 28, 30, 31, 34 Four of these studies consisted of a heterogeneous group of surgical patients, including cancer patients, 24, 26, 28, 31 one study reported on patients after major abdominal surgery 30 and one described patients undergoing elective heart surgery. 34

Five studies focused on oncological patients: three studies consisted of untreated ambu-latory (incurable) patients with a malignancy in the gastro-intestinal tract or lung, 25,29,33 one study described patients suffering from esophageal or gastric cancer with various extents of weight loss, 27 and one studied had included men suffering from prostate cancer. 32

All studies provided information about the used reference method (tritiated water dilution; deuterium dilution; bromide dilution; TBK; and DXA), the four manufacturers of the BIA devices, and their types: SF-BIA; BIS; and MF-BIA + BIS. In all but two studies 24, 26 the specific models of the devices were also described.

74

In seven studies (four surgical; three oncological) statistical regression equations were developed to estimate TBW, BCM, and ECW (see also Appendix 3). 24 - 29, 31 Resistance at 50 kHz, height, and body weight were variables often included in these newly developed regression equations.

The six studies evaluating the validity of BIA estimations (two surgical; four oncological) 25, 29 30, 32, 33, 34 used existing general equations to estimate TBW, FFM, and (%)FM (Appendix 4). 3, 35-46 The majority of the existing equations also included the resistance value at 50 kHz; height; and body weight in their set of variables. In addition, three studies (also) made use of an equation incorporated into the instrument’s software itself. 29, 33, 34

Figure 1 Flow of information through the different phases of the systematic review Figure 1 Flow of information through the different phases of the systematic review

Pubmed -Medline 2021 studiesCochrane 50 studies

Embase Ovid 372 studies Sum Search 1189 studies

C inahl 472 studies Other sources 3 studies

Total 4639 studies

838 duplicates removed

11 studies included in qualitative synthesis

Surgery n= 6Oncology n = 5

2596 studies excluded based on title

3801 studies screened on title

1142 studies excluded basedon abstract

1205 studies screened on abstract

52 studies excluded based on full text

63 studies screened on full text

75

Cha

pte

r 5Bi

oele

ctri

cal i

mp

edan

ce a

naly

sis

to e

stim

ate

bod

y co

mp

osit

ion

in s

urgi

cal a

nd o

ncol

ogic

al p

atie

nts:

a s

yste

mat

ic re

view

Tab

le 1

Cha

ract

eris

tics

of th

e in

clud

ed s

tudi

es (N

= 1

1)G

ener

al in

form

atio

nR

efer

ence

m

eth

od

eB

ioel

ectr

ical

imp

edan

ce (B

IA) e

stim

atio

n

Dev

ice

Equ

atio

n

Aut

hor

A

im

of s

tudy

Com

par

t-m

ent o

f st

udy

a

Pop

ulat

ion

bSu

rger

y (S

) or

onc

olog

y (O

) c

N d

Typ

e f

Man

ufac

-tu

rer g

Mod

elEq

uatio

n re

f, G

,F,M

hVa

riab

les

in e

quat

ion

R kH

z k

Gen

der

Hei

ght

Wei

ght

Age

Oth

ers

Un-

know

n l

Schr

oede

r24

Dev

elop

-m

ent

equa

tion

TBW

hete

roge

-ne

ous

surg

ical

gr

oup

S12

0TD

SF

-BIA

RJL

un-

know

nne

w

50x

Fred

rix 25

D

evel

op-

men

teq

uatio

n an

d M

easu

-re

men

t va

lidit

y

TBW

new

can

cer

pat

ient

s no

tr

eatm

ent

O

33D

DSF

-BIA

RJL

101

new

50

x

exis

ting

35 (G

)50

x

exis

ting

36

(G,M

,F)

50x

x

exis

ting

37

(G,M

,F)

50x

xx

(x)

Fear

on 26

D

evel

op-

men

teq

uatio

n

TBW

BCM

hete

roge

-ne

ous

surg

ical

gr

oup

S43

TD TBK

SF-B

IARJ

Lun

-kn

own

new

TBW

50x

x

new

TBK

50x

xx

Mih

olic

27

Dev

elop

-m

ent

equa

tion

ECW

canc

er

pat

ient

s w

ith w

eigh

t lo

ss

O21

BD

SF-B

IARJ

L10

3ne

w

-x

x

Han

nan

28

Dev

elop

-m

ent

equa

tion

TBW

ECW

hete

roge

-ne

ous

surg

ical

gr

oup

S43

TD BDM

F-BI

AX

itron

4000

Bne

w (4

) TBW

50 and

500

x(x

)(x

)

new

(2) E

CW

5 an

d 50

x

Sim

ons

29

Dev

elop

-m

ent

equa

tion

an

d M

easu

-re

men

t va

lidit

y

TBW

amb

ulat

ory

in

cura

ble

ca

ncer

p

atie

nts

O41

DD

SF-B

IARJ

L10

1ne

w (

2)50

x

exis

ting

inco

rpat

ed 3

7 (G

, M,F

)

50(x

)x

x(x

)

exis

ting

38 (G

)50

xx

xx

exis

ting

39

(G,M

,F)

50(x

)x

x(x

)

exis

ting

40 (G

)50

xx

exis

ting

42 (G

)50

x

76

Gen

eral

info

rmat

ion

Ref

eren

ce

met

ho

d e

Bio

elec

tric

al im

ped

ance

(BIA

) est

imat

ion

Dev

ice

Equ

atio

n

exis

ting

43

(M, F

)50

xx

x

exis

ting

44

(M, F

)50

xx

exis

ting

45 (G

)50

xx

exis

ting

46 (G

)50

xx

xx

Jens

en 30

M

easu

re-

men

t va

lidit

y

FFM

FMp

osto

p.

abdo

min

al

surg

ery

S 28

DXA

SF-B

IAH

TSA

nim

eter

exis

ting

39

(G,M

,F)

50(x

)x

x(x

)

Han

nan

31

Dev

elop

-m

ent

equa

tion

TBW

ECW

hete

roge

-ne

ous

surg

ical

gr

oup

S29

TD BDM

F-BI

Aap

-pr

oach

+BI

S ap

-pr

oach

Xitr

on40

00B

new

TBW

200

xx

x

new

EC

W5

xx

Smith

32

Mea

sure

-m

ent

valid

ity

% F

Mam

bul

ator

y p

rost

ate

canc

er

O38

DXA

SF-B

IARJ

L10

1Aex

istin

g 43

(M

,F)

50x

xx

exis

ting

45 (G

)50

xx

Elle

gård

33

Mea

sure

-m

ent

valid

ity

TBW

FFM

incu

r-ab

le a

nd

untr

eate

d ca

ncer

gi

-tra

ct

O13

2D

XABI

SX

itron

Hyd

ra

4200

exis

ting

inco

r-p

orat

ed

x

exis

ting

3 (G

)x

exis

ting

41 (G

) ze

ro

and

infin

-it

y

xx

van

Venr

ooij

34

Mea

sure

-m

ent

valid

ity

FFM

FMel

ectiv

e he

art

surg

ery

S26

DXA

BIS

Fres

eniu

s Bo

dy

Scou

tex

istin

g in

cor-

por

ated

x

a C

omp

artm

ent o

f stu

dy: T

BW =

tota

l bod

y w

ater

; BC

M =

bod

y ce

ll m

ass;

EC

W =

ext

ra c

ellu

lar w

ater

; FFM

= fa

t fre

e m

ass;

FM

= fa

t mas

s; %

FM

= p

erce

ntag

e fa

t mas

s.b P

opul

atio

n: h

eter

ogen

eous

sur

gica

l gro

up =

het

erog

eneo

us g

roup

of s

urgi

cal p

atie

nts

(can

cer,

ibd,

pan

crea

titis

); ne

w c

ance

r pat

ient

s no

trea

tmen

t = a

mb

ulat

ory,

pat

ient

s w

ith m

alig

nanc

y of

gi-t

ract

or l

ung

no c

hem

othe

rapy

or

radi

atio

n th

erap

y; c

ance

r pat

ient

s w

ith w

eigh

t los

s =

pat

ient

s w

ith e

sop

hage

al o

r gas

tric

can

cer w

ith v

ario

us e

xten

ts o

f wei

ght l

oss;

am

bul

ator

y in

cura

ble

can

cer p

atie

nts

= g

roup

of a

mb

ulat

ory

pat

ient

s w

ith in

cura

ble

and

unt

reat

ed

canc

er o

f the

gi-t

ract

; pos

top.

ab

dom

inal

sur

gery

= p

atie

nts

2-4

mon

ths

afte

r m

ajor

ab

dom

inal

sur

gery

; am

bul

ator

y p

rost

ate

canc

er =

gro

up o

f am

bul

ator

y lo

cally

adv

ance

d , l

ymp

h-no

de p

ositi

ve, o

r re

curr

ent

men

with

pro

stat

e ca

ncer

no

horm

one

ther

apy;

incu

rab

le a

nd u

ntre

ated

can

cer =

gro

up o

f pat

ient

s w

ith in

cura

ble

and

unt

reat

ed c

ance

r of t

he g

i-tra

ct; e

lect

ive

hear

t sur

gery

= p

atie

nts

unde

rgoi

ng e

lect

ive

open

hea

rt s

urge

ry (C

ABG

, hea

rt v

alve

).c Su

rger

y (S

) or o

ncol

ogy

(O):

S =

sur

gica

l pat

ient

s; O

= o

ncol

ogic

al p

atie

nts.

d N: n

umb

er o

f par

ticip

ants

in th

e st

udy.

e Re

fere

nce

met

hod:

TD

= tr

itium

dilu

tion

for T

BW; D

D =

deu

teriu

m d

ilutio

n fo

r TBW

; TB

K =

radi

oact

ive

40K

(= T

BK) f

or B

CM

; BD

= b

rom

ide

dilu

tion

for E

CW

; DXA

= d

ual e

nerg

y x-

ray

abso

rptio

met

ry fo

r FFM

and

FM

.

f Typ

e of

BIA

dev

ice:

SF

= s

ingl

e fr

eque

ncy

bio

elec

tric

al im

ped

ance

ana

lysi

s (S

F-BI

A);

MF

= m

ultip

le fr

eque

ncy

bio

elec

tric

al im

ped

ance

ana

lysi

s (M

F-BI

A);

BIS

= b

ioim

ped

ance

sp

ectr

osco

py.

g Man

ufac

ture

r of B

IA d

evic

e : R

JL s

yste

ms,

Det

roit,

USA

; Xitr

on te

chno

logi

es, S

an D

iego

, USA

; HTS

-eng

inee

ring,

Ode

nse,

Den

mar

k; F

rese

nius

Kab

i Med

ical

Car

e, B

ad H

omb

urg,

Ger

man

y.h E

quat

ion,

ref a

nd (G

,F,M

): ne

w =

dra

ftin

g ne

w e

quat

ion

; exi

stin

g +

num

ber

+ (G

,M,F

) =

use

of e

xist

ing

equa

tion

+ re

fere

nce

num

ber

and

an

equa

tion

for e

ntire

the

grou

p (G

), eq

uatio

n fo

r the

mal

e su

bgr

oup

(M),

equa

tion

for t

he

fem

ale

sub

grou

p (F

); ex

istin

g in

corp

orat

ed =

use

of e

quat

ion

inco

rpor

ated

in th

e in

stru

men

t’s s

oftw

are.

k R

kH

z: R

esis

tanc

e in

kH

z in

clud

ed in

equ

atio

n.

l unkn

own:

no

docu

men

tatio

n av

aila

ble

with

rega

rd to

the

used

equ

atio

n.

77

Cha

pte

r 5Bi

oele

ctri

cal i

mp

edan

ce a

naly

sis

to e

stim

ate

bod

y co

mp

osit

ion

in s

urgi

cal a

nd o

ncol

ogic

al p

atie

nts:

a s

yste

mat

ic re

view

Appendix 3 Newly drafted equations used in 7 studies

Author ref Reference method a

BIA b Equation

Equations for the estimation of total body water (TBW) in litres

Schroeder 24 TD SF-BIA TBW (L) = 1.04 (TBW estimated by SF-BIA) - 1.94

Fredrix 25 DD SF-BIA TBW (L) = 8.9 + 0.5 (height2) / resistance 50

Fearon 26 TD SF-BIA TBW (L) = 0.208 (weight) + 0.275 (height) – 0.0299 (resistance 50

) - 10.81

Hannan 28 TD MF-BIA TBW (L) = 0.497 (height)2 / resistance 50

+ 0.5 (antero-posterior thick-ness) + 0.275

TBW (L) = 0.45 (height)2 / resistance 500

+ 0.46 (antero-posterior thick-ness) + 0.0119 (height)2 / reactance

50 – 0.0106 (height)2 / reactance

500 - 1.04

TBW (L) = 0.446 (height)2 / resistance 50

+ 0.126 (weight) + 5.82

TBW (L) = 0.399 (height)2 / resistance 500

+ 0.114 (weight) + 5.69

Simons 29 DD SF-BIA TBW (L) normal weight = 9.64 + 0.516 (height2 / resistance50

)

TBW (L) underweight = 7.33 + 0.528 (height2 / resistance 50

)

Hannan 31 TD MF-BIA TBW (L) = 0.239 (height)2/ resistance 200

+ 0.189 (weight)+ 2.97 (gen-der:1 = male, 0 = female) + 5.46

BIS not described

Equations for the estimation of extra cellular water (ECW) in litres

Miholic 27 BD SF-BIA ECW (L) = - 0.587 + 0.0098 (height2 / reactance 50

) + 0.09 (weight) + 0.936 (phase angle

50)

Hannan 28 BD MF-BIA ECW (L) = 0.0119 (height)2 / resistance 50

+ 0.123 (height)2 / resistance

50 + 6.15

ECW (L) = 0.01 (height)2 / resistance 50

+ 0.165 (height)2 / resistance 5 +

5.75

Hannan 31 BD MF-BIA ECW (L) = 0.178 (height)2 / resistance 5 + 0.0688 (weight) + 3.77

BIS not described

Equations for the estimation of body cell mass (BCM) by total body count potassium (TBK) in grams

Fearon 26 TBK SF-BIA TBK (g) = 0.0055 (height)2 - 0.0597(weight) – 0.1783 (resistance 50

) + 1.22 (reactance

50) – 11.475 (total body potassium ) + 0.2203 (age) +

29.184

a Reference method : TD = tritiated water dilution; DD = deuterium dilution; BD = bromide dilution; TBK = total body count potassium b estimation: SF-BIA = single frequency bioelectrical impedance analysis; MF-BIA = multi frequency bioelectrical impedance analysis; BIS = bioimpedance spectroscopy

78

Bias in the individual studies Risk of bias with regard to design and performance may have occurred in the majority of the included studies. In only one study the inclusion and exclusion criteria, the time period in which patients were included and the recruitment methods were clearly described. 34 In addition, the conditions under which the BIA estimations were used to obtain body com-position estimates were not completely described in all studies.

One study included 120 subjects in total, with eight being healthy volunteers. Data from these healthy subjects was included in the presented results. 24 In another study the information of the entire study population, including the postoperative subgroup, was used in order to develop a new statistical regression equation for the estimation of ECW. 27 In two studies the equations to estimate FFM were used from the indirect estimations of TBW and %FM. 29, 32

Two studies both explicitly described the presence of oedema in a subpopulation. 28, 31 To reduce the risk of bias, the patients described to have oedema were excluded from the results.

Variability of newly developed and existing general equations The included studies describe 15 newly developed equations (10 for the estimation of TBW, four for ECW and one for BCM, Appendix 3), 25 general existing equations (14 for TBW, and 11 for FFM, Appendix 4) and three equations incorporated in the instrument’s software.

Just to illustrate the variation between the equations, the normal reference values of Kyle were included in the newly developed and existing general equations. 23 This was not possible for five of the 10 newly developed equations for TBW 24, 28, 31 and three of 25 gen-eral existing equations for TBW and FFM 3, 41 because these equations include variables for which normal reference values are unknown (e.g. resistance at 200 kHz or reactance at 500 kHz). Based on the variables included in the equations, a graphical representation was possible for five newly developed equations with regard to TBW as described in Appendix 3 and depicted in Figure 2. 25, 26, 28, 29 The newly developed equations show a variation up to 5 litres for the TBW estimations.

Figures could also be constructed for twelve out of fourteen of the (gender specific) existing equations for TBW and ten out of eleven for FFM (see Appendix 4). 24 28, 31 A wide variation between the various existing general equations was observed for TBW (up to 20 litres or kilograms) and for FFM (over 25 kg) (Figures 3 and 4).

The validity of BIA estimations Table 2 shows that in four of the six studies evaluating the validity of BIA estimations the difference in means and the relative difference in % (point estimate) could be described and/or calculated completely. 25, 32, 33, 34 In the remaining two studies these results could only partially be described or calculated as the necessary data were incomplete or missing. 29, 30

Compared to the reference methods deuterium dilution and DXA, the TBW compart-ment was mainly underestimated by the various BIA estimations. The calculated relative

79

Cha

pte

r 5Bi

oele

ctri

cal i

mp

edan

ce a

naly

sis

to e

stim

ate

bod

y co

mp

osit

ion

in s

urgi

cal a

nd o

ncol

ogic

al p

atie

nts:

a s

yste

mat

ic re

view

Figure 3 Presentation of the variability between general existing equations using normal reference values - total body water (TBW) in litres or kilograms

Application of the normal reference val-ues of Kyle for resis-tance, body weight, height, gender and age determined at 50 kHz in existing general equations to estimate total body water (TBW) as described in Ap-pendix 4. 23

x-axis describing various age-groups defined by Kyle. y-axis describing the estimated number of litres or kilograms of TBW.

Legend: 1.Lukaski 35 – group equation; 2. Kushner 36 - 1 – group equation; 3. Kushner 36 - 2 – male equation; 4. Kushner 36 - 3 – female equation; 5. van Loan 37 - 1 – group equation; 6. van Loan 37 - 2 – male equation; 7. van Loan 37 - 3 – female equation; 8. Lukaski 38 – group equation; 9. Heitmann 39 – 1 – group equation; 10. Heitmann 39- 2 – male equation; 11. Heitmann 39- 3 – female equation; 12. Kushner 40 - group equation.

Figure 2 Presentation of the variability between newly developed equations using normal reference values - total body water (TBW) in litres

Application of the normal reference val-ues of Kyle for resis-tance, body weight, height, gender and age determined at 50 kHz in the newly developed equations to estimate total body water (TBW) as described in Ap-pendix 3. 23

x-axis describing various age-groups defined by Kyle. y-axis describing the estimated number of litres of TBW.

Legend: 1. Fredrix 25; 2. Fearon 26; 3. Hannan 28 - nr 3; 4. Simons 29 - nr 1; 5. Simons 29 – nr 2.

80

difference ranged from -18.8% to + 7.2% (total range 26%). The smallest relative differ-ence (0.9%) was demonstrated in a group of incurable, underweight cancer patients 29 by means of a SF-BIA device and the Heitmann equation. 39 The largest underestimated BIA measurement (relative difference -18.8%) was demonstrated by using a SF-BIA device and the Deurenberg equation (45) in the same study among a normal weight subgroup. 29

FFM also tended to be underestimated by BIA. Relative differences ranged from -15.2% to + 3.8% (total range 19%). The smallest relative difference (3.7%) was demonstrated in a group of patients after elective cardiac surgery using a BIS device in combination with an equation incorporated into the instrument’s software. 34 The largest underestimated BIA measurements (relative difference -15.2%) was shown in patients suffering from incurable and untreated cancer of the gastro-intestinal tract. The BIA estimation was performed with a BIS device and an equation incorporated into the instrument’s software. 33

The FM estimations in the included studies were described in both %FM and kilo-grams of FM and this should be taken into account with regard to the interpretation of the results. A large variability in discrepancies between BIA and reference methods could be observed. Relative differences ranged from -15.7 % to 43.1% (total range 58.8%). The smallest relative difference (1.2%) was demonstrated in a group of patients undergoing elective cardiac surgery measured by means of a BIS device and an equation incorporated into the instrument’s software. 34 The largest overestimated BIA measurement (relative

Figure 4 Presentation of the variability between general existing equations using normal reference values - fat free mass (FFM) in kilograms

Application of the normal reference values of Kyle for resistance, body weight, height, gender and age de-termined at 50 kHz in existing general equations to esti-mate fat free mass (FFM) as described in Appendix 4. 23

x-axis describing various age-groups defined by Kyle. y-axis describ-ing the estimated number of kilograms of FFM.

Legend: 1. Heitmann 39 – 1 - group equation; 2. Heitmann 39 – 2 - male equation; 3. Heitmann 39 – 3 - female equa-tion; 4. Lukaski 42 – group equation; 5. Segal 43 -1 - male equation; 6. Segal 43- 2 - female equation; 7. Graves 44 - 1 - male equation; 8. Graves 44 – 2 - female equation; 9. Deurenberg 45 group equation; 10. Deurenberg 46 – group equation.

81

Cha

pte

r 5Bi

oele

ctri

cal i

mp

edan

ce a

naly

sis

to e

stim

ate

bod

y co

mp

osit

ion

in s

urgi

cal a

nd o

ncol

ogic

al p

atie

nts:

a s

yste

mat

ic re

view

Appendix 4 Existing general equations used in 13 studies

Reference GroupM / F a

Equations for the estimation of total body water (TBW) in litres

Ellis 3 G TBW (L) = fat free mass (kg) x 0.732

Lukaski 35 G TBW (L) = 2.03 + 0.63 (height cm2/resistance50

)

Kushner 36 G TBW (L) = 0.556 (height cm2 / resistance50

) + 0.0955 (weight kg) + 1.726

M TBW (L) = 0.396 (height cm2 / resistance50

) + 0.143 (weight kg) + 8.399

F TBW (L) = 0.382 (height cm2 / resistance50

) + 0.105 (weight kg) + 8.3154

Van Loan 37 G TBW (L) = 9.9868 + 0.000724 (height cm2) + 0.2822 (weight kg) - 0.0153 (resis-tance

50) - 2.3313 (gender; male = 0, female = 1) – 0.1319 (age years)

M TBW (L) = 8.40 + 0.3963 (height cm2/resistance50

) + 0.143 (weight kg)

F TBW (L) = 8.3148 + 0.3821(height cm2/resistance50

) + 0.1052 (weight kg)

Lukaski 38 G TBW (L) = 0.377 (height cm2/resistance50

) + 0.14 (weight kg) - 0.08(age years) + 2.9 (gender; male = 1, female = 0) + 4.65

Moissl 41 G TBW (L) = ρ ICW – (ρ ICW – ρ ECW) x (Rinfinity

/ R0) (2/3)

Equations for the estimation of total body water (TBW) in kilograms

Heitmann 39 G TBW (Kg) = 0.266 (height cm2/resistance50

) + 0.186 (weight kg) + 4.702 (gender; male = 1, female = 0) - 0.081(age years) + 11.03

M TBW (Kg) = 0.223 (height cm2/resistance50

) + 0.252 ( weight kg) + 8.799

F TBW (Kg) = 0.412 (height cm2/resistance50

) + 0.117 (weight kg) + 4.503

Kushner 40 G TBW (Kg) = 0.59 (height cm2/resistance50

) + 0.065 (weight kg) + 0.04

Equations for the estimation of fat free mass (FFM) in kilograms

Heitmannn 39 G FFM (Kg) = 0.295 (height cm2/resistance50

) + 0.204 (weight kg) + 5.009 (gender; male = 1, female = 0) - 0.076 (age years) + 0.227 (height in cm) – 17.04

M FFM (Kg) = 0.244 (height cm2/resistance50

) + 0.270 (weight kg) + 0.284 (height in cm) - 28.02

F FFM (Kg) = 0.411 (height cm2/resistance50

) + 0.141 (weight kg) + 0.267 (height in cm) - 28.61

Moissl 41 G FFM (Kg) = (δECW

x ECWbis

) + (δICW

x ICWbis

)

Lukaski 42 G FFM (Kg) = 0.810 (height cm2/resistance50

) + 6.39

Segal 43 M FFM (Kg) = 0.00132 (height cm2) - 0.04394 (resistance50

) + 0.3052 (weight kg) - 0.1676 (age years) + 22.66827

F FFM (Kg) = 0.00108 (height cm2) - 0.0209 (resistance50

) + 0.23199 (weight kg) - 0.06777 (age years) + 14.59453

Graves 44 M FFM (Kg) = 0.485(height cm2/resistance50

) + 0.338 (weight kg) + 5.32

F FFM (Kg) = 0.475(height cm2/resistance50

) + 0.295(weight kg) + 5.49

Deurenberg 45 G FFM (Kg) = (0.671 x height m2/resistance50

) + 3.1 (gender; male =1, female = 0) + 3.9

Deurenberg 46 G FFM (Kg) = 0.340 (height cm2/resistance50

) + 15.34(height in cm) + 0.273(weight kg) - 0.127(age years) + 4.56 (gender; male = 1, female = 0) - 12.44

a Group, Male or Female specific equation

82

Table 2 Difference between BIA estimations and reference methods expressed in difference in means and relative difference in 6 studies

Author ref

Reference method a

Type of device b

Reference equation c

BIA estimationmean (95% CI)

Reference method mean (95% CI)

Difference in means (95% CI)

Relative differ-ence in % (point estimate): d

Field of study: total body water (TBW)

Fredrix 25 reference method: DD Device: SF-BIA

35 35.6 L(33.5 to 37.7)

36.0 L (33.9 to 38.1 )

-0.4 L (-3.2 to 2.4)

-1.1

36 38.0 L (35.9 to 40.1)

36.0 L (33.9 to 38.1 )

2.0 L(-0.8 to 4.8)

5.6

37 32.6 L (30.5 to 34.7)

36.0 L (33.9 to 38.1 )

-3.4 L(-6.2 to -0.6)

-9.4

Simons29 normal weight eReference method: DDDevice: SF-BIA

37 33.2 L 38.2 L (35.9 to 40.5)

not available f -13.1

37 39.3 L 38.2 L (35.9 to 40.5)

not available f 2.9

38 32.2 L 38.2 L (35.9 to 40.5)

not available f -15.7

39 36.9 L 38.2 L (35.9 to 40.5)

not available f -3.4

40 37.3 L 38.2 L (35.9 to 40.5)

not available f -2.4

42 37.7 L 38.2 L (35.9 to 40.5)

not available f -1.3

43 34.0 L 38.2 L (35.9 to 40.5)

not available f -11.0

44 39.6 L 38.2 L (35.9 to 40.5)

not available f 3.7

45 31.0 L 38.2 L (35.9 to 40.5)

not available f -18.8

46 33.4 L 38.2 L (35.9 to 40.5)

not available f -12.6

Simons29 under-weight gReference method: DDDevice: SF-BIA

37 35.6 L 33.2 L (30.1 to 36.3)

not available f 7.2

37 29.1 L 33.2 L (30.1 to 36.3)

not available f -12.3

38 28.5 L 33.2 L (30.1 to 36.3)

not available f -14.2

39 33.5 L 33.2 L (30.1 to 36.3)

not available f 0.9

40 32.7 L 33.2 L (30.1 to 36.3)

not available f -1.5

42 33.9 L 33.2 L (30.1 to 36.3)

not available f 2.1

43 29.6 L 33.2 L (30.1 to 36.3)

not available f -10.8

83

Cha

pte

r 5Bi

oele

ctri

cal i

mp

edan

ce a

naly

sis

to e

stim

ate

bod

y co

mp

osit

ion

in s

urgi

cal a

nd o

ncol

ogic

al p

atie

nts:

a s

yste

mat

ic re

view

Author ref

Reference method a

Type of device b

Reference equation c

BIA estimationmean (95% CI)

Reference method mean (95% CI)

Difference in means (95% CI)

Relative differ-ence in % (point estimate): d

44 34.6 L 33.2 L (30.1 to 36.3)

not available f 4.2

45 28.4 L 33.2 L (30.1 to 36.3)

not available f -14.5

46 30.1 L 33.2 L (30.1 to 36.3)

not available f -9.3

Ellegård 33 Reference method: DXADevice: BIS

in 32.9 kg (31.6 to 34.2)

36.9 kg (35.6 to 38.2)

-4.0 kg(-5.8 to -2.2)

-10.8

41 34.6 kg (33.4 to 35.8)

36.9 kg (35.6 to 38.2)

-2.3 kg(-4.1 to -.0.6)

-6.2

Jensen 30 Reference method: DXADevice: SF-BIA

39 not available f not available f Difference ± SD: 1.4 kg ± 2.5

Field of study: fat free mass (FFM)

Ellegård 33 Reference method: DXADevice: BIS

in 42.9 kg (41.2 to 44.6)

50.6 kg (48.8 to 52.4)

-7.7 kg(-10.2 to -5.2)

-15.2

3 44.8 kg (43.0 to 46.6)

50.6 kg (48.8 to 52.4)

-5.8 kg(-8.3 to -3.3)

-11.5

41 45.7 kg (44.2 to 47.2)

50.6 kg (48.8 to 52.4)

-4.9 kg(-7.3 to -2.5)

-9.7

van Venrooij 34 Reference method: DXADevice: BIS

in-pre h 63.0 kg (58.6 to 67.4)

60.7 kg (56.5 to 64.9)

2.3 kg(-3.6 to 8.2)

3.8

in-post i 62.1 kg (57.5 to 66.7)

59.9 kg (55.8 to 64.0)

2.2 kg(-3.8 to 8.2)

3.7

Field of study: percentage fat mass (%FM) and fat mass (FM)

Smith 32 Reference method: DXADevice: SF-BIA

43 22.5 % (20.7 to 24.3)

26.7% (25.0 to 28.4)

-4.2%(-6.7 to -1.7)

-15.7

45 38.2 % (35.9 to 40.5)

26.7% (25.0 to 28.4)

11.5%(-8.7 to.14.3)

43.1

age rela-ted

35.4% (32.2 to 38.6)

26.7% (25.0 to 28.4)

8.7% (5.2 to 12.2)

32.6

van Venrooij 34 Reference method: DXADevice: BIS a

in-pre h 25.2 kg (20.3 to 30.1)

24.9 kg (20.8 to 29.0)

0.3 kg(-6.0 to 6.61)

1.2

in-post i 24.9 kg (20.7 to 29.1)

24.6 kg (20.9 to 28.3)

0.3 kg(-5.2 to 5.8)

1.2

a Reference method: DD = deuterium dilution; DXA = dual energy x-ray absorptiometry b Type of device: SF-BIA = single frequency bioelectrical impedance analysis; BIS = bioimpedance spectroscopy c Reference equation = reference number of the existing equation; in = use of equation incorporated in the

instrument’s softwared Relative difference in %: [(compartment estimated by BIA – compartment measured by reference method) /

compartment measured by reference method] x 100%e Normal weight: Simons et al. (1995) defined normal weight as body weight > 95% of ideal body weight f Not available = necessary data to describe and/or calculate the difference in means, the 95% CI for the differ-

ence in means, and relative difference in % (point estimate) between the BIA estimation and reference method were insufficient or not available

g Underweight weight: Simons et al. (1995) defined underweight as: body weight < 95% of ideal body weight. h pre = measurement in preoperative patients i post = measurement in postoperative patients

84

difference 43.1%) was demonstrated in men suffering from prostate cancer 32 by using a SF-BIA device and the Deurenberg equation. 45

Disscusion

Eleven studies (six surgery and five oncology) were consistent with the predetermined inclusion criteria and were included in this systematic review exploring the variability of the equations and investigating the validity of the BIA estimations. In answer to our research questions, the following results, conclusions and recommendations can be elab-orated from this systematic review.

Variability of equations This systematic review demonstrates the development and the application of a large num-ber of BIA equations. The 11 included studies describe 15 newly developed equations, 25 general existing equations and three equations incorporated in the instrument’s software.

In seven of the eleven studies new equations were developed. 24 - 29, 31 In our opinion the continuous development of new equations seems redundant, unnecessary and undesira-ble without determining whether existing general equations are useful or can be improved. Thirteen studies used existing general (gender specific) equations for the estimation of TBW, FFM and %FM, none specifically developed for oncological or surgical patients. 3, 35- 46 The European Society for Clinical Nutrition and Metabolism (ESPEN) advises the applica-tion of the Geneva equation based on bioimpedance data measured at 50 kHz. 47 However, we cannot assess the usefulness of the Geneva equation with regard to oncological or sur-gical patients as none of the studies included in this systematic review used this equation or were published before the Geneva equation in 2001. In three of the included studies 29, 33, 34 an equation was incorporated into the instru-ment’s software. These included equations contribute to the black box phenomenon as most manufacturers are not transparent about the variables included and/or the incor-poration of an existing general equation. This uncertainty makes it difficult to determine whether these types of equations are the most appropriate. A BIA estimation is derived from the combination of a device and an population-specific equation considered suitable for the patient being measured. As described by Kyle et al., an adequate equation is therefore of great importance for a valid estimation of a patient’s body composition. 5 When using Kyle’s fixed reference values in the newly developed and existing general equations, we found a substantial amount of variability between the equations with regard to TBW and FFM, which is not surprising in view of the diversity of variables and regression weights included into the equations. In addition, study population, research team, laboratory, and the modified analysis techniques and correction factors with regard to the reference methods are also of influence to variability. So, one can wonder if valid BIA estimates of body compartments are per definition pos-sible in light of the absence of measurement precision.

85

Cha

pte

r 5Bi

oele

ctri

cal i

mp

edan

ce a

naly

sis

to e

stim

ate

bod

y co

mp

osit

ion

in s

urgi

cal a

nd o

ncol

ogic

al p

atie

nts:

a s

yste

mat

ic re

view

The validity of the BIA estimationsThe review findings demonstrate both overestimation and underestimation of the differ-ent body compartments with all the used existing general equations combined with all devices compared to the reference methods.

The review findings demonstrate a tendency towards underestimation of both the TBW and the FFM compartment irrespective the BIA devices and the used existing gen-eral equations. Underestimation of the TBW may wrongly assume dehydration and could result in the adverse clinical decision to hydrate the patient. Underestimation of the FFM has major consequences as well; it may result in the postponement of a surgical procedure (given that a low preoperative FFM is a potential risk factor for postoperative complica-tions). 48 – 52 It may also result in a delay in the beginning of (neo) adjuvant treatment given that a low FFM contributes to more severe side effects, 53 – 55 and unnecessary physical therapy and/or dietary therapy may be started.

The FM compartment demonstrated a large variability; overestimation, underestima-tion, and adequate estimation were observed in the two studies evaluating this body compartment. An incorrect estimate of the FM is undesirable as it also estimates the FFM incorrectly. Overestimation of the FM implies underestimation of FFM and may result in adverse clinical decisions.

Unfortunately, the absence of data made it impossible to calculate the difference in means in the studies of Simons 29 and Jensen 30 evaluating TBW. In addition, the relative difference could not be calculated in the study of Jensen.

This systematic review demonstrates that, although not ideal, the Heitmann SF-BIA equation 39 leads to the best estimate of TBW. With regard to FM and FFM, the most pref-erable equation cannot specifically be identified as it was an equation incorporated into the Bodyscout instrument’s software and details about the variables included are not described. 34

It is remarkable that the Deurenberg SF-BIA equation produced 45 the poorest estima-tions of TBW and %FM (largest relative difference). These results raise the question whether the used equation or the used device (both studies used a SF-BIA) primarily contributed to the large differences. In contrast to the claims, that MF-BIA and BIS devices estimate more accurately than SF-BIA, this review does not confirm the superiority of MF-BIA or BIS over SF-BIA in sur-gical and oncological patients. 1-3, 5, 12 – 14 Our included studies demonstrated the small-est and the largest relative difference for TBW by means of SF-BIA. For FFM the smallest and the largest relative difference were calculated when using a BIS device. For FM, the smallest relative difference was calculated with a BIS device and the largest difference when a SF-BIA was used. According to our knowledge, among oncological and surgi-cal patients no studies have been performed measuring a certain body compartment using different BIA devices in combination with a reference method at the same point in time.

86

Other aspects of influence to BIA estimationsAlthough device and equations have the most prominent impact on the outcome of the BIA estimation, other aspects and conditions may also play a role. Exercise and food intake before the measurement probably also influence the results. 18 None of the included stud-ies described whether activity before the measurement was (not) allowed, five studies reported (briefly) about fasting before the measurement. 25, 27, 29, 33, 34 The literature also recommends to take into account that the position of the body during the measurement, movements during the measurement, the degree of degreasing of the skin for securing the electrodes, the type of electrodes and the skin temperature and condition are all, to a certain extent, influence on BIA outcome. 18 Seven out of the 11 studies provided informa-tion about the position of electrodes and/or the (supine) position of patient during the measurement; and only one also described the room temperature. 24 - 28, 30, 32 The phase of the menstrual cycle and the use of contraceptives may also influence the results; however, these details were typically not provided. 18

Limitations of the review To estimate the validity of BIA, we used strict inclusion criteria and this resulted in exclu-sion of a large number of studies. Clinical practice is considerably more stubborn and patients often present with oedema, overweight or serious underweight. This systematic review shows that there are considerable differences between BIA and reference methods in a population selected to have a ‘normal’ body composition. What can be expected from BIA estimations within the total population of surgical and oncologic patients, including those with a (suspected) altered body composition, disturbed fluid balance, severe malnu-trition or extreme obesity? Moreover, oncological patients undergoing surgery consist of a very large and diverse group of patients incomparable in terms of disease, prognosis and treatment. Neither surgical nor oncological patients can be seen as one group of patients, but include a wide palette of patients with a variety of diseases. Therefore, conclusions cannot simply be drawn for the entire subpopulations.

Only a part of the new and existing equations could be shown in figures to illustrate varia tion between the equations. However, given the content of the regression variables and their regression weight an entirely different picture and result is not in line with the expectations.

As only English articles were included in this review, this may have influenced the completeness of the total available studies with regard to the estimation of body compo-sition by BIA in surgical and oncological patients.

Implications for practice In our hospital, we routinely perform BIA estimations in oncological surgical patients visit-ing the outpatient clinic preceding and following therapeutic interventions. The large vari-ability and the weak validity of BIA estimations, as described in this systematic review, raise the question whether it makes sense to continue BIA estimations in individual patients.

87

Cha

pte

r 5Bi

oele

ctri

cal i

mp

edan

ce a

naly

sis

to e

stim

ate

bod

y co

mp

osit

ion

in s

urgi

cal a

nd o

ncol

ogic

al p

atie

nts:

a s

yste

mat

ic re

view

We believe that the continuation of the BIA estimations can only be useful under strict conditions and we advise to perform and interpret BIA estimations only longitudinally. A number of subsequent estimations performed with the same equation, the same device, and under the same circumstances may generate useful clinical information on changes in body composition within the individual patient; however, this has not been proven.

With regard to the equation, our results indicate that application of the Heitmann equation 39 seems to contribute to a valid estimation of TBW in patients with incurable cancer. For FFM and FM the most preferable equation cannot be indicated as the best estimate was obtained with an unknown, ‘black box’ equation. With regard to the type of device, no preference can be expressed because both overestimation and underestima-tion was observed for each compartment irrespective of the device.

Future In future, it is desirable to obtain more knowledge about the application of the raw bioim-pedance data (e.g. resistance, reactance, impedance and the phase angle at one or more frequencies) or vector analysis. 10, 15, 16, 56 - 64 To overcome lack of clarity with regard to the equations, the direct use of raw impedance data or vector analysis may prove to more use-ful clinically. There is a growing number of studies that suggest that these data may be useful to predict nutritional status and/or clinical outcomes. Reference values for these raw bioimpedance parameters must be developed based on age, gender, and ethnicity; but in the meantime, the normal reference values for some bioimpedance parameters measured in large groups of healthy adults measured by Kyle et al. can be used as a first step. 23, 57

In addition, there is a need for coordination with regard to the used reference methods and the performance of these measurements between researchers.

Conclusions

This review explored the variability of regression equations used in the BIA estimations and evaluated the validity of BIA estimations in adult surgical and oncological patients. In the newly developed and the existing general equations, substantial variability was found between the equations for total body water and fat free mass. As measurement precision was absent, no valid estimate of the body compartments could be demonstrated. BIA mainly underestimated total body water and the fat free mass irrespective the used equation or device. Estimations of the fat mass body compartment varied widely. Our results indicate that application of the Heitmann equation contributes, to some extent, to a valid estimation of TBW. We conclude that BIA estimations in individual patient care can only be useful when BIA estimations are performed longitudinally and under strict conditions.

88

Appendix 1 Background information about the measurement of body compartments by reference methods and bioelectrical impedance analysis

Body compartments

2 compart-ment model

The total body weight = fat mass (FM) + fat free mass (FFM)

Multi com-partment models

Anatomic level. Body weight = TBO + TBC + TBH + TBN + TBCa + TBP +TBK + TBCl + TBNa + TBMg + etc.Molecular level. Body weight = TBW + FM + TBPr + OM+ STMCellular level. Body weight = FM + BCM + ECW + extracellular solids Functional – tissue system level. Body weight = adipose tissue (FM + cells) SM + bone (mineral + fluid + marrow) + other tissues.

total body water (TBW)

TBW = ECW + ICW. 60 % of the body weight consists of water in nonobese subjects. In vivo measurement techniques for TBW are tritiated water dilution, deuterium dilution and oxygen-18 dilution). The estimation error of each dilution is < 1 kg.

Intra cellular water (ICW)

ICW = TBW - ECW

Extra cellular water (ECW)

ECW = TBW - ICW

Fat free mass (FFM)

FFM = bone minerals, visceral proteins, ICW and ECW. It is assumed that the average hydration of FFM varies in age: newborns 80%, 10-yr old children 75%, and healthy adults 73%. Calculation of FFM in healthy well hydrated adults is: FFM = TBW / 0.732.

Fat mass (FM) FM = total body weight - FFM. OR:FM = total body weight – (fat-free extracellular solids (FFECS) +BCM +ECW)

Lean body mass (LBM)

LBM = ECW + body cell mass (BCM) + FFECS

Body cell mass (BCM)

As 98% of the total body potassium is found within the cells, it provides a good estimate of the BCM. Described ratios of total body potassium (TBK) to BCM: a. 120 mmol kg −1, corres-ponding with BCM (g) = 8.33 × TBK (mmol) for adults. b. 9.18 ± 0.09 for healthy adults.

Reference methods

Hydrostatic weighing

Hydrostatic weighing (hydro densitometry or underwater weighing) is the classic method to measure body composition. It uses the Archimedes principle to determine total body volume by measuring the difference between a subject’s weight in water and that in air and thus determining whole-body density. This technique typically requires the subject to be completely submerged underwater while exhaling maximally (yielding residual lung volume) to minimize the effect of buoyancy from lung air. The aim of hydrostatic weighing is to measure the density of the body. Body density = Wa / (((Wa - Ww) / Dw) - (RV + 100cc)).Wa = body weight in air (kg), Ww = body weight in water (kg), Dw = density of water, RV = residual lung volume, and 100cc is the correction for air trapped in the gastrointestinal tract. The body density (D) can be converted to FFM, FM and LBM. The limitation of hydrostatic weighing includes time, expense, and technical expertise and is therefore not usable in clinical practice.

Deuterium dilution

Deuterium (2H) is a stable isotope of hydrogen with a natural abundance in the oceans of earth. Deuterium dilution (heavy water, 2H20), is water enriched with the hydrogen isotope deuterium. Heavy water for research and commercial use is referred to as deuterium oxide. Deuterium is a stable isotope and can be used to measure total body water (TBW) because it behaves like water. Before and several hours after drinking deuterium dilution, a urine, saliva or blood sample is taken and deuterium enrichment of the samples is measured by using gas chromatography or mass spectroscopy. The deuterium space is calculated, then corrected for exchange of hydrogen with the non-aqueous compartment (4 %).

Tritiated wa-ter dilution

Tritiated water dilution (super-heavy water or really heavy water, 3H20) is a form of water where the usual hydrogen atoms are replaced with tritium. Tritiated water dilution can be used to measure total body water (TBW in a similar fashion to deuterium dilution). Tritiated water dilution is measured by ß-counting.

Bromide dilution

Bromide is the most commonly used tracer to measure the volume of extracellular water (ECW). Before and x hours after drinking a sodium bromide solution, blood and/or urine samples are taken. Bromide enrichment of samples is most frequently measured by high-pressure liquid chromatography. The ECW space is calculated and corrected for the Donnan equilibration (5%) and intracellular distribution (10 %).

89

Cha

pte

r 5Bi

oele

ctri

cal i

mp

edan

ce a

naly

sis

to e

stim

ate

bod

y co

mp

osit

ion

in s

urgi

cal a

nd o

ncol

ogic

al p

atie

nts:

a s

yste

mat

ic re

view

Total body potassium

Total body potassium (TBK) is primarily used to measure body cell mass (BCM), but is also used for to estimate FFM. The TBK is externally measured by detection of the naturally occur-ring radioactive isotope 40K with a constant fraction (0.012%) of the total body potassium. The subject is lying in a supine position on a bed in a shielded room (steel walls surrounded by concrete). The 40K signal is recorded, corrected for background, and measures the amount of natural occurring 40K.

Dual energy X-ray ab-sorptiometry (DXA)

Dual-energy X-ray absorptiometry (DXA) uses is based on the fact that the attenuation of X-rays through bone, lean tissue, and fat are different due to their differences in densities and chemical composition. With increasing photon energy, the differences in the attenuation properties for these tissues decrease. DXA estimates bone mineral content (BMC), lean body mass (LBM) and fat mass (FM). Fat free mass (FFM) can be calculated by adding lean tissue and BMC analysed by to whole body scans.

Bioelectrical impedance – principles and methods

The principle of bio-electrical impedance analysis (BIA)

Measuring the body compartments by bioelectrical impedance analysis is based on the as-sumption that the body can be modelled as a isotropic cylindrical conductor with its length proportional to the subject’s height. Impedance is based on measuring resistance (R) and reactance (Xc) of an alternating electrical current in the human body. Intracellular fluids, body fluids and electrolytes behave as electrical conductors (e.g. resistance R, subdivided in extracellular resistance (Re and intracellular resistance (Ri), and cell membranes act as electrical condensers and are involved in capacitance (e.g. reactance Xc).

Performance and conditi-ons of BIA

Ideally the impedance estimation is performed under standardized conditions; patients must be measured with an empty bladder, not wearing metal containing accessories and shoes or socks must be removed from the body side that is being measured. Patients are ly-ing relaxed in the supine position on an examination table or bed with their legs separated and their arms touching neither their torso nor the examination table. Traditionally, the evaluations are conducted on the patients’ non-dominant body side. After cleaning the skin with alcohol, the electrodes are placed on the dorsal surface of the hand and foot; proximal to the metacarpophalangeal and metatarsophalangeal joints, between the distal prominen-ces of the radius and ulna at the wrist and between the medial and lateral malleoli at the ankle. For avoiding electrode polarization and minimizing the effects of the impedance of skin beneath the electrodes, the so called four-electrode technique is often used; one pair of electrodes for passing current into the body and a second pair for sensing the resulting voltage drop.

Single-frequency bioelectrical impedance analysis (SF-BIA)

SF-BIA is by far the most widely available bioimpedance methodology, which involves the application of a bioelectrical current for the measurement of impedance at a single frequency, typically 50 kHz. BIA analysis at 50 kHz is based on empirical theory. Because a 50 kHz current may not penetrate completely into the cells, SF-BIA methods actually do not measure the entire ICW volume. Therefore, SF-BIA does not strictly measure TBW, but extrapolates to obtain a TBW estimate. The fat free mass (FFM) is indirectly calculated by SF-BIA according to the assumption that the FFM is constantly hydrated at 73% (FFM = TBW / 0.732).

Multiple-fre-quency BIA (MF-BIA)

MF-BIA is based on empirical linear regression models and measures at a few frequencies: A MF-BIA device obtains data at low frequencies (e.g. 1 or 5 kHz) for the quantification of ex-tracellular water (ECW) and at a high frequency (typically 50, 100, 200, or 500 kHz) to deter-mine (TBW) as the current can pass through the cell membrane at these higher frequencies. ICW is assessed by subtraction of TBW – ECW.

Bioimped-ance Spec-troscopy (BIS)

BIS measures impedance over the entire spectrum of frequencies, from 5 to 1000 kHz and is based on mathematical and physical modelling (e.g. cole-cole plot) and mixture equations (e.g. Hanai theory). The modelling procedure by BIS involves fitting the spectral data to the so called Cole-Cole model using nonlinear curve fitting. From this fitting the resistance at zero and infinite fre-quency are extrapolated. These values are used to calculate the resistance of the ICW and ECW. Next, Hanai mixture equations are used to calculate ECW and ICW with use of resistivity constants. Resistance at very low frequency (<1 kHz) is used to estimate extracellular water (ECW) TBW is calculated as the sum of ECW and ICW at very high frequencies (> 5 MHz).

References 1 - 14, 17 - 22, 26, 28, 39, 65

Abbreviations: BCM = body cell mass; BMC = bone mineral content; D = body density; Dw = density of water; DXA = dual-energy X-ray absorptiometry; ECW = extracellular water; FFM = fat free mass = bone minerals, vis-ceral proteins = ICW and ECW; FM = fat mass = total body weight – FFM = total body weight – (fat-free extracel-lular solids +BCM +ECW); ICW = intracellular water; LBM = ECW + BCM + fat-free extracellular solids; OM = bone minerals (OM); R = resistance; SM = skeletal muscle; RV = residual lung volume; STM = soft tissue mineral; TBC = = total body carbon; TBCa= total body calcium; TBCl= total body chlorine; TBH= total body hydrogen; TBK = total body potassium; TBMg = total body magnesium; TBN = total body nitrogen; TBNa= total body sodium; TBO = total body oxygen; TBP = total body phosphorus; TBPr = total body protein; TBW = total body water; Wa = body weight in air; Ww = body weight in water; Xc = reactance.

90

References1. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Gómez JM , Heitmann BL et al. Bioelectrical

impedance analysis-part I: review of principles and methods. Clin Nutr. 2004; 23: 1226-43. 2. De Lorenzo A, Andreoli A, Matthie J, Withers P. Predicting body cell mass with bioimpedance by using

theoretical methods: a technological review. J Appl Physiol. 1997; 82: 1542-58. 3. Ellis K J. Human body composition: in vivo methods. Physiol Rev. 2000; 80: 649-80. 4. Mourtzakis M, Prado CM, Lieffers JR, Reiman T, McCargar LJ, Baracos VE. A practical and precise approach to

quantification of body composition in cancer patients using computed tomography images acquired during routine care. Appl Physiol Nutr Metab. 2008; 33: 997-1006.

5. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Manuel Gómez J et al. Bioelectrical impedance analysis-part II: utilization in clinical practice. Clin Nutr. 2004; 23: 1430-53.

6. Lukaski HC. Requirements for clinical use of bioelectrical impedance analysis (BIA). Ann N Y Acad Sci. 1999; 873: 72-6.

7. Lukaski HC. Methods for the assessment of human body composition: traditional and new. Am J Clin Nutr. 1987; 46: 537-56.

8. Cox-Reijven PLM. The validation of bio-electrical impedance spectroscopy (BIS) for measuring body composition in patients. Thesis. Maastricht: ISBN 90-5681-17-9, 2002.

9. Dittmar M. Reliability and variability of bioimpedance measures in normal adults: effects of age, gender, and body mass. Am J Phys Anthropol. 2003; 122: 361-70.

10. Selberg O, Selberg D. Norms and correlates of bioimpedance phase angle in healthy human subjects, hospitalized patients, and patients with liver cirrhosis. Eur J Appl Physiol. 2002; 86: 509-16.

11. van Marken Lichtenbelt WD, Westerterp KR, Wouters L, Luijendijk SC. Validation of bioelectrical-impedance measurements as a method to estimate body-water compartments. Am J Clin Nutr. 1994; 60: 159-66.

12. Jaffrin MY, Morel H. Body fluid volumes measurements by impedance: A review of bioimpedance spectroscopy (BIS) and bioimpedance analysis (BIA) methods. Med Eng Phys. 2008; 30: 1257-69.

13. Earthman C, Traughber D, Dobratz J, Howell W. Bioimpedance spectroscopy for clinical assessment of fluid distribution and body cell mass. Nutr Clin Pract. 2007; 22: 389-405.

14. Buchholz AC, Bartok C, Schoeller DA. The validity of bioelectrical impedance models in clinical populations. Nutr Clin Pract. 2004; 19: 433-46.

15. Gupta D, Lis CG, Dahlk SL, Vashi PG, Grutsch JF, Lammersfeld CA. Bioelectrical impedance phase angle as a prognostic indicator in advanced pancreatic cancer. Br J Nutr. 2004; 92: 957-62.

16. Barbosa-Silva MC, Barros A J, Wang J, Heymsfield SB, Pierson RN. Bioelectrical impedance analysis: population reference values for phase angle by age and sex. Am J Clin Nutr. 2005; 82: 49-52.

17. Mattsson S, Thomas BJ. Development of methods for body composition studies. Phys Med Biol. 2006; 7:51(13):R203-28.

18. Brodie D, Moscrip V, Hutcheon R. Body composition measurement: a review of hydrodensitometry, anthropo-metry, and impedance methods. Nutrition. 1998 Mar; 14: 296-310.

19. Dutch Malnutrition Steering Group, Amsterdam, The Netherlands. http://www.stuurgroepondervoeding.nl [Accessed July 1st 2008].

20. Moher D, Liberati A, Tetzlaff J, Altman D G; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009; 151: 264-9.

21. Heymsfield SB, Wang Z, Visser M, Gallagher D, Pierson RN. Techniques used in the measurement of body composition: an overview with emphasis on bioelectrical impedance analysis. Am J Clin Nutr. 1996; 64: 478S-484S.

22. Ellis KJ, Shypailo RJ, Pratt JA, Pond WG. Accuracy of dual-energy x-ray absorptiometry for body-composition measurements in children. Am J Clin Nutr. 1994; 60: 660-5.

23. Kyle UG, Genton L, Slosman DO, Pichard C. Fat-free and fat mass percentiles in 5225 healthy subjects aged 15 to 98 years. Nutrition. 2001; 17: 534-41.

24. Schroeder D, Christie P M, Hill G L. Bioelectrical impedance analysis for body composition: clinical evaluation in general surgical patients. JPEN J Parenter Enteral Nutr. 1990; 14: 129-33.

25. Fredrix EW, Saris WH, Soeters PB, Wouters EF, Kester AD, von Meyenfeldt MF, Westerterp KR. Estimation of body composition by bioelectrical impedance in cancer patients. Eur J Clin Nutr. 1990; 44: 749-52.

26. Fearon KC, Richardson RA, Hannan J, Cowan S, Watson W, Shenkin A, Garden OJ. Bioelectrical impedance analysis in the measurement of the body composition of surgical patients. Br J Surg. 1992; 79: 421-3.

27. Miholic J, Reilmann L, Meyer H J, Körber H, Dieckelmann A, Pichlmayr R. Estimation of extracellular space and

91

Cha

pte

r 5Bi

oele

ctri

cal i

mp

edan

ce a

naly

sis

to e

stim

ate

bod

y co

mp

osit

ion

in s

urgi

cal a

nd o

ncol

ogic

al p

atie

nts:

a s

yste

mat

ic re

view

blood volume using bioelectrical impedance measurements. Clin Investig. 1992; 70: 600-5.28. Hannan WJ, Cowen SJ, Fearon KC, Plester CE, Falconer JS, Richardson RA. Evaluation of multi-frequency bio-

impedance analysis for the assessment of extracellular and total body water in surgical patients. Clin Sci (Lond). 1994; 86: 479-85.

29. Simons JP, Schols AM, Westerterp KR, ten Velde GP, Wouters EF. The use of bioelectrical impedance analysis to predict total body water in patients with cancer cachexia. Am J Clin Nutr. 1995; 61: 741-5.

30. Jensen MB, Hermann AP, Hessov I, Mosekilde L. Components of variance when assessing the reproducibility of body composition measurements using bio-impedance and the Hologic QDR-2000 DXA scanner. Clin Nutr. 1997; 16: 61-5.

31. Hannan WJ, Cowen SJ, Plester C, Fearon KC. Proximal and distal measurements of extracellular and total body water by multi-frequency bio-impedance analysis in surgical patients. Appl Radiat Isot. 1998; 49: 621-2.

32. Smith MR, Fuchs V, Anderson EJ, Fallon MA, Manola J. Measurement of body fat by dual-energy X-ray absorptiometry and bioimpedance analysis in men with prostate cancer. Nutrition. 2002; 18: 574-7.

33. Ellegård LH, Ahlén M, Körner U, Lundholm KG, Plank LD, Bosaeus IG. Bioelectric impedance spectroscopy underestimates fat-free mass compared to dual energy X-ray absorptiometry in incurable cancer patients. Eur J Clin Nutr. 200; 63: 794-801.

34. van Venrooij LM, Verberne HJ, de Vos R, Borgmeijer-Hoelen MM, van Leeuwen PA, de Mol BA. Preoperative and postoperative agreement in fat free mass (FFM) between bioelectrical impedance spectroscopy (BIS) and dual-energy X-ray absorptiometry (DXA) in patients undergoing cardiac surgery. Clin Nutr. 2010; 29: 789-94.

35. Lukaski HC, Johnson PE, Bolonchuk WW, Lykken GI. Assessment of fat-free mass using bioelectrical impedance measurements of the human body. Am J Clin Nutr. 1985; 41: 810-7.

36. Kushner RF, Schoeller DA. Estimation of total body water by bioelectrical impedance analysis. Am J Clin Nutr. 1986; 44: 417-24.

37. Van Loan M, Mayclin P. Bioelectrical impedance analysis: is it a reliable estimator of lean body mass and total body water? Hum Biol. 1987; 59: 299-309.

38. Lukaski HC, Bolonchuk WW. Estimation of body fluid volumes using tetrapolar bioelectrical impedance measurements. Aviat Space Environ Med. 1988; 59: 1163-9.

39. Heitmann BL. Prediction of body water and fat in adult Danes from measurement of electrical impedance. A validation study. Int J Obes. 1990; 14: 789-802.

40. Kushner RF, Schoeller DA, Fjeld CR, Danford L. Is the impedance index (ht2/R) significant in predicting total body water? Am J Clin Nutr. 1992; 56: 835-9.

41. Moissl UM, Wabel P, Chamney PW, Bosaeus I, Levin NW, Bosy-Westphal A et al. Body fluid volume determination via body composition spectroscopy in health and disease. Physiol Meas. 2006; 27: 921-33.

42. Lukaski HC, Bolonchuk WW, Hall CB, Siders WA. Validation of tetrapolar bioelectrical impedance method to assess human body composition. J Appl Physiol. 1986; 60: 1327-32.

43. Segal KR, Van Loan M, Fitzgerald PI, Hodgdon JA, Van Itallie TB. Lean body mass estimation by bioelectrical impedance analysis: a four-site cross-validation study. Am J Clin Nutr. 1988; 47: 7-14.

44. Graves JE, Pollock ML, Colvin AB, Van Loan M, Lohman TG. Comparison of different bioelectrical impedance analyzers in the prediction of body composition. Am J Hum Biol 1989; 1: 603-11.

45. Deurenberg P, van der Kooij K, Evers P, Hulshof T. Assessment of body composition by bioelectrical impedance in a population aged greater than 60 y. Am J Clin Nutr. 1990; 51: 3-6.

46. Deurenberg P, van der Kooy K, Leenen R, Weststrate JA, Seidell JC. Sex and age specific prediction formulas for estimating body composition from bioelectrical impedance: a cross-validation study. Int J Obes. 1991; 15: 17-25.

47. Kyle UG, Genton L, Karsegard L, Slosman DO, Pichard C. Single prediction equation for bioelectrical impedance analysis in adults aged 20-94 years. Nutrition. 2001; 17: 248-53.

48. Hassen TA, Pearson S, Cowled PA, Fitridge RA. Preoperative nutritional status predicts the severity of the systemic inflammatory response syndrome (SIRS) following major vascular surgery. Eur J Vasc Endovasc Surg. 2007; 33: 696-702.

49. CBO Guideline Perioperative Nutrition. Dutch Institute for Healthcare Improvement. Utrecht, The Netherlands. http:// www.cbo.nl/Downloads/114/rl_periovoed_07.pdf [Accessed 2007].

50. Wolfe RR. Is the double-blind randomized trial the most valid experimental approach to evaluating treatment modalities in critical ill patiënts? Curr Opin Clin Nutr Metab Care. 1998; 1: 185-187.

51. Windsor JA, Hill GL. Weight loss with physiologic impairment. A basic indicator of surgical risk. Ann Surg. 1988; 207: 290-6.

52. Awad S, Lobo DN. What’s new in perioperative nutritional support? Curr Opin Anaesthesiol. 2011; 24: 339-48. 53. Bower MR, Martin RC. Nutritional management during neoadjuvant therapy for esophageal cancer. J Surg

92

Oncol. 2009; 100: 82-7. 54. Prado CM, Baracos VE, McCargar LJ, Reiman T, Mourtzakis M, Tonkin K, Mackey JR, Koski S, Pituskin E, Sawyer

MB. Sarcopenia as a determinant of chemotherapy toxicity and time to tumor progression in metastatic breast cancer patients receiving capecitabine treatment. Clin Cancer Res. 2009; 15: 2920-6.

55. Prado CM, Lieffers JR, McCargar LJ, Reiman T, Sawyer MB, Martin L, Baracos VE. Prevalence and clinical implica-tions of sarcopenic obesity in patients with solid tumours of the respiratory and gastrointestinal tracts: a population-based study. Lancet Oncol. 2008; 9: 629-35.

56. Piccoli A, Pillon L, Dumler F. Impedance vector distribution by sex, race, body mass index, and age in the United States: standard reference intervals as bivariate Z scores. Nutrition. 2002; 18: 153-67.

57. Kyle U G, Genton L, Hans D, Pichard C. Validation of a bioelectrical impedance analysis equation to predict appendicular skeletal muscle mass (ASMM). Clin Nutr. 2003; 22: 537-43.

58. Barbosa-Silva MC, Barros AJ, Post CL, Waitzberg DL, Heymsfield SB. Can bioelectrical impedance analysis identify malnutrition in preoperative nutrition assessment? Nutrition. 2003; 19: 422-6.

59. Barbosa-Silva MC, Barros AJ. Bioelectric impedance and individual characteristics as prognostic factors for post-operative complications. Clin Nutr. 2005; 24: 830-8.

60. Barbosa-Silva MC, Barros AJ. Bioelectrical impedance analysis in clinical practice: a new perspective on its use beyond body composition equations. Curr Opin Clin Nutr Metab Care. 2005; 8: 311-7.

61. Cardinal TR, Wazlawik E, Bastos JL, Nakazora LM, Scheunemann L. Standardized phase angle indicates nutri-tional status in hospitalized preoperative patients. Nutr Res. 2010; 30: 594-600.

62. Gupta D, Lis CG, Dahlk SL, King J, Vashi PG, Grutsch JF et. al. The relationship between bioelectrical impedance phase angle and subjective global assessment in advanced colorectal cancer. Nutr J. 2008; 30: 19.

63. Itobi E, Stroud M, Elia M. Impact of oedema on recovery after major abdominal surgery and potential value of multifrequency bioimpedance measurements. Br J Surg. 2006; 93: 354-61.

64. Castillo Martínez L, Colín Ramírez E, Orea Tejeda A, Asensio Lafuente E, Bernal Rosales LP, Rebollar González V et. al. Bioelectrical impedance and strength measurements in patients with heart failure: comparison with functional class. Nutrition. 2007; 23: 412-8.

65. Schoeller DA. Hydrometry. In: Human Body Composition, edited by Roche AF, Heymsfield SB, Lohman TG. Champaign, IL: Human Kinetics, 1996, pp 25–43.

Chapter 6

Presence and persistence of nutrition-related symptoms during the first year following esophagectomy with gastric tube reconstruction in clinically disease-free patients

E.B. HaverkortJ.M. Binnekade

O.R.C. BuschM.I. van Berge Henegouwen

R.J. de Haan D.J. Gouma

World J Surg. 2010 Dec; 34(12):2844-52

94

Abstract

Background Esophagectomy with gastric tube reconstruction results in a variety of postoperative nutrition-related symptoms that may influence the patient’s nutritional status. Methods We developed a 15-item questionnaire, focusing on the nutrition-related complaints the first year after an esophagectomy. The questionnaire was filled out the first week after discharge and 3, 6, and 12 months after surgery. The use of enteral nutrition, meal size and frequency, social aspects related to eating, defecation pattern, and body weight were recorded at the same time points. We analysed the relationship between the baseline characteristics and the number of nutrition-related symptoms, as well as the relationship between those symptoms and body weight with linear mixed models.Results We found no significant within-patient change for the total number of nutrition-related symptoms (P = 0.67). None of the baseline factors were identified as predictors of the com-plaint scores. The most frequently experienced complaints were early satiety, postprandial dumping syndrome, inhibited passage due to high viscosity, reflux, and absence of hun-ger. One year after surgery, meal sizes were still smaller, the social aspects of eating were influenced negatively, and patients experienced an altered stool frequency. Directly after the surgical procedure 78% of the patients lost weight, and the entire postoperative year the mean body weight remained lower (P = 0.47). We observed no association between the complaint scores and body weight (P = 0.15). Conclusions After an esophagectomy, most patients struggle with nutrition-related symptoms, are confronted with nutrition-related adjustments and a reduced body weight.

Key wordsNutrition; Surgery; Esophagus; Esophagectomy; Postoperative; Gastric tube.

95

Cha

pte

r 6Pr

esen

ce a

nd p

ersi

sten

ce o

f nut

ritio

n-re

late

d sy

mpt

oms d

urin

g th

e fir

st y

ear f

ollo

win

g es

opha

gect

omy

with

gas

tric

tube

Introduction

An esophagectomy with gastric tube reconstruction, mainly performed as a potentially curative treatment in patients with cancer of the esophagus or cardia, is often related to a number of postoperative nutrition-related complaints. 1–5 Dysphagia, pain during alimen-tation, hoarseness, reflux, early satiety, abnormal gastric emptying, dumping syndrome, increased stool frequency, and fluctuations in bodyweight are frequently reported com-plaints after this surgical procedure. 6–17

These nutrition-related complaints might affect the nutritional status negatively by increasing the risk of malnutrition and reducing the quality of life. 18–21 Little is known about the short-term and long-term occurrence of or changes in these symptoms. There is also no detailed information about nutrition-related adjustments (e.g., the need for enteral nutrition, altered meal size and meal frequency, social aspects, stool consistency and stool frequency) or nutritional status in terms of body weight after esophagectomy.

Most previous studies evaluated the nutrition-related symptoms only once or twice in the first postoperative year. Furthermore, these studies did not use specific questionnaires to evaluate these nutrition-related symptoms systematically and in greater depth. 7, 9, 12, 13,

15, 22 In the present study, we addressed the following research questions: (1) Does

this patient have nutrition-related symptoms and have these symptoms increased or decreased during the first year after esophagectomy? (2) Are nutrition-related adjust-ments necessary after esophagectomy? (3) What is the patient’s nutritional status in terms of body weight and how has this status changed during the first year after esophagec-tomy? (4) Are nutrition-related symptoms significantly influenced by patient characteris-tics and surgery-related characteristics? (5) Is there a significant correlation between the nutrition-related symptoms and postoperative bodyweight?

Patients and methods

All consecutive postoperative patients who underwent a transhiatal or transthoracic esophagectomy with gastric tube reconstruction for both malignant and benign dis-eases at the surgical department of the Academic Medical Center, University Hospital of Amsterdam (The Netherlands) were considered eligible for inclusion in this study. Patients were screened for inclusion during a 4-year period (2002–2005).

The following exclusion criteria were applied: patients suffering from a diabetes-related neuropathy, a neurological disease, ulcerative colitis, Crohn’s or celiac disease, and patients with a proven allergy. Patients who were unable to speak and/or read Dutch were also excluded, as they were unable to fill out the questionnaire.

The main goal of our study was to describe the nutritional symptoms directly related to the surgical procedure. Therefore, patients were excluded if they had a proven recurrence of a malignant disease or had another life-threatening disease.

96

The protocol was approved by the institutional review board (Ethics Committee) of the Academic Medical Center at the University of Amsterdam. All patients were informed and gave written consent.

Surgical procedure The esophageal resection was carried out either via the transhiatal or the transthoracic approach and the reconstruction of the digestive tract was performed by a gastric tube. 6, 23, 24 Although a cervical gastroesophageal anastomosis has a leakage rate with less devastating consequences, the percentage of benign strictures is higher. However, the gastric tube is generally preferred over a colonic interposition because there is a lower morbidity rate and a better postoperative quality of life. 25 Esophageal resections are rarely performed for benign diseases such as achalasia, because gastric tube reconstruction still causes a variety of symptoms.

Baseline assessmentsPatient-related baseline data (age, sex, presence of malignancy, neoadjuvant treatment, admission duration), physical status (preoperative bodyweight, body mass index [BMI], co-morbidity, preoperative complaints), and surgery-related characteristics (American Society of Anesthesiologists [ASA] classification, type of surgical procedure, postoperative complications) were collected during the first postoperative week from electronic medical records and dietician records.

Postoperative nutritional careAccording to our hospital’s standard guidelines, postoperative patients were initially fed through a surgically placed jejunostomy feeding tube. Tube feeding uses a complete liq-uid enteral formula, which is given during the period when the patient is not allowed to eat or drink or is otherwise unable to meet nutritional needs. Seven days after the start of tube feeding, possible cervical leakage was evaluated by x-swallow. Oral food intake began with small, frequent meals during the day only if no leakage was found. The amount of tube feeding was tailored to the oral intake achieved and guided by a dietician. Sip feeding units (small portions of highly concentrated drinks) were prescribed as supple-mentary feeding. Sip feeding is prescribed in case of an insufficient intake of nutrients, and is a complete enteral formula to be taken orally as a beverage rich in energy, proteins, and micronutrients.

Follow-up assessmentsFor the purpose of the present study, a 15-item feeding questionnaire was developed at our hospital. This questionnaire focused on the nutrition-related symptoms experienced by individual patients. The aim of the questionnaire was to describe the patient’s com-plaints during the preceding week and covered the following items: (1) dysphagia and reflux (8 items) and (2) stomach content and dumping syndrome (7 items). The total score

97

Cha

pte

r 6Pr

esen

ce a

nd p

ersi

sten

ce o

f nut

ritio

n-re

late

d sy

mpt

oms d

urin

g th

e fir

st y

ear f

ollo

win

g es

opha

gect

omy

with

gas

tric

tube

ranged from 0 (no complaints) to 15 (suffering from all complaints, see Appendix 1). The feeding questionnaire was pre-tested on 15 patients, adjusted, and then re-tested on another sample of 10 patients before the start of the study.

Patients were asked to fill out the questionnaire at home before each planned follow-up assessment during the first week after discharge, and then at 3, 6, and 12 months after surgery. Follow-up assessments took place in the outpatient clinic. A dietician collected the questionnaires and discussed the reported nutrition-related symptoms with the patients. In addition, nutrition-related adjustments in terms of enteral nutrition, meal size and meal frequency, the nutrition-related social aspects, stool consistency and stool fre-quency, and the nutritional status in terms of body weight were recorded at this session.

Meal sizes were compared to the preoperative situation and classified as ‘‘equal,’’ ‘‘increased, ’’or ‘‘decreased.’’ Meal frequency was defined as the number of meals, snacks, and/or beverages a patient consumed during a 24 h period (day and night). These meals were categorized as <3 times, 3 times, 4–6 times, or >6 times per 24 h.

Appendix 1 The 15 item feeding questionnaire

No. Did you experience last week …Section 1. Dysphagia and reflux1 … pain during swallowing2 … swallowing the wrong way3 … the viscosity inhibits the passage around the cervical suture4 … a lump in your throat5 … the need to retch as food sticks around the cervical suture6 … food returning to your mouth after swallowing as it sticks around the cervical suture7 … reflux of food and / or stomach contain towards your mouth8 … a painful feeling and / or a feeling of burning behind your breast-bone

Section 2. Stomach content and dumping syndrome9 … early satiety10 … absence of hunger11 … food returning to your mouth after eating too much12 … nausea13 … vomiting14 … one or more complaints an hour to 1 hour after a meal a

15 … one or more complaints 1–2 hours after a meal b

Items were scored in terms of presence (1 point) or absence (0 point).

a For example, the need to rest, belly pain, nausea, dizziness, exhausting, blushing, diarrhea, palpitation of the heart, sweatingb For example, paleness, hunger, restless, dizziness, headache, yawning, palpitation of the heart, sweating, trem-bling

98

The social aspects related to eating and drinking were evaluated by three questions: (1) experiencing fear of eating or drinking, (2) experiencing pleasure in eating and drink-ing, and (3) going out to dinner (to a restaurant or to family/friends). Fear was categorized as ‘‘never,’’ ‘‘seldom,’’ ‘‘often,’’ or ‘‘always.’’ Pleasure in eating and drinking and going out to dinner were compared to the preoperative situation and were categorized as ‘‘no(t),’’ ‘‘less,’’ ‘‘equal,’’ or ‘‘more.’’

The defecation pattern was compared to the preoperative situation. It was divided into frequency (<1 stool, 1 stool, 2 stools, >2 stools [including frequency]) within a 24-h period and consistency (thin as water, slush, solid, or other [including description]).

In order to calculate weight loss, patients were asked to estimate their weight one month prior to each study assessment.

Statistical analyses Baseline patient characteristics, physical status, surgery related characteristics, the pres-ence of nutrition-related symptoms, intake of enteral nutrition (tube feeding and sip feed-ing), meal size, meal frequency, body weight, nutrition-related social aspects, and defeca-tion patterns were summarized using descriptive statistics.

The Kolmogorov-Smirnov test was used to test whether the patient-related and surgery-related characteristics were normally distributed. Differences between pro-portions and mean scores were analyzed with the X2 test or the two group t-test when appropriate.

The repeated data structure of the nutrition-related complaint scores (both total and subscale scores) were analyzed with a linear mixed model (LMM). In this approach we included all complaint scores as measured at the various follow-up visits. Hence, in case of relapse of the malignant disease or death, the data of the patients (before these events) were also included in the mixed models. The LMM was repeated after including the baseline patient and surgery-related characteristics (age, sex, preoperative tube feed-ing, neo-adjuvant treatment, surgical procedure, and postoperative complications) that were assumed to predict the complaint scores.

The course of body weight during the first postoperative year was also evaluated by a linear mixed model with and without the complaint scores as covariate. All analyses were executed in SPSS version 16.0 (SPSS Corp. Chicago IL).

Results

During the period of the study (2002–2005), 140 patients underwent an esophagectomy with gastric tube reconstruction. Of these patients, 134 (96%) were eligible for the study. In total, 96 patients took part in the study. Sixty of them (63%) completed the one-year follow-up. Forty-one patients filled out the questionnaire at all time points during the first postoperative year (Figure 1). The baseline characteristics of the 96 patients who took part in the study are summarized in Table 1.

99

Cha

pte

r 6Pr

esen

ce a

nd p

ersi

sten

ce o

f nut

ritio

n-re

late

d sy

mpt

oms d

urin

g th

e fir

st y

ear f

ollo

win

g es

opha

gect

omy

with

gas

tric

tube

Patient and surgery-related characteristics did not significantly differ between the patients who completed the study (n = 60) and those who dropped out (n = 36) during the one-year follow-up.

Nutrition-related symptomsOn average, patients suffered from seven nutrition-related complaints during the four fol-low-up moments (Table 2). Linear mixed modeling showed no significant within-patient change in the number of symptoms over time for the total population (P = 0.67). There was

Figure 1 Study flow chart Figure 1 Study flow chart

4 lost to follow up 7 were excluded from further analyses because of relapse of the disease

1 week after the introduction oral foods / drinks:

96 participants 80 patients filled out the questionnaire

Esophagectomy with gastric tube reconstruction in 140 patients

3 months after surgery : 85 participants 76 patients filled out the questionnaire

6 months after surgery : 76 participants 69 patients filled out the questionnaire

12 months after surgery : 60 participants 59 patients filled out the questionnaire

6 exclusions due to criteria 1 suffering of a neurological disease 1 severe chyle leakage 4 not speaking the Dutch language

34 declined to participate 4 died after signing written informed consent but before entering the study

2 lost to follow up 6 were excluded from further analyses because of relapse of the disease 1 died

1 was excluded from further analyses because of another life -threatening disease 15 were excluded from further analyses

100

Table 1 Baseline characteristics of the patient group (N = 96)

Total population(N = 96)

Completers one year follow-up (n = 60)

Non- completers(n = 36)

Completers vs Non- completers (P value)

Patient-related characteristics

Age, years a 62 (10) 60 (10) 63 (10) 0.27

Male gender b 73 (76) 48 (80) 25 (69) 0.23

Presence of malignancy b 93 (97) 58 (97) 35 (97) 1.00

Neoadjuvant treatment b 1.00

Chemotherapy 11 (11) 6 (10) 5 (14)

Chemotherapy and radiotherapy 18 (19) 12 (20) 6 (17)

Admission duration a 20 (8) 19 (9) 20 (7) 0.64

Physical status-related characteristics

Preoperative body weight a 79.0 (15.8) 79.9 (16.2) 78.3 (15.3) 0.56

Preoperative weight loss 49 (51) 33 (55) 16 (44) 0.40

BMI a 26.0 (3.9) 26.1 (3.9) 26.2 (4.2) 0.80

Underweight BMI< 18.5 < 65 y years

4 (4) 4 (7) 0 (0)

Underweight BMI < 20.0 > 65 years 1 (1) 0 (0) 1(3)

Obesity BMI > 30 13 (14) 9 (15) 4 (11)

Co-morbidity b

Cardiovascular diseases 46 (48) 27 (45) 19 (53) 0.80

Pulmonary diseases 21 (22) 13 (22) 9 (25) 0.80

Diseases of the urinary tract 19 (20) 13 (22) 6 (17) 0.79

Renal diseases 6 (6) 5 (8) 1 (3) 0.41

Diabetes mellitus 6 (6) 3 (5) 3 (8) 1.00

Preoperative complaints b

Dysphagia 69 (72) 41 (68) 28 (78 0.49

Epigastric/retrosternal pain 24 (25) 14 (23) 10 (28) 0.47

Hiccups 9 (9) 4 (7) 5 (14) 0.72

Anorexia 2 (2) 1 (2) 1 (3) 1.00

Preoperative tube-feeding 7 (7) 5 (8) 2 (6) 1.00

Surgery-related characteristics

ASA classification b

ASA 3 20 (21) 10 (17) 10 (28) 0.33

ASA 4 1 (1) 1 (2)

Surgical procedure b 0.73

Transthoracic esophageal resec-tion

50 (52) 32 (53) 18 (50)

Transhiatal esophageal resection 46 (48) 28 (47) 18 (50)

Postoperative complications b 60 (63) 35 (58) 25 (69) 0.27

BMI body mass index, ASA American Society of Anesthesiologists scorea Mean (SD)b n (%)

101

Cha

pte

r 6Pr

esen

ce a

nd p

ersi

sten

ce o

f nut

ritio

n-re

late

d sy

mpt

oms d

urin

g th

e fir

st y

ear f

ollo

win

g es

opha

gect

omy

with

gas

tric

tube

also no time related change in the two subscale scores concerning dysphagia/reflux (P = 0.80) and stomach content/dumping syndrome (P = 0.71).

In addition, none of the baseline factors appeared to be a significant predictor for the complaint scores: age (P = 0.84), sex (P = 0.20), preoperative tube-feeding (P = 0.54), neo-adjuvant treatment (P = 0.09), surgical procedure (P = 0.14), and postoperative complica-tions (P = 0.78).

The five most frequent nutrition-related symptoms experienced by the patients were early satiety, postprandial dumping syndrome, inhibited passage due to high viscosity, reflux, and absence of hunger (Table 3). During the entire postoperative year a large number (about 90%) of the patients experienced early satiety, whereas about 75% of the patients suffered from postprandial dumping syndrome. About 50% of the study population experienced hunger one year after the operation. There was no difference between the type of symptoms in patients who completed the study and those who dropped out.

Table 2 Mean number (±SD) of nutrition-related complaints and average body weight of the patients mea-sured at four time points during the one-year follow-up

1 week(n = 80)

3 months(n = 76)

6 months(n = 69)

12 months(n = 59)

P Value a

Total sum score of complaints b 7.0 (3.0) 7.3 (2.9) 7.4 (3.1) 7.4 (3.6) 0.67

Subscale

Dysphagia and reflux c 3.3 (1.9) 3.4 (2.0) 3.5 (2.0) 3.5 (2.4) 0.80

Stomach content and umping syndrome d 3.7 (1.5) 3.9 (1.5) 3.8 (1.7) 3.9 (1.8) 0.71

Body weight 74.3 (13.7) 74.2 (12.7) 74.7 (12.1) 73.9 (12.7) 0.47

a P values based on linear mixed modelsb Range total subscore: 0 points (complaints free) to 15 points (suffering from all complaints)c Subscale dysphagia and reflux: 0 points to 8 points (suffering from all complaints)d Subscale stomach content and dumping syndrome: 0 points to 7 points (suffering from all complaints)

Table 3 Percentage of patients who experienced the five most frequent nutrition-related complaints during the one-year follow-up

Complaint a 1 week(n = 80)

3 months(n = 76)

6 months(n = 69)

12 months(n = 59)

Early satiety 71 (89) 66 (87) 60 (87) 53 (90)

Postprandial dumping syndrome 59 (74) 59 (78) 54 (78) 44 (75)

Inhibited passage due to high viscosity 42 (53) 48 (63) 41 (59) 37 (63)

Reflux of food/fluid 48 (60) 41 (54) 45 (65) 36 (61)

Absence of hunger 61 (76) 43 (57) 39 (57) 30 (51)

a All values are expressed as n (%)

102

Nutrition-related adjustments: enteral nutritionAfter discharge from the hospital, tube feeding was continued in 38 of the 80 patients (48%) who filled out the questionnaire during the first week after discharge. Their mean intake was 990 ml (range: 500–2,000 ml) and was mainly administered (79%) during the night. Six months after surgery only two patients still used nightly tube feeding to achieve their nutritional goals, and 12 months after surgery there was one such patient.

Sip feeding was started in 25% of the study population after discharge, and 12 months after surgery it was still used by 22%. Mean daily consumption of sip feeding was 300 ml (range: 70–600 ml) during the first 3 postoperative months and remained 230 ml (range: 40–400 ml) at 12 months.

Nutrition-related adjustments: meal sizes and meal frequency Meal sizes were reduced compared to preoperative measurements according to 99% of the patients the first week after discharge; 12 months after surgery 92% of the patients still reported eating smaller meals compared to the preoperative situation. In addition, the number of meals remained high during the entire postoperative year; 12 months after the operation 58% of the patients had 3–6 meals a day, and 27% consumed as many as 6–9 snacks and small meals.

Nutrition-related adjustments: social aspects One year after surgery, 16% of the study population still experienced eating or drinking in the company of others as unpleasant, and more than 50% of the patients went out less often to eat at a restaurant or with family or friends. The same proportion experienced a range from less pleasure to no pleasure in eating and drinking in general.

Nutrition-related adjustments: stool consistency and stool frequencyWith regard to stool consistency, 81% of the patients defined their stool consistency before the surgical procedure as ‘‘solid.’’ Only 2% defined it as being ‘‘slush.’’

Although the consistency at 12 months was defined as ‘‘solid’’ by 48% of the patients, ‘‘slush’’ by 22%, and as ‘‘variable’’ by 30% (solid–slush, slush–thin as water), no significant change in stool consistency was demonstrated (P 0.17). In this group of survivors of the first postoperative year, their preoperative and postoperative stool consistencies were not significantly different.

Before surgery, a stool frequency of 1 stool per day was common in 65% of the total study population, 4% had >2 stools and 10% had <1. One year after surgery, 45% of the remaining participants had one stool per day, 8% had >2 per day, and almost 30% had <1 per day. Patients who survived the first postoperative year showed a significant alteration in postoperative stool frequency (increase or decrease) compared to the preoperative fre-quency (P < 0.001).

103

Cha

pte

r 6Pr

esen

ce a

nd p

ersi

sten

ce o

f nut

ritio

n-re

late

d sy

mpt

oms d

urin

g th

e fir

st y

ear f

ollo

win

g es

opha

gect

omy

with

gas

tric

tube

Nutritional status: body weightAt the time of discharge from the hospital, 75 patients (78%) had lost body weight (mean: 6.2 ± 5.6 kg) compared to their preoperative body weight. During the entire postoperative year, the mean body weight of the total population remained stable at a reduced level (P = 0.47) (see Table 2). No significant association between the complaint scores and body weight was observed (P = 0.15).

Discussion

This prospective longitudinal cohort study shows that patients who underwent an esophagectomy with gastric tube reconstruction suffered from a number of persistent, nutrition-related symptoms during the entire first postoperative year. Early satiety, post-prandial dumping, inhibited passage due to high viscosity, reflux of food and/or fluids, and the absence of hunger were the most frequently reported nutrition-related symptoms. We demonstrated that time (in this case the first 12 postoperative months) is unrelated to the number of nutrition-related symptoms. The persistent symptoms could not be explained by a range of patient or surgery-related characteristics.

The large majority of patients ate smaller meals with a relatively high frequency dur-ing the first year of follow-up. After one year, the surgical procedure still influenced the nutrition-related social aspects of eating, and patients also experienced an altered stool frequency. A reduction of body weight occurred directly after the surgical procedure, and the majority of patients were unable to return to their preoperative weight. The weight reduction could not be explained by the nutrition-related complaints.

Our findings differ from those reported in the literature; we observed a number of per-sistent complaints, while other studies described a decrease in the nutrition-related symp-toms that were experienced. 6, 7, 13, 14, 26 These differences could be explained by the design and execution of our study. Our primary aim was to evaluate the nutrition-related symp-toms that patients experienced. In other studies, these nutritional aspects were secondary and were therefore not described or studied in detail. In some studies, retrospective data were used, or data were collected through postal surveys, without patient contact. 8, 11, 22,

27, 28 If the symptoms of postoperative patients were assessed by caregivers, only a maxi-mum of two nutrition-related symptoms were scored. 1, 7, 9 In addition, the symptoms were evaluated only once or twice during the first postoperative year, at non-comparable time points, and mainly by using generic quality-of-life instruments (RSC,EORTC QLQ-C30, EOS 24, Moss SF-20, SF-36). These multidimensional measures assess various health domains and do not specifically focus on nutrition-related symptoms. 7–9, 11, 14, 15, 22, 27, 28

We developed a questionnaire to evaluate the experienced nutrition-related symp-toms prospectively at specific time points during the first postoperative year. After the questionnaires were completed, the responses were always discussed with the patients, and unclear answers or contradictions were clarified.

Another explanation for the difference in findings compared to the literature could be

104

that, with our instrument, questions could only be answered by yes (presence of symp-tom) or no (absence of symptom), making a gradual alteration of a symptom less obvious. Some of the generic quality-of-life instruments have scaled answer options, making small changes in the experienced nutrition-related symptoms clearer.

A relationship between the nutrition-related symptoms and the type of surgery has been suggested in other studies. 6, 7, 10, 27, 29 As we expected, we found no difference in the number and severity of symptoms between patients who underwent a transhiatal or a transthoracic procedure, making this explanation doubtful in our population.

The findings of Ryan et al. 16, who described an insufficient oral intake at discharge in 60% of their patients after esophagectomy, was confirmed in our study. However, in our study, almost 75% of patients were dependent on enteral nutrition after discharge at some point in time. To our knowledge, our study is the first that has demonstrated the continuing need for nutritional support during a long postoperative period, when early satiety inhibits the consumption of sufficient quantities of food. In accordance with the literature, our patients tended to eat smaller but more frequent meals to compensate for this phenomenon; in cases where tube feeding was mainly used during the first 3 postoperative months, over 20% of the study population still needed sip feeding 2 months after the operation to achieve their nutritional targets. 11

Our research does not support the hypothesis that tube feeding inhibits oral intake, suppresses appetite, and increases satiety in patients after esophagectomy. Stratton et al. showed that tube feeding did not reduce oral intake significantly and had only a slightly negative influence on appetite. 30 According to our hypothesis, the reduced intake and early satiety after esophagectomy are mainly caused by the surgical recon-struction itself.

Although preoperative loss of body weight did occur in 50% of our study population, the mean preoperative BMI was still 26.0 (± 3.9). This finding was also reported by Steyn et al., who noted that preoperative weight loss in patients with esophageal carcinoma did not result per se in underweight, due to the prevalence of overweight or obesity in the Western world. 31 Nevertheless, we must be vigilant with regard to undesired preoperative weight loss in this population, as it associated with postoperative morbidity and therefore needs to be identified and treated at the earliest possible preoperative stage. 18–21

After surgery, both patients and caregivers are focused on undesired weight loss and/or fluctuations in body weight. 8, 10–12, 14, 25, 32 In the present study, we showed that weight loss occurs directly after surgery, followed by a stable reduced postoperative weight. Similar to the findings of Moraca and McLarty, only 25% of our population returned to their preoperative weight one year after surgery. 8, 11 This may explain why the five preop-eratively malnourished patients in this study were incapable of gaining weight postopera-tively and remained malnourished during the entire postoperative year.

Our follow-up time was probably not long enough to show adaptation, a reduction of nutrition-related symptoms, and an increase in body weight. Although it is assumed that a physically and emotionally stable situation is achieved 6 months after the operation, it

105

Cha

pte

r 6Pr

esen

ce a

nd p

ersi

sten

ce o

f nut

ritio

n-re

late

d sy

mpt

oms d

urin

g th

e fir

st y

ear f

ollo

win

g es

opha

gect

omy

with

gas

tric

tube

is our clinical experience that an actual maximum physical status occurs only 2 or 3 years after the surgery. 7, 13

Studies in patients with esophageal malignancies are invariably handicapped by high mortality and relapse rates of the study subjects; about 40% of the patients dropped out of our study mainly due to recurrence, and ultimately a substantial number of these patients died. However, these rates are comparable with those in the literature, where dis-ease recurrence affects 30% of the patients during the first postoperative year. 7

Regarding our research design, because a substantial number of patients with incomplete follow-up could introduce methodological flaws, we investigated the nutrition-related com-pliant scores and body weights within the framework of a linear mixed regression technique.

To our knowledge, this is the first study describing nutrition-related symptoms of patients, adjustments, and body weight at multiple, specific time points during a one-year follow-up period after esophagectomy. However, it is unknown whether suffering from a serious number of nutrition-related symptoms after this operation leads to altered food choices and indirectly influences the intake of macro- and micronutrients. Therefore, a study to address the ability of patients to reach their recommended daily intakes of nutri-ents is being performed.

In conclusion, the present study shows that in the first year after an esophagec-tomy with gastric tube reconstruction, the majority of patients struggle with persistent nutrition-related symptoms, nutrition-related adjustments in terms of meal size, meal frequency, nutrition-related social aspects, and altered stool frequency. They must also struggle to achieve a sustained body weight. Therefore, at specific postoperative time points, both surgeon and dietitian should inform the patient about the occurrence of postoperative nutrition-related symptoms, which could be persistent. In addition, they should also systematically assess the specific symptoms of each patient, and evaluate the nutritional status of that patient in terms of body weight, to improve the patient’s quality of life and to prevent malnutrition.

Acknowledgments The authors gratefully acknowledge Joyce T. Haver, RD, Department of Dietetics of the Aca-de mic Medical Centre, University of Amsterdam, for her help and support during the study.

References1. O’Rourke I, Tait N, Bull C et al (1995) Oesophageal cancer: outcome of modern surgical management. Aust N

Z J Surg 65: 11–16.2. Bartels H, Stein HJ, Siewert JR (1998) Preoperative risk analysis and postoperative mortality of oesophagectomy

for resectable oesophageal cancer. Br J Surg 85:840–844.3. Parshad R, Singh RK, Kumar A et al (1999) Adenocarcinoma of distal esophagus and gastroesophageal

junction: long-term results of surgical treatment in a North Indian Center. World J Surg 23: 277–283.4. Lerut T, Coosemans W, Decker G et al (2004) Extended surgery for cancer of the esophagus and

gastroesophageal junction. J Surg Res 117:58–63.

106

5. Lagarde SM, de Boer JD, ten Kate FJ et al (2008) Postoperative complications after esophagectomy for adenocarcinoma of the esophagus are related to timing of death due to recurrence. Ann Surg 247:71–76.

6. Finley FJ, Lamy A, Clifton J et al (1995) Gastrointestinal function following esophagectomy for malignancy. Am J Surg 169: 471–475.

7. Wainwright D, Donovan JL, Kavadas V et al (2007) Remapping the body: learning to eat again after surgery for esophageal cancer. Qual Health Res 17:759–771.

8. Moraca RJ, Low DE (2006) Outcomes and health-related quality of life after esophagectomy for high-grade dysplasia and intramucosal cancer. Arch Surg 141:545–549.

9. Blazeby JM, Williams MH, Brookes ST et al (1995) Quality of life measurement in patients with oesophageal cancer. Gut 37:505–508.

10. Gawad KA, Hosch SB, Bumann D et al (1999) How important is the route of reconstruction after esophagectomy: a prospective randomized study. Am J Gastroenterol 94:1490–1496.

11. McLarty AJ, Deschamps C, Trastek VF et al (1997) Esophageal resection for cancer of the esophagus: long-term function and quality of life. Ann Thorac Surg 63:1568–1572.

12. De Leyn P, Coosemans W, Lerut T (1992) Early and late functional results in patients with intrathoracic gastric replacement after oesophagectomy for carcinoma. Eur J Cardiothorac Surg 6:79–84.

13. Ludwig DJ, Thirlby RC, Low DE (2001) A prospective evaluation of dietary status and symptoms after near-total esophagectomy without gastric emptying procedure. Am J Surg 181: 454–458.

14. Baba M, Aikou T, Natsugoe S et al (1997) Appraisal of ten-year survival following esophagectomy for carcinoma of the esophagus with emphasis on quality of life. World J Surg 21:282–285.

15. Conroy T, Marchal F, Blazeby JM (2006) Quality of life in patients with oesophageal and gastric cancer: an overview. Oncology 70:391–402.

16. Ryan AM, Rowley SP, Healy LA et al (2006) Post-oesophagectomy early enteral nutrition via a needle catheter jejunostomy: 8-year experience at a specialist unit. Clin Nutr 25:386–393.

17. Page RD, Oo AY, Russell GN et al (2002) Intravenous hydration versus naso-jejunal enteral feeding after esophagectomy: a randomised study. Eur J Cardiothorac Surg 22:666–672.

18. Baldwin C, Weekes CE (2008) Dietary advice for illness-related malnutrition in adults. Cochrane Database of Systematic Reviews Issue 1. Art. No.: CD002008. doi:10.1002/14651858.CD002008. pub3.

19. Wilson MMG, Vaswani S, Liu D et al (1998) Prevalence and causes of undernutrition in medical outpatients. Am J Med 104:56–63.

20. Stratton RJ, Green CJ, Elia M (2003) Disease related malnutrition: an evidence-based approach to treatment, 1st edn. CABI Publishing, Cambridge.

21. Guideline Perioperative Nutrition. Dutch Institute for Healthcare Improvement CBO, Utrecht (2007) http://www.cbo.nl/product. richtlijnen/folder2002102312843/rl_periovoed_07.pdf.

22. De Boer AG, van Lanschot JJ, van Sandick JW et al (2004) Quality of life after transhiatal compared with extended transthoracic resection for adenocarcinoma of the esophagus. J Clin Oncol 22:4202–4208.

23. Hulscher JB, Tijssen JG, Obertop H et al (2001) Transthoracic versus transhiatal resection for carcinoma of the esophagus: a meta-analysis. Ann Thorac Surg 72:306–313.

24. Hulscher JB, van Sandick JW, de Boer AG et al (2002) Extended transthoracic resection compared with limited transhiatal resection for adenocarcinoma of the esophagus.NEngl JMed 347:1662–1669.

25. Cense HA, Visser MR, van Sandick JW et al (2004) Quality of life after colon interposition by necessity for esophageal cancer replacement. J Surg Oncol 88:32–38.

26. Hölscher AH, Voit H, Buttermann G et al (1988) Function of the intrathoracic stomach as esophageal replacement. World J Surg 12:835–844.

27. Viklund P, Lindblad M, Lagergren J (2005) Influence of surgeryrelated factors on quality of life after esophageal or cardia cancer resection. World J Surg 29:841–848.

28. van Knippenberg FC, Out JJ, Tilanus HW et al (1992) Quality of life in patients with resected oesophageal cancer. Soc Sci Med 35:139–145.

29. Nakamura M, Kido Y, Hosoya Y et al (2007) Postoperative gastrointestinal dysfunction after 2-field versus 3-field lymph node dissection in patients with esophageal cancer. Surg Today 37:379–382.

30. Stratton RJ, Stubbs RJ, Elia M (2003) Short-term continuous enteral tube feeding schedules did not suppress appetite and food intake in healthy men in a placebo-controlled trial. J Nutr 133: 2570–2576.

31. Steyn RS, Grenier I, Darnton SJ et al (1995) Weight gain as an indicator of response to chemotherapy for oesophageal carcinoma. Clin Oncol (R Coll Radiol) 7:382–384.

32. Nishihira T, Watanabe T, Ohmori N et al (1984) Long-term evaluation of patients treated by radical operation for carcinoma of the esophagus. World J Surg 8:778–785.

Chapter 7

Suboptimal intake of nutrients after esophagectomy with gastric tube reconstruction

E.B. HaverkortJ.M. Binnekade

R.J. de Haan O.R.C. Busch

M.I. van Berge HenegouwenD.J. Gouma

J Acad Nutr Diet. 2012 Jul;112(7):1080-7

108

Abstract

Esophagectomy with gastric tube reconstruction results in a variety of postoperative nutrition-related complaints that may impair nutritional intake and nutritional status. The aim of this study was to determine to what extent patients reached the recommended intake of various nutrients at six and twelve months after esophagectomy. It was also ana-lyzed whether a suboptimal intake could be explained by the most clinically significant nutrition related complaints after esophagectomy. In a prospective cohort study (2002-2006), the nutrient intake of 96 patients, recorded in preprinted nutritional diaries, was compared with the recommended energy intake in The Netherlands and Recommended Daily Allowance of protein and micronutrients. Energy and protein intake remained below recommendations in 24% and 7% of the patients, respectively. Less than 10% of the patients had a sufficient intake of all micronutrients. Folic acid, vitamin D, copper, calcium and vitamin B-1 were the micronutrients most often reported to have a suboptimal intake. Multivariate logistic regression, corrected for preoperative epigastrical pain and energy intake, showed that the number of nutrition-related complaints was not an independent risk factor for the presence of a suboptimal intake of nutrients (adjusted odds ratio = 1.11; 95% CI 0.94 to 1.31; P = 0.22). This study shows that the intake of micronutrients remains below recommendations in the majority of patients twelve months after esophagectomy. This problem requires special attention and care by dietitians. Key wordsNutrition; Esophagectomy; Nutrients; Vitamins; Minerals.

109

Cha

pte

r 7Su

bop

tim

al in

take

of n

utri

ents

aft

er e

sop

hag

ecto

my

wit

h g

astr

ic tu

be

reco

nstr

ucti

on

Introduction

In patients who undergo a resection of the esophagus or cardia (esophagectomy), the pre-ferred reconstruction is a gastric tube. 1-8 This procedure results in severe changes in the gas-trointestinal anatomy (eg, the cardia is resectioned and the remaining part of the esopha-gus is reconnected to the stomach) and results in nutrition-related complaints such as early satiety, postprandial dumping syndrome, dysphagia, reflux, absence of hunger, altered stool frequency and fluctuations in body weight. 8, 9 – 20 A previous study indicated that these are persistent complaints. 21 In contrast to findings reported in the literature, at twelve months postoperative, the majority of patients in this study still experienced nutrition-related com-plaints. These complaints can result in a persistently different or adapted nutritional intake, which can negatively influence patients’ nutritional status and, in turn, their recovery. 22, 23

No studies have yet determined whether the altered anatomy actually affects nutrient intake after an esophagectomy. During nutrition-related follow-up, registered dietitians and surgeons focus mainly on counseling to maintain body weight. 11, 13 – 15, 17, 19, 24 It was previously shown that approximately three quarters of the patients had lost weight – six kg on average – at discharge from the hospital, relative to their preoperative body weight. 21 However, bodyweight by itself does not give any information about adequate intake of proteins and/or micronutrients (ie, vitamins, minerals and trace-elements). Watanabe and colleagues 25 assumed that adequate protein intake also implied adequate micronutrient intake. However, no studies are available on either protein intake or micronutrient intake after esophagectomy, so this assumption could be erroneous. Nutrient deficiencies result in a broad range of physical and psychological symptoms. It has been demonstrated that a long-term suboptimal intake negatively affects body-weight, nutritional status, and quality of life. 11, 13 – 17, 19, 20, 22 – 24, 26 Symptoms of nutrient defi-ciency may not be evident to registered dietitians, surgeons and other caregivers. 23

If it can be determined to what extent the nutrient intake of esophagectomy patients does not meet the recommendations as defined by the Health Council of the Netherlands (described in Figure 1, item Definition of malnutrition) 27-30 then a protocol can be drafted to improve postoperative nutritional care, ultimately resulting in improved recovery. In the present study, the following questions where therefore addressed: Do patients reach the intake recommendations for nutrients (ie, energy, proteins, micronutrients) at six and twelve months after esophagectomy with gastric tube reconstruction? If an inadequate intake occurs, which nutrients are most frequently suboptimal in the diet? Can nutrition-related complaints explain a serious suboptimal intake of nutrients?

Methods

Setting and patients A prospective cohort study was conducted at the Academic Medical Center, Amsterdam, The Netherlands (2002-2006). This is a tertiary care university-affiliated hospital with 1,000

110

beds, specialized in the treatment of gastrointestinal oncological diseases (ie, esophagus, pancreas, bile-tract, liver, and colon). Consecutive patients 18 years of age and older who underwent an esophagectomy were invited to participate in this study, which was approved by the Institutional Review Board of the hospital. All patients gave their written informed consent. Exclusion criteria included inability to speak and/or read Dutch, suffering from dia-betic-related neuropathy, neurological disease, ulcerative colitis, Crohn or celiac disease, severe postoperative chyle leakage, and proven allergy. Because recurrence of malignant disease can result in loss of appetite, reduced intake, increased nutrition-related com-plaints and unwanted weight loss, patients with a proven recurrence of the disease were also excluded from further analysis. 1, 31

Baseline assessmentsBaseline characteristics (ie, age, sex, presence of malignancy, neo-adjuvant treatment), physical status (ie, preoperative body weight, body mass index, comorbidity, preopera-tive nutritional complaints) and surgery-related characteristics (ie, American Society of Anesthesiologists [ASA] classification defining preoperative fitness, surgical procedure, postoperative complications, and admission duration) were collected from medical and dietetic records.

Prescription of proteins and energy The postoperative nutritional care protocol is described in detail in Figure 1. To preserve body cell mass, 1.5-1.7 grams of protein/kg/24 hours during the first six months postoper-ative was prescribed. 32-37 After this period, the protein prescription was reduced to 1.2–1.3 grams/kg/24 hours, because patients were expected to be less catabolic by then. Energy requirements were calculated with the Harris and Benedict equation plus 30% for meta-bolic stress and activity. 38, 39 Stable body weight is important because weight loss >10% implies loss of body cell mass, including muscle mass. 35, 36, 40-42

Tube feeding and oral foods All postoperative patients were fed by polymeric tube-feeding during the first week post-operative. At day seven postoperative, if anastomotic leakage was excluded, food was gradually introduced (liquid meals for one day, solid foods thereafter). If the oral intake was insufficient at discharge, tube-feeding was continued or sip-feed-ing introduced. Detailed information about modification and discontinuation of tube-feeding and sip-feeding is also shown in Figure 1.

Follow-up after discharge and evaluation of intakeAfter discharge, patients visited the departments of surgery and dietetics of the outpa-tient clinic at least once every three months during the entire first postoperative year. At three months postoperative, patients were instructed how to fill out the pre-printed

111

Cha

pte

r 7Su

bop

tim

al in

take

of n

utri

ents

aft

er e

sop

hag

ecto

my

wit

h g

astr

ic tu

be

reco

nstr

ucti

on

Fig

ure

1 P

roto

col p

osto

per

ativ

e nu

triti

onal

car

e an

d p

roce

dure

s in

pat

ient

s af

ter e

sop

hage

ctom

y w

ith g

astr

ic tu

be

reco

nstr

uctio

n

Item

ref

eren

ce

In g

ener

al

Car

e d

uri

ng

the

stu

dy

per

iod

Ad

apte

d c

are,

intr

od

uce

d a

fter

th

e st

ud

y

Pro

tein

, en

erg

y an

d

mic

ron

u-

trie

nts

p

resc

rip

tio

n

27-3

0, 3

2-39

Pres

crip

tion

of p

rote

ins

≤ 6

mo

pos

top

erat

ive:

1.5

- 1.

7 g

of p

ro-

tein

/kg

bod

y w

eigh

t/24

h.

Pres

crip

tion

of p

rote

ins

> 6

mon

ths

pos

top

erat

ive:

1.2

– 1

.3 g

of

pro

tein

/kg

bod

y w

eigh

t/24

h.

In c

ase

of o

verw

eigh

t or o

bes

ity:

pro

tein

pre

scrip

tion

is a

dap

ted

and

calc

ulat

ed b

ased

on

BMI a

= 2

7.Pr

escr

iptio

n of

ene

rgy:

Har

ris a

nd B

ened

ict e

quat

ion

(198

4) +

30%

(2

0% d

isea

se +

10%

act

ivit

y).

RD b

: Cal

cula

tion

of p

rote

in a

nd e

nerg

y ne

ed.

Pres

crip

tion

of m

icro

nutr

ient

s: re

-co

mm

enda

tions

as

defin

ed b

y th

e H

ealt

h C

ounc

il of

the

Net

herl

ands

. RD

: Cal

cula

tion

of th

e m

icro

nu-

trie

nts

need

s.

Tub

e fe

e-d

ing

Su

rgeo

n: S

urgi

cally

pla

ced

need

le je

juno

stom

y fe

edin

g-tu

be.

RD

: Cal

cula

tion

of p

olym

eric

tub

e-fe

edin

g ac

cord

ing

to p

rote

in

and

ener

gy n

eed.

N

urse

: Day

1 p

osto

per

ativ

e: 4

0 m

L/h,

incr

ease

d w

ith 2

0 m

L/h

ever

y 6-

8 h

until

requ

irem

ents

wer

e m

et. I

n ca

se o

f int

estin

al c

omp

lain

ts

(eg,

cra

mp

s, p

ain,

nau

sea)

: no

incr

ease

or t

emp

orar

y de

crea

se o

f en

tera

l pum

p fo

r 1-2

h, t

hen

cont

inua

tion

acco

rdin

g to

pro

toco

l.

Intr

od

uc-

tio

n o

f ora

l fo

od

s af

ter

surg

ery

Radi

olog

ist a

nd s

urge

on: D

ay 7

pos

top

erat

ive

eval

uatio

n an

a-st

omot

ic le

akag

e by

wat

er s

olub

le c

ontr

ast x

-ray

; no

leak

age:

in

trod

uctio

n of

liqu

id fo

ods.

RD

: Day

8 p

osto

per

ativ

e: in

trod

uctio

n of

sol

id fo

ods.

Pro

tein

- and

en

ergy

-ric

h ad

vice

, 6-8

tim

es/d

.A

t dis

char

ge: O

ral i

ntak

e ≤

50%

pre

scrib

ed e

nerg

y an

d/or

pro

tein

am

ount

: mai

ntai

ning

tub

e-fe

edin

g. O

ral i

ntak

e >

50%

- <

100%

of

pre

scrib

ed e

nerg

y an

d/or

pro

tein

am

ount

: int

rodu

ctio

n si

p-f

ee-

ding

(pro

tein

or e

nerg

y en

riche

d de

pen

ding

on

need

s).

Follo

w-

up

aft

er

dis

char

ge

In g

ener

al: (

Com

bin

ed) v

isits

to th

e ou

tpat

ient

clin

ic d

epar

tmen

ts

of s

urge

ry a

nd d

iete

tics

ever

y 3

mo

first

pos

top

erat

ive

year

. RD

: In

case

of t

ube-

feed

ing

afte

r dis

char

ge: e

valu

atio

n or

al in

take

an

d tu

be-

feed

ing

ever

y ot

her w

eek

(by

tele

pho

ne).

Surg

eon:

For

eva

luat

ion

of p

hysi

cal r

ecov

ery

and

in re

spon

se to

co

mp

lain

ts (e

g, p

ain,

dys

pha

gia,

gas

tro-

inte

stin

al c

omp

lain

ts):

m

edic

al e

xam

inat

ions

(eg,

dila

tion,

CT

c , MRI

d ) and

/or t

reat

men

t (e

g. p

resc

riptio

n of

med

icat

ion)

. H

ome

care

nur

se (I

f nec

essa

ry o

r des

ired)

: ass

istin

g p

atie

nt b

y ad

min

iste

ring

tub

e-fe

edin

g.

RD: 3

mo

pos

top

erat

ive:

inst

ruct

ion

on h

ow

to fi

ll ou

t pre

-prin

ted

nutr

ition

al d

iary

. 6

and

12 m

onth

s p

osto

per

ativ

e: e

valu

atio

n of

p

re-p

rinte

d nu

triti

onal

dia

ry.

RD: E

valu

atio

n an

d ca

lcul

atio

n of

in

take

(mac

ro a

nd m

icro

nutr

ient

s),

eval

uatio

n of

bod

y w

eigh

t (ga

in o

r lo

ss),

and

BMI.

Prac

tical

adv

ice

on m

acro

nutr

ient

s an

d m

icro

nutr

ient

s an

d p

osto

per

a-tiv

e nu

triti

on-r

elat

ed c

omp

lain

ts.

Surg

eon:

Info

rmat

ion

on s

urge

ry

and

pos

top

erat

ive

nutr

ition

-rel

ated

co

mp

lain

ts.

112

Fig

ure

1 P

roto

col p

osto

per

ativ

e nu

triti

onal

car

e an

d p

roce

dure

s in

pat

ient

s af

ter e

sop

hage

ctom

y w

ith g

astr

ic tu

be

reco

nstr

uctio

n

Item

ref

eren

ce

In g

ener

al

Car

e d

uri

ng

the

stu

dy

per

iod

Ad

apte

d c

are,

intr

od

uce

d a

fter

th

e st

ud

y

Red

uct

ion

o

f en

tera

l n

utr

itio

n

RD: M

odifi

catio

n of

tub

e-fe

edin

g in

cas

e of

incr

ease

d or

al in

take

. D

isco

ntin

uatio

n tu

be-

feed

ing:

ora

l int

ake

of p

rote

ins

≥ 7

5% o

f p

resc

ribed

am

ount

; sip

-fee

ding

: con

tinue

d un

til in

take

is s

uffici

ent

afte

r bei

ng p

resc

ribed

in c

ase

of in

adeq

uate

inta

ke o

r los

s of

fe

edin

g tu

be

(rem

oval

/ob

stru

ctio

n).

Eval

uat

ion

o

f in

take

by

pre

-pri

nte

d

nu

trit

ion

al

dia

ry 43

-46

RD

: 3 m

o p

osto

per

ativ

e: in

stru

ctio

n ho

w

to fi

ll ou

t ora

l int

ake

in n

utrit

iona

l dia

ry. 6

an

d 12

mo

pos

top

erat

ive:

dia

ry w

as s

ent t

o p

atie

nt b

y m

ail t

wo

wee

ks in

adv

ance

. D

iarie

s w

ere

fille

d ou

t at h

ome

durin

g 3

d (2

wee

kday

s an

d 1

wee

kend

day

) one

wee

k b

efor

e p

lann

ed fo

llow

-up

ass

essm

ent.

C

olle

ctio

n an

d ev

alua

tion

of d

iary

dur

ing

visi

ts to

out

pat

ient

clin

ic.

Att

emp

ts to

redu

ce m

is-r

epor

ting:

fille

d ou

t di

ary

was

com

bin

ed w

ith 2

4-h

reca

ll (fo

ods

cons

umed

ove

r the

pre

viou

s 24

hou

rs);

food

freq

uenc

y qu

estio

nnai

re (e

valu

atin

g fr

eque

ncy

of fo

ods

and

amou

nts

cons

umed

ov

er p

revi

ous

mon

th);

diar

y ev

alua

tion;

sh

owin

g p

ictu

res

of s

ervi

ngs

to p

atie

nt to

he

lp th

em e

stim

ate

the

amou

nts

eate

n;

mea

surin

g th

e co

nten

t of g

lass

es, c

ups,

b

owls

and

mug

s by

pat

ient

at h

ome.

Ob

ject

ive

clin

ical

as-

sess

men

t 32-3

6

RD: B

ody

wei

ght i

n ki

logr

ams

usin

g Se

ca 8

88, a

nd h

eigh

t in

cent

i-m

eter

s (o

nly

first

vis

it) u

sing

Sec

a st

adio

met

er 2

22, w

ithou

t sho

es

and

in li

ght i

ndoo

r clo

thin

g (S

eca

GM

BH).

Mea

surin

g b

ody

com

pos

ition

(bod

y ce

ll m

ass)

by

usin

g b

ioel

ectr

ical

im

ped

ance

ana

lysi

s.

Defi

nit

ion

of

mal

nu

trit

ion

32

,34

Mal

nutr

ition

was

op

erat

iona

lized

in a

ccor

danc

e w

ith th

e N

atio

nal

Dut

ch G

uide

line

on P

erio

per

ativ

e N

utrit

ion

2007

: Inv

olun

tary

w

eigh

t los

s of

≥5%

with

in 1

mon

th; a

nd/o

r Inv

olun

tary

wei

ght l

oss

of ≥

10%

with

in 6

mon

ths;

and

/or A

BM

I <18

.5.

RD: 6

mo

pos

top

erat

ive:

pre

oper

ativ

e b

ody

wei

ght w

as u

sed

in o

rder

to c

alcu

late

% o

f w

eigh

t los

s w

ithin

6 m

onth

s. T

ime

poi

nt

6 an

d 12

mo

pos

top

erat

ive:

pat

ient

s es

ti-m

ated

thei

r wei

ght 1

mo

bef

ore

stud

y as

-se

ssm

ent

in o

rder

to c

alcu

late

% o

f wei

ght

loss

with

in 1

mo.

a BM

I = B

ody

Mas

s In

dex,

cal

cula

ted

as k

g/m

2 , b RD

= re

gist

ered

die

titia

n, c

CT

= X

-ray

Com

put

ed T

omog

rap

hy, d

MRI

= M

agne

tic R

eson

ance

Imag

ing

113

Cha

pte

r 7Su

bop

tim

al in

take

of n

utri

ents

aft

er e

sop

hag

ecto

my

wit

h g

astr

ic tu

be

reco

nstr

ucti

on

nutritional diary. At six and twelve months postoperative, patients recorded their intake in these diaries during two weekdays and one weekend day during the week before follow-up assessment. To optimize correct nutritional reports, the diaries were evaluated during the visits to the outpatient clinic, and pictures of food servings were shown to ensure a more exact estimate of the amounts eaten. In addition, patients were required to undergo a 24-h recall, an interview about their food and beverage consumption during the preceding 24 hours, and had to fill out a food-frequency questionnaire (Figure 1).

IntakeThe nutritional diaries were imported into a digital program ‘Voeding’ (version 2, release 70, iSoft). The software then calculated the intake of the following nutritional compo-nents: energy, proteins, vitamins A, D, E, B-1, B-2, B-3, B-6, B-12, C, folic acid (vitamin B-11), calcium, magnesium, phosphorus, copper, iron, selenium and zinc. An average intake for the three reported days before the follow-up assessment was calculated for each nutrient. Because dietary self-reporting is prone to error, 43-46 a margin of 10% below the recommen-dations and RDAs was used to define suboptimal intake. For example, patients with an intake <90% of calculated energy requirements and <90% of the Recommended Dietary Allowances (RDAs) of protein and micronutrients were considered as patients with a suboptimal intake. 28-31 If a patient’s intake showed more than four nutrients with inadequate intake (<90%), the patient was considered to have a serious suboptimal intake, placing him at higher risk for nutritional deficiency.

Self-reported nutrition-related complaintsIn a previous study on nutrition-related complaints after esophagectomy, patients reported early satiety, postprandial dumping syndrome, inhibited passage, reflux and the absence of hunger as the most bothersome complaints. 21 This set of five items was used to assess the number of nutrition-related complaints.

Data analysisContinuous normally distributed variables were expressed as mean (± SD), and categori-cal variables were expressed as n (%). Time-related change in nutrient intake between six and twelve months after surgery was analyzed using McNemar’s test. Multivariate logistic regression was used to analyze the association between the number of nutrition-related complaints and the presence of a serious suboptimal intake of nutrients (more than four inadequate nutrients). Because the goal was to quantify the net effect of nutrition-related complaints on nutrients, the association was corrected for statistically significant univari-ate baseline characteristics and energy intake. Effect sizes were expressed in odds ratios with their corresponding 95% confidence intervals. All analyses were performed with Statistical Package for Social Sciences software (version 16.0, 2007, SPSS Corporation, Chicago, Illinois, USA).

114

Results

In total, 96 patients participated in the study (Figure 2). Of this total, 60 patients (63%) completed the one-year follow-up, and 36 patients dropped out during this period (78% because of recurrent disease). Baseline characteristics did not differ signifi-cantly (P values >0.05) between the patients who completed the study and those who dropped out. The diaries were completed by 70 patients at six months post operation and by 59 patients at twelve months post operation. At both time points, 54 patients had completed the diaries. Table 1 shows the patients’ baseline characteristics, indicating a representative sample of the target population seen in the Academic Medical Center.

Figure 2 Study flow chart. Description of participants throughout the study evaluating the intake of nutrients after esophagectomy with gastric tube reconstruction

Figure 2 Study flow chart. Description of participants throughout the study evaluating the intake of nutrients after esophagectomy with gastric tube reconstruction

6 lost to follow-up 13 were excluded from further analyses because of relapse of the disease 1 died

1 week after the introduction oral foods / drinks :

96 included participants

Esophagectomy with gastric tube reconstruction in 140

consecutive patients

6 months after surgery : 76 participants 70 patients filled out the nutritional diary

12 months after surgery : 60 participants 59 patients filled out the nutritional diary

6 exclusions due to criteria 1 suffering of a neurological disease 1 severe chyle leakage 4 not speaking the Dutch language

34 declined to participate 4 died after signing written informed consent but before entering the study

1 was excluded from further analyses because of another life -threatening disease 15 were excluded from further analyses

115

Cha

pte

r 7Su

bop

tim

al in

take

of n

utri

ents

aft

er e

sop

hag

ecto

my

wit

h g

astr

ic tu

be

reco

nstr

ucti

onIntake of energy and protein At twelve months postoperative, 24% of the patients did not reach their energy goal and 7% had a protein intake below the recommendation (Table 2). To safeguard sufficient energy and protein intake, 48% of the population continued tube-feeding after hospital discharge (mean daily intake = 1715 Kcal ± 575 and 65 g pro-tein ± 20). At six months postoperative, only two patients still used nightly tube-feeding and only one after twelve months. Sip-feeding was started in 25% of the patients after discharge (mean daily intake = 230 ml, 280 Kcal ± 125 and 14 g of protein ± 8) and was still being used by 22% at twelve months postoperative. Mean body weight did not substantially change the first postoperative year. At six months postoperative the mean body weight was 74.7 kg (± 12.1) and this was 73.9 kg (± 12.7) at twelve months post operation.

Table 1 Baseline characteristics of the patient who underwent esophagectomy with gastric tube reconstruc-tion (N = 96)

Age (y), mean (SDa) 62 (10 )

Male sex, n 73

Presence of malignancy, n 93

Neo-adjuvant treatment (chemotherapy and/or radiotherapy), n 29

Physical status related characteristics

Preoperative body weight, mean (SD) 79.0 (15.8)

BMI b, mean (SD) 26.0 (3.9)

Underweight BMI < 18.5 / obesity BMI > 30, n 5 / 13

Co-morbidity, top 3, n

Cardiovascular diseases 46

Pulmonary diseases 21

Diseases of the urinary tract 19

Preoperative complaints, top 3, n

Weight loss 49

Dysphagia 69

Epigastrical/retrosternal pain 24

Surgery-related characteristics

ASA c classification ≥ 3, n 21

Transthoracic esophageal resection, n 50

Transhiatal esophageal resection, n 46

Occurrence of postoperative complications, n 60

Admission duration, mean (SD) 20 (8)

a SD = standard deviation b BMI = body mass index c ASA = American Society of Anesthesiologists

116

Intake of micronutrients On average, patients reported a suboptimal intake of three vitamins and two minerals/trace elements. At twelve months post operation, intakes of folic acid (85% of the patients), vitamin D (61%), copper (56%), calcium (49%) and vitamin B-1 (48%) were most often reported as patients who completed their diaries at both time points (n = 54) showed no significant time-related change (range P values = 0.05 to 1.00) for any of the micronutri-ents (data not shown). At twelve months post operation, only four patients (7%) had a suf-ficient intake of all calculated micronutrients, whereas 23 patients (39%) were considered to still have a serious suboptimal intake (more than four inadequate nutrients).

Table 2 Patients with an intake of macro- and micronutrients < 90% of the nutrient recommendations, 6 and 12 months after esophagectomy with gastric tube reconstruction

6 months after surgery (n = 70)a

12 months after surgery (n = 59)a

n (%)

Macronutrients

Energy (KJ) 16 (23) 14 (24)

Proteins 6 (9) 4 (7)

Vitamins

Vitamin A 21 (30) 18 (31)

Vitamin D b 41 (59) 36 (61)

Vitamin E 17 (24) 16 (27)

Vitamin B-1 b 26 (37) 28 (48)

Vitamin B-2 20 (29) 14 (24)

Vitamin B-3 15 (21) 16 (27)

Vitamin B-6 16 (23) 18 (31)

Folic acid b 58 (83) 50 (85)

Vitamin B-12 10 (14) 10 (17)

Vitamin C 16 (23) 15 (25)

Minerals and trace elements

Calcium b 32 (46) 29 (49)

Magnesium 24 (34) 20 (34)

Phosphorus 0 (0) 1 (2)

Copper b 42 (60) 33 (56)

Iron 13 (19) 7 (12)

Selenium 18 (26) 19 (32)

Zinc 25 (36) 18 (31)

a An average intake during 3 days (2 weekdays and 1 weekend day) the week prior to the follow-up assess-ment was calculated for each nutrient. An average intake <90% of RDAs, corrected for age and gender, was defined as suboptimal intakeb The nutrients most often reported to have a suboptimal intake

117

Cha

pte

r 7Su

bop

tim

al in

take

of n

utri

ents

aft

er e

sop

hag

ecto

my

wit

h g

astr

ic tu

be

reco

nstr

ucti

on

Nutrition-related complaints and serious suboptimal intake In the univariate logistic regression model, the baseline characteristic preoperative epi-gastrical pain (crude odds ratio = 3.99; 95% CI: 1.13 to 14.01) and energy intake (crude odds ratio = 5.39; 95% CI: 1.68 to 17.29) were found to be associated with a serious sub-optimal intake at twelve months postoperative. Multivariate regression, adjusted for both baseline variables, no longer showed an independent impact of the number of nutrition-related complaints on the presence of a serious suboptimal intake of nutrients (adjusted odds ratio = 1.11; 95% CI: 0.94 to 1.31; P 0.22).

Discussion

This study demonstrates that the majority of esophagectomy patients do not reach the minimum RDA for most micronutrients at six and twelve months postoperative. The micro nutrients with the most frequently reported suboptimal intake were folic acid, vita-min D, copper, calcium and vitamin B-1. No statistically significant associations between nutrition-related complaints and suboptimal intake could be demonstrated. However, it is plausible that the number of patients identified as at risk of an inade-quate nutrition after esophagectomy is an underestimation of the actual problem. There are three reasons for this. First, a conservative cut-off point was chosen, defining an intake <90% of recommendations as suboptimal. Although the focus of this analysis was on die-tary intake approximating the RDA after esophagectomy, the analysis included an estima-tion of errors made by the patient and registered dietitian in recording and assessment of dietary intake. Earlier studies described both under-reporting and over-reporting of nutrients in studies using nutritional diaries to evaluate the nutritional intake, but a summary of the literature shows a consistent under-reporting of the energy intake by approximately 10%. 44, 46 No (consistent) results with regard to misreporting of micronutrients could be demon-strated. Therefore, based on misreporting of energy intake the margin of 10% was built in to evaluate intake. Second, the oral intake at six and twelve months postoperative was evaluated when patients were assumed to have reached a physically and emotionally stable situation. It generally requires three to nine months for patients to regain a defined eating pattern after this surgical procedure. 15, 19 Finally, only patients without relapse of the malignancy were followed the complete first year after esophagectomy. In all likelihood, the patients excluded due to recurrence had an intake even more suboptimal than the full-term participants.

LimitationsBecause the main goal was to quantify the net effect of nutrition-related complaints caused by the surgical procedure, patients experiencing from co-morbidities that affect the intestinal tract were excluded. Also excluded were patients who were unable to read

118

or speak Dutch, expecting that this would reduce the accuracy of filling out the diaries and increase misreporting. In addition, 34 patients declined to participate mainly because they assumed participation would be too stressful emotionally. Due to the possible effects of these exclusions and refusals, extrapolating and generalizing the results to the total population undergoing esophagectomy should be done with caution. Another limitation of this study was the decision to use the five most clinically sig-nificant nutrition related complaints by the patients to explain the suboptimal intake of nutrients. 21 However, most bothersome complaints do not necessarily increase the risk of a nutrient deficiency. It is probable that other complaints that are experienced as less bothersome, such as an altered defecation pattern (resulting in a reduced uptake of nutri-ents), could lead to greater risk of developing a nutrient deficiency. Finally, the follow-up time of this study was limited. Twelve months might not be long enough to show an improved nutrient intake. According to the present clinical experience, the maximum improvement in physical status occurs two to three years after surgery.

Conclusions The present study shows patients undergoing esophagectomy are at risk for micronutri-ent deficiency due to suboptimal intake up to twelve months after surgery. This implies that intake of micronutrients should be adequately monitored (and corrected when nec-essary) in post-esophagectomy patients, in addition to body weight. As a consequence of this study, an adapted nutritional support protocol was drafted (Figure 1). The adapted protocol first describes that attention is needed regarding optimum intake of energy and proteins. To achieve these goals, tube-feeding and sip-feeding must be continued, for a longer period of time if necessary, until sufficient oral intake of energy and proteins is safe-guarded. In addition, extra attention should be given to the micronutrient intake. During tube-feeding and sip-feeding, the RDAs of micronutrients are assured, but after reduction or discontinuation of enteral nutrition, this needs extra attention. It is recommended to evaluate the intake of micronutrients every three months at least until twelve months postoperative and if necessary, to supplement specific nutrients.

References 1. Parshad R, Singh RK, Kumar A, et al. Adenocarcinoma of distal esophagus and gastroesophageal junction:

long-term results of surgical treatment in a North Indian Center. World J Surg. 1999;23(3):277- 283.2. Bartels H, Stein HJ, Siewert JR. Preoperative risk analysis and postoperative mortality of oesophagectomy for

resectable oesophageal cancer. Br J Surg. 1998;85(6):840-844.3. Lerut T, Coosemans W, Decker G, et al. Extended surgery for cancer of the esophagus and gastroesophageal

junction. J Surg Res. 2004;117(1):58-63.4. Lagarde SM, de Boer JD, ten Kate FJ, et al. Postoperative complications after esophagectomy for adeno-

carcinoma of the esophagus are related to timing of death due to recurrence. Ann Surg. 2008;247(1):71-76.5. Hulscher JB, Tijssen JG, Obertop H, et al. Transthoracic versus transhiatal resection for carcinoma of the

esophagus: a meta-analysis. Ann Thorac Surg. 2001;72(1):306-313.

119

Cha

pte

r 7Su

bop

tim

al in

take

of n

utri

ents

aft

er e

sop

hag

ecto

my

wit

h g

astr

ic tu

be

reco

nstr

ucti

on

6. O’Rourke I, Tait N, Bull, et al. Oesophageal cancer: outcome of modern surgical management. Aust N Z J Surg. 1995;65(1):11-16.

7. Hulscher JB, van Sandick JW, de Boer AG, et al. Extended transthoracic resection compared with limited transhiatal resection for adenocarcinoma of the esophagus. N Engl J Med. 2002 Nov 21;347(21):1662-1669.

8. Finley FJ, Lamy A, Clifton J, et al. Gastrointestinal function following esophagectomy for malignancy. Am J Surg. 1995;169(5):471-475.

9. Blazeby JM, Williams MH, Brookes ST, et al. Quality of life measurement in patients with oesophageal cancer. Gut. 1995;37(4):505-508.

10. Moraca RJ, Low DE. Outcomes and health-related quality of life after esophagectomy for high-grade dysplasia and intramucosal cancer. Arch Surg. 2006 Jun;141(6):545-549.

11. Baba M, Aikou T, Natsugoe S, et al. Appraisal of ten-year survival following esophagectomy for carcinoma of the esophagus with emphasis on quality of life. World J Surg. 1997;21(3):282-285.

12. Conroy T, Marchal F, Blazeby JM. Quality of life in patients with oesophageal and gastric cancer: an overview. Oncology. 2006;70(6):391-402.

13. McLarty AJ, Deschamps C, Trastek VF, et al. Esophageal resection for cancer of the esophagus: long-term function and quality of life. Ann Thorac Surg. 1997;63(6):1568-1572.

14. De Leyn P, Coosemans W, Lerut T. Early and late functional results in patients with intrathoracic gastric replacement after oesophagectomy for carcinoma. Eur J Cardiothorac Surg. 1992;6(2):79-84.

15. Ludwig DJ, Thirlby RC, Low DE. A prospective evaluation of dietary status and symptoms after near-total esophagectomy without gastric emptying procedure. Am J Surg. 2001 May;181(5):454-458.

16. van Knippenberg FC, Out JJ, Tilanus HW, et al. Quality of Life in patients with resected oesophageal cancer. Soc Sci Med. 1992;35(2):139-145.

17. Ryan AM, Rowley SP, Healy LA, et al. Post-oesophagectomy early enteral nutrition via a needle catheter jejunostomy: 8-year experience at a specialist unit. Clin Nutr. 2006;25(3):386-393.

18. Page RD, Oo AY, Russell GN, et al. Intravenous hydration versus naso-jejunal enteral feeding after esophagectomy: a randomised study. Eur J Cardiothorac Surg. 2002;22(5):666-672.

19. Wainwright D, Donovan JL, Kavadas V, et al. Remapping the body: learning to eat again after surgery for esophageal cancer. Qual Health Res. 2007;17(6):759-771.

20. Gawad KA, Hosch SB, Bumann D, et al. How important is the route of reconstruction after esophagectomy: a prospective randomized study Am J Gastroenterol. 1999;94(6):1490-1496.

21. Haverkort EB, Binnekade JM, Busch OR, et al. Presence and persistence of nutrition-related symptoms during the first year following esophagectomy with gastric tube reconstruction in clinically disease-free patients. World J Surg. 2010;34(12):2844-2852.

22. Baldwin C, Weekes CE. Dietary advice for illness-related malnutrition in adults. Cochrane Database of Systematic Reviews 2008, Issue 1. Art. No.: CD002008. DOI: 10.1002/14651858.CD002008.pub3.

23. Stratton RJ, Green CJ, Elia M. Disease related malnutrition: an evidence-based approach to treatment, 1st ed. Cambridge, UK: CABI Publishing; 2003.

24. Martin L, Lagergren P. Long-term weight change after oesophageal cancer surgery. Br J Surg. 2009;96(11):1308-1314.

25. Watanabe R, Hanamori K, Kadoya H, et al. Nutritional intakes in community-dwelling older Japanese adults: high intakes of energy and protein based on high consumption of fish, vegetables and fruits provide sufficient micronutrients, J Nutr Sci Vitaminol. 2004;50(3):184-195.

26. McWhirter JP, Pennington CR. Incidence and recognition of malnutrition in hospital. BMJ. 1994;308(6934):945–948.

27. The Health Council. Sufficient intake of vitamins and minerals, 2009/06. The Hague, The Netherlands: The Health Council of The Netherlands; 2009.

28. The Health Council. Dietary standards, 2000/12. The Hague, The Netherlands: The Health Council of The Netherlands; 2000.

29. The Health Council. Optimal use of folic acid, 2008/02. The Hague, The Netherlands: The Health Council of The Netherlands; 2008.

30. The Health Council. Dietary standards: energy, proteins, fat and digestible carbohydrates, 2001/19. The Hague, The Netherlands: The Health Council of The Netherland; 2001.

31. Steyn RS, Grenier I, Darnton SJ, et al. Weight gain as an indicator of response to chemotherapy for oesophageal carcinoma. Clin Oncol (R Coll Radiol). 1995;7(6):382-384.

32. CBO Guideline. National Dutch Guideline on Perioperative Nutrition, 2007. Dutch Institute for Healthcare Improvement CBO, Utrecht, the Netherlands. CBO Web site. http://www.cbo.nl/product/richtlijnen/ folder 20021023121843/rl_periovoed_07.pdf. Accessed September, 2007.

120

33. Sauerwein HP, Serlie MJ. Optimal nutrition and its potential effect on survival in critically ill patients. Neth J Med. 2010;68(3):119-122.

34. Guideline malnutrition. Screening and treatment of malnutrition. Steering Committee Malnutrition, Amsterdam, The Netherlands. Web site. www.stuurgroepondervoeding.nl. Accessed October, 2009. Improved and adapted May, 2010 and June, 2011.

35. Windsor JA, Hill GL. Weight loss with physiologic impairment. A basic indicator of surgical risk. Ann Surg. 1988 Mar;207(3):290-296.

36. Windsor JA, Hill GL. Grip strength: a measure of the proportion of protein loss in surgical patients. Br J Surg. 1988;75(9):880-882.

37. Symons TB, Sheffield-Moore M, Wolfe RR, et al. A moderate serving of high-quality protein maximally stimulates skeletal muscle protein synthesis in young and elderly subjects. J Am Diet Assoc. 2009;109(9):1582-1586.

38. Weijs PJ, Kruizenga HM, van Dijk AE, et al. Validation of predictive equations for resting energy expenditure in adult outpatients and inpatients. Clin Nutr. 2008;27(1):150-157.

39. Sauerwein HP, Strack van Schijndel RJ. Perspective: How to evaluate studies on peri-operative nutrition? Considerations about the definition of optimal nutrition for patients and its key role in the comparison of the results of studies on nutritional intervention. Clin Nutr. 2007;26(1):154-158.

40. Heymsfield SB, McManus C, Stevens V, et al. Muscle mass: reliable indicator of protein-energy malnutrition severity and outcome. Am J Clin Nutr. 1982;35(5 Suppl):1192-1199.

41. Bourdel-Marchasson I, Joseph PA, Dehail P, et al. Functional and metabolic early changes in calf muscle occurring during nutritional repletion in malnourished elderly patients. Am J Clin Nutr. 2001;73(4):832-838.

42. Lopes J, Russell DM, Whitwell J, et al. Skeletal muscle function in malnutrition. Am J Clin Nutr. 1982;36(4):602-610.

43. Penn L, Boeing H, Boushey CJ, et al. Assessment of dietary intake: NuGO symposium report. Genes Nutr. 2010;5(3):205-213.

44. Larsson CL, Westerterp KR, Johansson GK. Validity of reported energy expenditure and energy and protein intakes in Swedish adolescent vegans and omnivores. Am J Clin Nutr. 2002;75(2):268-274.

45. Hagfors L, Westerterp K, Sköldstam L, et al. Validity of reported energy expenditure and reported intake of energy, protein, sodium and potassium in rheumatoid arthritis patients in a dietary intervention study. Eur J Clin Nutr. 2005;59(2):238-245.

46. Poslusna K, Ruprich J, de Vries JH, et al. Misreporting of energy and micronutrient intake estimated by food records and 24 hour recalls, control and adjustment methods in practice. Br J Nutr. 2009;101 Suppl 2:S73-85.

Chapter 8

General discussion

123

Cha

pte

r 8G

ener

al d

iscu

ssio

n

Since the 1970s, there has been growing awareness that nutrition has a positive impact on surgical patients and consequently on their surgical outcome. More generally, human nutrition is now considered to be so important that a recent publication even suggested that diets could replace or reduce drug treatment.1 Although the influence of dietary treat-ment on the course of disease may not be as large as described in this article, nutrition undeniably plays a role in the treatment of surgical patients. The importance of nutrition in relation to surgery is shown by the number of articles published: in 1960 no papers at all on this topic had been listed in PubMed, while in 1980, 370 articles could be found, and in 2010 this number had increased to more than 2150. Nowadays, dietetics is largely practice-based. The effectiveness and efficiency of dietary interventions has not been demonstrated for the majority of diseases. To further improve the quality of patient care it is therefore of great importance to improve the scientific credibility of dietetics based on evidence-based nutritional recommendations, methods and techniques.

The primary goal of this thesis was to contribute to evidence-based dietetics, specifi-cally focusing on the value of nutritional assessment in preoperative and postoperative patients undergoing major abdominal surgery.

The studies in this thesis, addressed the lack of knowledge about nutritional assess-ment and dietary interventions in this group of patients. Working on the various aspects of this thesis also brought to light a number of new uncertainties and misstatements that need further attention and consideration. In this final chapter we will summarize and dis-cuss our main findings. Furthermore, implications for clinical practice will be described and direction for future research will be discussed.

Prevalence of malnutrition among preoperative surgical patientsRecent epidemiological data on malnutrition among clinical surgical patients demon-strated a prevalence between 5% and 55%. 2-10 However, prevalence data on malnutrition among preoperative outpatients is still lacking. The results of our studies indicate that malnutrition occurs in 25% of the patients selected for major abdominal surgery (Chapter 4) and in about 6% of the general surgical population of outpatients (Chapters 2 and 3). In addition, these studies also showed that the living situation and household composi-tion (e.g. patient cannot rely on voluntary care or is living alone) is related to the risk of malnutrition (Chapter 2). For clinical practice, this means that dieticians should also take these psychosocial risk factors into account, otherwise management of malnutrition in the preoperative and postoperative phases will be suboptimal. Therefore, we argue for a more multidimensional approach with a focus on both physical and psychosocial aspects.

Definition of malnutritionWorldwide there is no consensus on the definition of malnutrition. 11 - 15 In this thesis mal-nutrition was operationalized according to the National Dutch Guideline on Perioperative nutrition (CBO richtlijn Perioperatief voedingsbeleid) by (a) involuntary weight loss of

124

≥ 5% within one month; and/or (b) involuntary weight loss of ≥ 10% within six months, and/or (c) a body mass index < 18.5.16 However, this definition includes some elements that are still under discussion; a critical loss of body protein must occur before vital physi-ological functions are lost (e.g. perceptible reduction of muscle strength and muscle function), and we doubt that a body weight loss of 5% in one month generally results in increased health risk. 17 - 19 In addition, it is our opinion that in patients undergoing major abdominal surgery, the definition of malnutrition should be based on low fat-free mass. A low fat-free mass is a potential risk factor for adjuvant chemo-radiation therapy, as well as postoperative complications such as infections, leakage of the anastomoses, abscesses and re-operation. 16, 20 - 27

The current definition of malnutrition also has led to discussion on the lower limit of the body mass index. The National Dutch Guideline on Perioperative Nutrition generally uses a lower limit of 18.5. But in specific diseases (e.g. COPD or cardiac disease) and for patients > 65 years of age, is it already common to use a lower body mass index limit of 20.0. Future research is needed to unravel to what extent a lower body mass index limit increases the health risk.

Prevalence of obesity in patients undergoing major abdominal surgeryIn any case, it is clear that we should avoid malnutrition in patients scheduled for major abdominal surgery. However, we must also pay attention to the proportion of the surgical population who presents with overweight (BMI 25.0 –29.9) or even obesity (BMI ≥ 30) as this also entails health risks in the preoperative and postoperative phase. Recent studies have described a prevalence of 10% - 34% obesity in their outpatients. 28 - 34 In Chapter 4 we demonstrated a mean BMI of 25.2 kg/m2 (± 3.8) in 123 patients undergoing major abdomi-nal surgery and although 25% of them were identified as malnourished, 11% were obese. In the Chapters 6 and 7 we studied 96 preoperative patients scheduled for esophagec-tomy. Their mean BMI was 26.0 kg/m2 (± 3.9), and 14% of this population was identified as obese. More research is needed to study the health risks, and the precise nutritional needs, of overweight and obese patients undergoing major abdominal surgery during the preoperative and postoperative phase.

Need for nutrients in the preoperative and postoperative phase of major abdomi-nal surgery In various chapters of this thesis the importance of optimal nutrition during illness in the preoperative phase and postoperative phase is discussed. Based on best practice, described in the National Dutch Guideline on Perioperative Nutrition, we recommended 1.5 grams of protein/kg body weight /24 hours. To estimate the energy requirements, we calculated the Harris and Benedict equation (1984) plus 30%. 16 However, the precise pro-tein and energy requirements during illness is uncertain; current recommendations are based on limited data from small studies. 35

We do not know whether these requirements safeguard the preservation of the fat-

125

Cha

pte

r 8G

ener

al d

iscu

ssio

n

free mass, especially in surgical oncology patients scheduled for major abdominal sur-gery. With regard to micronutrients (vitamins, minerals and trace elements), the Dutch recommendations for the healthy adult population are used, even though little research has been conducted on these nutrients during illness. Given the importance of an appro-priate advice with regard to energy, macronutrients and micronutrients, future studies should clarify whether and how long the need for nutrients in patients undergoing major abdominal surgery must be modified and whether subgroups should be determined with respect to BMI, ethnicity, sex, age, illness, surgical procedure and/or additional treatment.

Self-report of anthropometric dataIt is frequently assumed that self-report of weight and height generates inaccurate infor-mation. This will probably be the case for certain patient groups (e.g. obese patients and patients who suffer from an eating disorder), but our study in a general population of pre-operative outpatients showed that the use of self-reported anthropometric data is a relia-ble and valid method to screen for malnutrition (Chapter 2). A high level of agreement was found between self-reports and clinical assessments for height, weights, calculated BMI and classification of nutritional status. Moreover, when we compared the self-reported data with the frequently used screening tools Short Nutritional Assessment Questionnaire (SNAQ), Malnutrition Universal Screening Tool (MUST), and Mini Nutritional Assessment (MNA), the diagnostic accuracy of self-reports proved to be better. 17, 36 - 39

As self-reported data provide in an efficient way to screen for malnutrition in a general population of preoperative outpatients, we advise the use of self-reported data instead of conventional screenings tools in preoperative outpatients.

The results of this study have been used to prepare a form to screen all patients for mal-nutrition at the outpatient clinic of GIOCA (Gastro-Intestinal Oncology Center Amsterdam) at the Academic Medical Center.

Handgrip strength measured with a dynamometerCurrently handgrip strength measurement is a frequently used standard method to iden-tify patients at risk of malnutrition, but it is unclear which reference values should be used. In Chapter 3 we investigated whether handgrip strength, measured with a dynamometer, can be used as a method for screening for malnutrition in adult preoperative outpatients by applying the algorithms of Álvares-da-Silva, Klidjian, Matos, and Webb. 17, 40 – 42 However, none of the four algorithms derived from handgrip strength measurements was found to have sufficient diagnostic accuracy to introduce this method as a systematic institutional screening tool to detect malnutrition in individual adult preoperative elective outpatients.

A plausible explanation for this poor accuracy may be the reference values that are used. When using reference values in a specific target population, the characteristics of the target population in relation to the base population in which the values were estab-lished are often disregarded. 17, 40 - 44 In recent years, hand grip strength values of healthy native Dutch were collected by the University Medical Center Maastricht with the purpose

126

of drafting normal hand grip strength values for the Dutch population.45 In the future, it is therefore recommended to use these normal values when performing hand grip strength measurements in the Netherlands.

Interestingly, the normal handgrip strength values as defined by Mathiowetz, which are based on cut-off points, are often used to identify patients at risk of malnutrition or post-operative complications.46 However, these reference values were originally not de fined to assess these risks, but were intended to determine the need of hand strengthening. According to Mathiowetz, it is therefore incorrect and undesirable to apply these normal values for determining malnutrition or predicting postoperative complications (personal communication).

The measurement of handgrip strength itself is another issue that needs further atten-tion. To compare studies that use handgrip strength, a strict procedure should be followed as described by Mathiowetz.46, 47 In addition, the posture and motivation of the patient affects the outcome and should also be taken into account.

Finally, as it is unclear when and how sudden changes of handgrip strength are meas-urable during intervention or between preoperative and postoperative observations. More research is needed to determine the optimal time frame (days or weeks) needed to observe clinical relevant changes in handgrip strength.

Bioelectrical impedance analysis measurementBesides handgrip strength, bioelectrical impedance analysis (BIA) measurements are often performed in clinical practice. BIA is alleged to be a simple, easy and non-invasive method to estimate body compartments such as fat-free mass and fat mass during illness, recovery and treatment. 48 – 56 Frequently used BIA devices are single-frequency bioelectrical imped-ance analysis (SF-BIA), multi-frequency bioelectrical impedance analysis (MF-BIA), and bioimpedance spectroscopy (BIS). In Chapter 4 we evaluated the measurement concurrence between SF-BIA and BIS. Although the results showed good intraclass correlation coefficients between SF-BIA and BIS, the devices are not interchangeable. Compared to SF-BIA, BIS classified a larger pro-portion of the patients as suffering from a body composition outside the normal range in terms of low fat-free mass and high fat mass. Consequently, health care professionals should be aware that such devices may differ in their measurements of body composition, and could therefore affect clinical decision making in terms of starting physical therapy, dietary therapy or postponing a surgical procedure. 16, 18, 57

Our systematic review in Chapter 5 indicated that the validity of BIA devices among sur-gical and oncological (surgical) patients can be questioned; the differences between body compartments measured by both a reference method (regarded as the gold standard) and a BIA device turned out to be considerable. 48 – 50, 52, 53, 55, 58 The review also demonstrated that the estimations made by a MF-BIA device or BIS device are not more accurate than those made by a SF-BIA device. We advise health care professionals to continue measuring body composition with BIA

127

Cha

pte

r 8G

ener

al d

iscu

ssio

n

in surgical and oncological patients (also those treated with surgery), but under strict con-ditions. The measurements should be performed with the same device, using the same equation and under the same circumstances. Single measurements providing a set of absolute data on body composition are not useful because of the deviations relative to the reference methods. Multiple measurements under the same conditions, however, can pro-vide useful clinical information on changes in body composition for individual patients.

To overcome the present lack of clarity about the equations used in BIA measurements, more knowledge is needed about using raw bioimpedance data (including resistance, reactance, impedance and the phase angle) and/or vector analysis. 59-69

Nutrition-related symptoms after major abdominal surgeryDieticians should be aware of the physiological impact and consequences of major abdominal surgical procedures on the human body and must have sufficient knowledge and expertise to safeguard optimal nutritional care. Much research has been done on the best surgical methods and techniques, but these studies do not always address dietary aspects. Current expert-based dietary recommendations do not always lead to optimal and efficient patient care, and more nutrition-related research on the postoperative effects of major abdominal surgery is urgently needed. For this reason we performed a study on nutrition after esophagectomy with gastric tube reconstruction (esophagectomy).

In Chapter 6, we investigated patients’ experience nutrition-related complaints during the first year after esophagectomy and studied the changes in these symptoms. In addi-tion, we evaluated the necessity of nutrition-related adjustments as well as the patients’ nutritional status in terms of body weight.

The results show that patients suffer from a number of persistent, nutrition-related complaints during the entire first postoperative year. Early satiety, postprandial dump-ing, inhibited passage due to high viscosity, reflux of food and/or fluids, and the absence of hunger were the most frequently reported nutrition-related complaints. We demon-strated that the number of nutrition-related complaints was stable over time and could not be explained by a range of patient or surgery-related characteristics. One year post-operatively, the large majority of patients still needed to eat smaller meals with a relatively high frequency, had an altered stool frequency and still experienced the negative influ-ence of their changed food intake on their social life. A reduction of body weight occurred directly after the surgical procedure, and the majority of patients were unable to return to their preoperative body weight within one year after surgery. The weight reduction was not associated with the nutrition-related complaints.

In Chapter 7 we reported on our study – in the same cohort of patients – on the extent to which the intake of energy, proteins, and micronutrients 6 and 12 months after esophagec-tomy meets the recommendations as defined by the Health Council of the Netherlands. 70-73 In addition, we evaluated which nutrients are most frequently suboptimal in the diet and studied if nutrition-related complaints could explain a serious suboptimal intake of nutrients. The results showed that the majority of esophagectomy patients did not reach

128

the minimum recommended dietary allowances (RDA) for most micronutrients at 6 and 12 months postoperative and were therefore at risk for micronutrient deficiency. The micro-nutrients with the most frequently reported suboptimal intake were folic acid, vitamin D, copper, calcium and vitamin B1. The number of nutrition-related complaints was not an independent risk factor for the presence of a suboptimal intake of nutrients.

Based on this study we wrote an adapted AMC-nutritional support protocol to guide optimal intake of energy and proteins. To achieve these goals, tube-feeding and sip-feeding must be continued, for a longer period of time if necessary, until sufficient oral intake of energy and proteins is assured. In addition, we recommended evaluating not only the intake of energy and protein, but also the intake of micronutrients. This should be done at least every 3 months until 12 months postoperative, and if necessary, specific nutrients should be supplemented. The results of this study, described in Chapters 6 and 7, formed the basis of the uniform, national guideline ‘Nutritional advice after esophagec-tomy with gastric tube reconstruction’, by the ‘Chirurgisch Overleg Diëtisten Acade-mische Ziekenhuizen (CHIODAZ)’ and can be downloaded from the website of the Dutch Association of Dieticians (www.dietist.nl).

Implications for clinical practice and future researchIn this final section we summarise our main findings, describe the implications of the results and discuss direction for future research.

Implications Self-reported patient anthropometric data was shown to be a reliable method to screen for malnutrition (Chapter 2). These self-reports have high diagnostic accuracy, so we rec-ommend introducing self-reporting of anthropometric data in the general surgical outpa-tient population as an efficient method to screen for malnutrition.

In contrast, handgrip strength measured with a dynamometer demonstrated low accu-racy when used to screen for malnutrition in the general surgical outpatient population, irrespective the equation used (Chapter 3). We therefore advise against using this method for screening purposes; handgrip strength measurement should only be used in longitu-dinal studies to evaluate changes in strength during disease or treatment. As handgrip strength is greatly influenced by the behavioural, mental and physical state of the patient, the measurement should be carried out according to a strict procedure as described by Mathiowetz.46 For a proper evaluation of the results, the characteristics of the base popula-tion and the target population must be comparable, and it is advisable to use national reference values for Dutch studies.

Based on the results of Chapter 4, we now understand that various BIA devices do not generate the same results. Dieticians and other caregivers should be aware of this variation, which may influence their clinical decision making. The results of our system-atic review (Chapter 5) also indicate that BIA measurements in surgical and oncological patients are less valid than expected according to the statements of the manufacturers. At

129

Cha

pte

r 8G

ener

al d

iscu

ssio

n

the AMC, we routinely performed BIA estimations in patients scheduled for major abdomi-nal surgery preceding and following therapeutic interventions. Based on the results of the systematic review, we no longer consider a single BIA measurement to be useful. We now use BIA measurements only longitudinally, i.e. to evaluate changes in body composition using the same device and same equation under the same circumstances.

To optimize nutritional care during the first year after esophagectomy, our adapted AMC nutritional support protocol should be followed by dieticians (Chapters 6 and 7). This includes a more intensive monitoring of nutritional goals in terms of proteins, energy, and micronutrients with special attention for folic acid, vitamin D, copper, calcium and vita-min B1. In addition, the national guideline for patients after esophagectomy (‘Nutritional advice after esophagectomy with gastric tube reconstruction’, endorsed by CHIODAZ should be used to inform the patients.

Future researchThe current economic crisis has compelled all health care disciplines to work with smaller budgets and fewer resources. In this time of scarce funding, nutrition-related research is often regarded as a costly item that can be easily cut. However, it should be realized that saving money on the development of evidence-based dietary advices stands in the way of optimal patient care and in the long term will result in additional costs due to unnecessary, improper or excessive interventions and methods.

In the future, dieticians and medical doctors must continue to pay attention to screen-ing and treatment of malnutrition, but must also be aware that a substantial number of patients undergoing major abdominal surgery are classified as obese. More research is needed to evaluate the influence of increased BMI on the side-effects of adjuvant chemo-radiation therapy and on the occurrence of postoperative complications.

To provide optimal nutritional and dietary advice before and after surgery in patients undergoing major abdominal surgery, dieticians must rely on evidence concerning the need for energy, macronutrients (proteins, fats and carbohydrates), micronutrients (vita-mins, minerals, trace elements) and fluids. More research is needed to evaluate the opti-mal need for nutrients during adjuvant treatments, especially in patients with an abnor-mal body composition in terms of malnutrition and obesity.

Nutritional advice should focus on patients’ fat-free mass rather than total body weight; the preservation of fat-free mass during treatment has a large influence on recovery, pre-vention of side effects, and postoperative complications in patients undergoing major abdominal surgery. Future research should demonstrate whether the fat-free mass values measured in large groups of healthy adults by the research group of Kyle can be used as normal reference values. 74, 75

Regarding handgrip strength measurements by dynamometer, a study should be per-formed to evaluate the added value of these measurements when used longitudinally in patients undergoing major abdominal surgery. If it appears that the continuation of hand-grip strength measurements in this group of patient has added value, efforts should be

130

made to collect more reference values of healthy subjects in the Netherlands. In addition, a study should be conducted to determine the time frame in which changes with regard to handgrip strength measurements by dynamometer are observable during disease and treatment in patients undergoing major abdominal surgery.

Considering the results of our systematic review, it is desirable to obtain more knowl-edge about the application of the raw bioimpedance data (e.g. resistance, reactance, impedance and the phase angle at one or more frequencies) and/or vector analysis. Direct use of these values obtained by BIS measurements will probably overcome the lack of clar-ity with regard to the various equations used in BIA estimations.

In our two studies evaluating the nutritional complaints and nutrient intake after esophagectomy, we followed patients during the first postoperative year. However, it is also necessary to obtain more knowledge about the long-term complaints and intake, for at least 5 years after esophagectomy.

Finally, to further improve patient care after major abdominal surgery, an inventory of complaints and nutrients intake is also necessary in patients who have undergone a stom-ach resection, pylorus preserving pancreaticoduodenectomy or partial liver resection.

References1. Thiele I, Swainston N, Fleming RM, Hoppe A, Sahoo S, Aurich MK et.al. A community-driven global reconstruc-

tion of human metabolism. Nat Biotechnol. 2013 May;31(5):419-25. 2. Baldwin C, Weekes CE. Dietary advice with or without oral nutritional supplements for disease-related

malnutrition in adults. Cochrane Database Syst Rev. 2011 Sep 7;(9):CD002008.3. Bozzetti F, Gavazzi C, Miceli R, Rossi N, Mariani L, Cozzaglio L, Bonfanti G, Piacenza S. Perioperative total

parenteral nutrition in malnourished, gastrointestinal cancer patients: a randomized, clinical trial. JPEN J Parenter Enteral Nutr. 2000 Jan-Feb;24(1):7-14.

4. Halfens RJ, Meijers JM, Du Moulin MF, van Nie NC, Neyens JC, Schols JM. Rapportage resultaten Landelijke Prevalentiemeting Zorgproblemen 2010. Retreived from http://www.nld.lpz-um.eu/media/text/home/LPZ_rapport_2010_.pdf

5. The Malnutrition Prevalence Group. Edington J, Boorman J, Durrant ER, Perkins A, Giffin CV, James R, Thomson JM, Oldroyd JC, Smith JC, Torrance AD, Blackshaw V, Green S, Hill CJ, Berry C, McKenzie C, Vicca N, Ward JE, Coles SJ. Prevalence of malnutrition on admission to four hospitals in England. Clin Nutr. 2000 Jun;19(3):191-5.

6. Corish CA, Kennedy NP. Undernutrition in hospitals. Br J Nutr. 2001 Apr;85(4):509-10.7. Kelly IE, Tessier S, Cahill A, Morris SE, Crumley A, McLaughlin D, McKee RF, Lean ME. Still hungry in hospital:

identifying malnutrition in acute hospital admissions. QJM. 2000 Feb;93(2):93-8.8. Kyle UG, Pirlich M, Schuetz T, Luebke HJ, Lochs H, Pichard C. Prevalence of malnutrition in 1760 patients at

hospital admission: a controlled population study of body composition. Clin Nutr. 2003 Oct;22(5):473-81.9. Allison S P. Malnutrition, disease and outcome. Nutrition.2000 Jul-Aug;16(7-8):590-3.10. McWhirter JP, Pennington CR. Incidence and recognition of malnutrition in hospital. BMJ. 1994 Apr

9;308(6934):945-8.11. Meijers JM, van Bokhorst-de van der Schueren MA, Schols JM, Soeters PB, Halfens RJ. Defining malnutrition:

mission or mission impossible? Nutrition. 2010 Apr;26(4):432-40. 12. Lochs H, Allison SP, Meier R, Pirlich M, Kondrup J, Schneider S, van den Berghe G, Pichard C. Introductory

to the ESPEN Guidelines on Enteral Nutrition: Terminology, definitions and general topics. Clin Nutr. 2006 Apr;25(2):180-6.

13. Stratton RJEM, Green CJ. Scientific criteria for defining malnutrition. In: Disease- related malnutrition, 1st edn. Wallingford, UK: CABI Publishing, 2003: 1-22.

14. Soeters PB, Reijven PL, van Bokhorst-de van der Schueren MA, Schols JM, Halfens RJ, Meijers JM, van Gemert

131

Cha

pte

r 8G

ener

al d

iscu

ssio

n

WG. A rational approach to nutritional assessment. Clin Nutr. 2008 Oct;27(5):706-16. 15. Norman K, Pichard C, Lochs H, Pirlich M. Prognostic impact of disease-related malnutrition. Clin Nutr. 2008

Feb;27(1):5-15. 16. CBO Guideline Perioperative Nutrition. Dutch Institute for Healthcare Improvement. Utrecht, The Netherlands,

2007 (Accessed 2007, at www.cbo.nl/Downloads/114/rl_periovoed_07.pdf). 17. Klidjian AM, Foster KJ, Kammerling RM, Cooper A, Karran SJ. Relation of anthropometric and dynamometric

variables to serious postoperative complications. Br Med J. 1980 Oct 4;281(6245):899-901.18. Windsor JA, Hill GL. Weight loss with physiologic impairment. A basic indicator of surgical risk. Ann Surg. 1988

Mar;207(3):290-6.19. Windsor JA, Hill GL. Grip strength: a measure of the proportion of protein loss in surgical patients. Br J Surg.

1988 Sep;75(9):880-2.20. Wolfe RR. Is the double-blind randomized trial the most valid experimental approach to evaluating treatment

modalities in critical ill patiënts? Curr Opin Clin Nutr Metab Care. 1998 March; 20 (1): 185-186.21. Clowes GH Jr, George BC, Villee CA Jr, Saravis CA. Muscle proteolysis induced by a circulating peptide in

patients with sepsis or trauma. N Engl J Med. 1983 Mar 10;308(10):545-5222. Pichard C, Kyle UG, Morabia A, Perrier A, Vermeulen B, Unger P. Nutritional assessment: lean body mass

depletion at hospital admission is associated with an increased length of stay. Am J Clin Nutr. 2004 Apr;79(4):613-8.

23. Prado CM, Lieffers JR, McCargar LJ, Reiman T, Sawyer MB, Martin L, Baracos VE. Prevalence and clinical implications of sarcopenic obesity in patients with solid tumours of the respiratory and gastrointestinal tracts: a population-based study. Lancet Oncol. 2008 Jul;9(7):629-35.

24. van Venrooij LM, de Vos R, Zijlstra E, Borgmeijer-Hoelen MM, van Leeuwen PA, de Mol BA. The impact of low preoperative fat-free body mass on infections and length of stay after cardiac surgery: a prospective cohort study. J Thorac Cardiovasc Surg. 2011 Nov;142(5):1263-9.

25. van Venrooij LM, Verberne HJ, de Vos R, Borgmeijer-Hoelen MM, van Leeuwen PA, de Mol BA. Postoperative loss of skeletal muscle mass, complications and quality of life in patients undergoing cardiac surgery. Nutrition. 2012 Jan;28(1):40-5.

26. Tan BH, Birdsell LA, Martin L, Baracos VE, Fearon KC. Sarcopenia in an overweight or obese patient is an adverse prognostic factor in pancreatic cancer. Clin Cancer Res. 2009 Nov 15;15(22):6973-9

27. Visser M, van Venrooij LM, Wanders DC, de Vos R, Wisselink W, van Leeuwen PA, de Mol BA The bioelectrical impedance phase angle as an indicator of undernutrition and adverse clinical outcome in cardiac surgical patients. Clin Nutr. 2012 Dec;31(6):981-6.

28. Melis M, Weber J, Shridhar R, Hoffe S, Almhanna K, Karl RC, Meredith KL. Body mass index and perioperative complications after oesophagectomy for adenocarcinoma: a systematic database review. BMJ Open. 2013 May 2;3(5).

29. Watanabe M, Ishimoto T, Baba Y, Nagai Y, Yoshida N, Yamanaka T, Baba H. Prognostic Impact of Body Mass Index in Patients with Squamous Cell Carcinoma of the Esophagus. Ann Surg Oncol. 2013 Jun 25.

30. Scarpa M, Cagol M, Bettini S, Alfieri R, Carraro A, Cavallin F, Trevellin E, Saadeh LM, Ruol A, Vettor R, Ancona E, Castoro C. Overweight patients operated on for cancer of the esophagus survive longer than normal-weight patients. J Gastrointest Surg. 2013 Feb;17(2):218-27.

31. Melis M, Weber JM, McLoughlin JM, Siegel EM, Hoffe S, Shridhar R, Turaga KK, Dittrick G, Dean EM, Karl RC, Meredith KL. An elevated body mass index does not reduce survival after esophagectomy for cancer. Ann Surg Oncol. 2011 Mar;18(3):824-31.

32. Morgan MA, Lewis WG, Hopper AN, Escofet X, Harvard TJ, Brewster AE, Crosby TD, Roberts SA, Clark GW. Prognostic significance of body mass indices for patients undergoing esophagectomy for cancer. Dis Esophagus. 2007;20(1):29-35.

33. Dandona M, Linehan D, Hawkins W, Strasberg S, Gao F, Wang-Gillam A. Influence of obesity and other risk factors on survival outcomes in patients undergoing pancreaticoduodenectomy for pancreatic cancer. Pancreas. 2011 Aug;40(6):931-7.

34. Khan S, Sclabas G, Reid-Lombardo K, Sarr MG, Nagorney D, Kendrick ML, Que FG, Donohue JH, Huebner M, Lohse C, Farnell MB. Does body mass index/morbid obesity influence outcome in patients who undergo pancreatoduodenectomy for pancreatic adenocarcinoma? J Gastrointest Surg. 2010 Nov;14(11):1820-5.

35. Weijs PJ, Kruizenga HM, van Dijk AE, van der Meij BS, Langius JA, Knol DL, Strack van Schijndel RJ, van Bokhorst-de van der Schueren MA. Validation of predictive equations for resting energy expenditure in adult outpatients and inpatients. Clin Nutr. 2008 Feb;27(1):150-7.

36. Kruizenga HM, Seidell JC, de Vet HC, Wierdsma NJ, van Bokhorst-de van der Schueren MA. Development and validation of a hospital screening tool for malnutrition: the short nutritional assessment questionnaire

132

(SNAQ). Clin Nutr. 2005 Feb;24(1):75-82.37. Vellas B, Villars H, Abellan G, Soto ME, Rolland Y, Guigoz Y, Morley JE, Chumlea W, Salva A, Rubenstein LZ, Garry

P. Overview of the MNA--Its history and challenges. J Nutr Health Aging. 2006 Nov-Dec;10(6):456-63; 38. Guigoz Y. The Mini Nutritional Assessment (MNA) review of the literature--What does it tell us? J Nutr Health

Aging. 2006 Nov-Dec;10(6):466-85. 39. Elia M. (editor) The ‘MUST’ report. A report by the Malnutrition Advisory Group of the British Association for

Parenteral and Enteral Nutrition (BAPEN). Nutritional screening for adults: a multidisciplinary responsibility. Development and use of the‘ Malnutrition Universal Screening Tool’ (‘MUST’) for adults 2003.

40. Álvares-da-Silva MR, Reverbel da Silveira T. Comparison between handgrip strength, subjective global assessment, and prognostic nutritional index in assessing malnutrition and predicting clinical outcome in cirrhotic. Nutrition. 2005 Feb;21(2):113-7.

41. Matos LC, Tavares MM, Amaral TF. Handgrip strength as a hospital admission nutritional risk screening method. Eur J Clin Nutr. 2007 Sep;61(9):1128-35.

42. Webb AR, Newman LA, Taylor M, Keogh JB. Hand grip dynamometry as a predictor of postoperative complications reappraisal using age standardized grip strengths. JPEN J Parenter Enteral Nutr. 1989 Jan-Feb;13(1):30-3.

43. Bohannon RW, Peolsson A, Massy-Westropp N, Desrosiers J, Bear-Lehman JB. Reference values for adult grip strength measured with a Jamar dynamometer: a descriptive meta-analysis. Physiotherapy 2006 March;92: (1):11–15.

44. Jakobsen LH, Rask IK, Kondrup J. Validation of handgrip strength and endurance as a measure of physical function and quality of life in healthy subjects and patients. Nutrition. 2010 May;26(5):542-50.

45. Reijven P.L.M. Nederlandse normaalwaarden handknijpkracht Afdeling Diëtetiek Maastricht Universiteits Medisch Centrum (algoritme- onderzoek- functionele parameters). (Accessed 2011 www.nutritional-assessment.azm.nl).

46. Mathiowetz V, Kashman N, Volland G, Weber K, Dowe M, Rogers S. Grip and pinch strength: normative data for adults. Arch Phys Med Rehabil. 1985 Feb;66(2):69-74.

47. Mathiowetz V, Weber K, Volland G, Kashman N. Reliability and validity of grip and pinch strength evaluations. J Hand Surg Am. 1984 Mar;9(2):222-6.

48. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Gómez JM, Heitmann BL, Kent-Smith L, Melchior JC, Pirlich M, Scharfetter H, Schols AM, Pichard C. Bioelectrical impedance analysis--part I: review of principles and methods. Clin Nutr. 2004 Oct;23(5):1226-43.

49. De Lorenzo A, Andreoli A, Matthie J, Withers P. Predicting body cell mass with bioimpedance by using theo-retical methods: a technological review. J Appl Physiol (1985). 1997 May;82(5):1542-58.

50. Ellis KJ.Human body composition: in vivo methods. Physiol Rev. 2000 Apr;80(2):649-80.51. Selberg O, Selberg D. Norms and correlates of bioimpedance phase angle in healthy human subjects, hospita-

lized patients, and patients with liver cirrhosis. Eur J Appl Physiol. 2002 Apr;86(6):509-16. 52. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Manuel Gómez J, Lilienthal Heitmann B, Kent-Smith

L, Melchior JC, Pirlich M, Scharfetter H, M W J Schols A, Pichard C. Bioelectrical impedance analysis-part II: utilization in clinical practice. Clin Nutr. 2004 Dec;23(6):1430-53.

53. Cox-Reijven PL. The validation of bio-electrical impedance spectroscopy (BIS) for body composition in patients. Thesis. Maastricht: ISBN 90-5681-17-9, 2002.

54. Aghdassi E, Arendt B, Salit IE, Allard JP. Estimation of body fat mass using dual-energy x-ray absorptiometry, bioelectric impedance analysis, and anthropometry in HIV-positive male subjects receiving highly active antiretroviral therapy. JPEN J Parenter Enteral Nutr. 2007 Mar-Apr;31(2):135-41.

55. Earthman C, Traughber D, Dobratz J, Howell W. Bioimpedance spectroscopy for clinical assessment of fluid distribution and body cell mass. Nutr Clin Pract. 2007 Aug;22(4):389-405.

56. Lukaski HC, Johnson PE, Bolonchuk WW, Lykken GI. Assessment of fat-free mass using bioelectrical impedance measurements of the human body. Am J Clin Nutr. 1985 Apr;41(4):810-7.

57. Wolfe RR. Is the double-blind randomized trial the most valid experimental approach to evaluating treatment modalities in critical ill patiënts? Curr Opin Clin Nutr Metab Care. 1998 March; 20 (1): 185-186.

58. Buchholz AC, Bartok C, Schoeller DA. The validity of bioelectrical impedance models in clinical populations. Nutr Clin Pract. 2004 Oct;19(5):433-46.

59. Selberg O, Selberg D. Norms and correlates of bioimpedance phase angle in healthy human subjects, hospitalized patients, and patients with liver cirrhosis. Eur J Appl Physiol. 2002 Apr;86(6):509-16.

60. Gupta D, Lis CG, Dahlk SL, Vashi PG, Grutsch JF, Lammersfeld CA. Bioelectrical impedance phase angle as a prognostic indicator in advanced pancreatic cancer. Br J Nutr. 2004 Dec;92(6):957-62.

61. Barbosa-Silva MC, Barros AJ, Wang J, Heymsfield SB, Pierson RN Jr. Bioelectrical impedance analysis: popula-

133

Cha

pte

r 8G

ener

al d

iscu

ssio

n

tion reference values for phase angle by age and sex. Am J Clin Nutr. 2005 Jul;82(1):49-52.62. Piccoli A, Pillon L, Dumler F. Impedance vector distribution by sex, race, body mass index, and age in the

United States: standard reference intervals as bivariate Z scores. Nutrition. 2002 Feb;18(2):153-67.63. Barbosa-Silva MC, Barros AJ, Post CL, Waitzberg DL, Heymsfield SB. Can bioelectrical impedance analysis

identify malnutrition in preoperative nutrition assessment? Nutrition. 2003 May;19(5):422-6.64. Barbosa-Silva MC, Barros AJ. Bioelectric impedance and individual characteristics as prognostic factors for

post-operative complications. Clin Nutr. 2005 Oct;24(5):830-8.65. Barbosa-Silva MC, Barros AJ. Bioelectrical impedance analysis in clinical practice: a new perspective on its use

beyond body composition equations. Curr Opin Clin Nutr Metab Care. 2005 May;8(3):311-7.66. Cardinal TR, Wazlawik E, Bastos JL, Nakazora LM, Scheunemann L. Standardized phase angle indicates

nutritional status in hospitalized preoperative patients. Nutr Res. 2010 Sep;30(9):594-600.67. Gupta D, Lis CG, Dahlk SL, King J, Vashi PG, Grutsch JF, Lammersfeld CA. The relationship between bioelectrical

impedance phase angle and subjective global assessment in advanced colorectal cancer. Nutr J. 2008 Jun 30;7:19.

68. Itobi E, Stroud M, Elia M. Impact of oedema on recovery after major abdominal surgery and potential value of multifrequency bioimpedance measurements. Br J Surg. 2006 Mar;93(3):354-61.

69. Castillo Martínez L, Colín Ramírez E, Orea Tejeda A, Asensio Lafuente E, Bernal Rosales LP, Rebollar González V, Narváez David R, Dorantes García J. Bioelectrical impedance and strength measurements in patients with heart failure: comparison with functional class. Nutrition. 2007;23(5):412-8.

70. The Health Council. Dietary standards, 2000/12. The Hague, The Netherlands: The Health Council of The Netherlands; 2000.

71. The Health Council. Dietary standards: energy, proteins, fat and digestible carbohydrates, 2001/19. The Hague, The Netherlands: The Health Council of The Netherland; 2001.

72. The Health Council. Sufficient intake of vitamins and minerals, 2009/06. The Hague, The Netherlands: The Health Council of The Netherlands; 2009

73. The Health Council. Optimal use of folic acid, 2008/02. The Hague, The Netherlands: The Health Council of The Netherlands; 2008.

74. Kyle UG, Genton L, Slosman DO, Pichard C. Fat-free and fat mass percentiles in 5225 healthy subjects aged 15 to 98 years. Nutrition. 2001 Jul-Aug;17(7-8):534-41.

75. Kyle UG, Schutz Y, Dupertuis YM, Pichard C. Body composition interpretation. Contributions of the fat-free mass index and the body fat mass index. Nutrition. 2003; 19: 597-604.

Summary

English summary

136

Chapter 1 is the general introduction and presents the background and outline of the thesis. It provides an overview of nutrition and nutritional needs during illness in gen-eral and after major abdominal surgery in particular. The risk of malnutrition for patients undergoing surgery was described, frequently used tools to screen for malnutrition were addressed and the measurement of body composition was discussed.

Preoperative screening for malnutrition is mandatory in the Netherlands. Chapter 2 of the thesis concerns how accurately self-reported anthropometric data on body weight, height and the derivative body mass index can detect malnutrition in adult preoperative outpa-tients. In a cross-sectional study of 488 patients, self-reported data on weight and height were compared with measured data and three commonly used malnutrition screen-ing tools in the Netherlands: MNA, MUST and SNAQ. Interobserver agreement between self-reported data and clinical assessments in relation to the anthropometric indicators was high (intraclass correlation coefficient range: 0.97-0.99). The diagnostic accuracy of self-reported data to detect malnourishment was high: self-reports had a sensitivity rate of 0.97 (95% confidence interval (CI) 0.83-0.99) and specificity rate of 0.98 (95% CI 0.96-0.99). Relative to the results of the screening instruments, the diagnostic accuracy of self-reported data was found to be better in terms of sensitivity, specificity and predic-tive values. We concluded that the use of self-reported anthropometric data is a simple, highly sensitive method for screening for preoperative malnutrition in elective surgical outpatients.

Chapter 3 reports on our study of whether handgrip strength measured with a dynamom-eter can be used to screen for malnutrition in adult preoperative outpatients. A number of algorithms based on handgrip strength measurements are available to detect malnutri-tion. Frequently used algorithms are those proposed by Álvares-da-Silva, Klidjian, Matos, and Webb. In a cross-sectional study of 504 patients, the results of the four algorithms were compared with the reference standard of malnutrition (defined as ≥ 5% involuntary weight loss within 1 month and/or ≥ 10 % involuntary weight loss within 6 months and/or a body mass index <18.5). Although the Klidjian algorithm showed the highest sensitivity rate (0.79; 95% CI 0.62-0.90) relative to the other algorithms, all algorithms lacked suffi-cient diagnostic accuracy to introduce handgrip strength as systematic screening tool to detect malnutrition in individual preoperative patients.

As impaired body composition plays an important role in the occurrence of postop-erative complications in terms of infections, leakage of the anastomoses, abscesses, re-operation, increased length of hospital stay, re-admission, and mortality, it is important to assess the patients’ preoperative body composition. Bioelectrical impedance analysis (BIA) is a commonly used, easy and non-invasive method for measuring body compo-sition. Chapter 4 describes the relative reliability of two BIA devices to estimate fat-free mass and fat mass.

137

Eng

lish

sum

mar

y

In a prospective observational study among 123 consecutive adult, elective surgical patients scheduled for major abdominal surgery, the measurement agreements for the continuous variables between the single-frequency bioelectrical impedance analysis device (SF-BIA) and the bioimpedance spectroscopy device (BIS) were analysed by means of intraclass correlation coefficients. The agreement was ‘almost perfect’ for fat mass in kg (0.86, 95% CI 0.80 to 0.90); and ‘substantial’ for fat-free mass in kg (0.78, 95% CI 0.70 to 0.84). However, the mean differences between SF-BIA and BIS measurements were substantial. The agreements between both devices for the dichotomous variables, calculated using Cohen’s Kappa statistics, was ‘substantial’ for fat mass index (0.67, 95% CI 0.51 to 0.83), and ‘slight’ for fat-free mass index (0.19, 95% CI 0.01 to 0.37). The BIS device classified a larger proportion of patients as having a body composition outside the normal range. A low fat-free mass index was found in 16% of the patients when estimated with the SF-BIA device versus 47% with the BIS device; a high fat mass index was found in 71% of the patients with the SF-BIA device versus 81% with the BIS device. We therefore concluded that these two BIA devices are not interchangeable.

Doubts about the reliability of BIA estimations led to a systematic review on this topic. The aim of this systematic review, which is described in Chapter 5, was to explore the vari-ability of statistical equations used in the BIA estimations and to evaluate the validity of BIA estimations in adult surgical and oncological patients. Eleven studies (six surgery, five oncology) met the selection criteria. To illustrate variability between the equations, fixed normal reference values were entered into newly developed and existing equations, and the results were plotted in fig-ures. Substantial variability was found between the equations for both total body water (newly developed equations: up to 5 litres, existing equations: up to 20 litres or kilograms) and fat-free mass (over 25 kg). The validity of BIA estimations was studied by comparing BIA measurements with the appropriate reference methods. BIA mainly underestimated total body water (relative difference ranged from -18.8% to + 7.2%) and fat-free mass (rela-tive differences ranged from -15.2% to + 3.8%). Estimates of the fat mass compartment varied widely (relative difference ranged from -15.7 % to 43.1%). The results of this review indicate that the absence of measurement precision pre-cludes a valid estimate of a body compartment if used as a single incidental measurement. Continuation of the BIA estimations in patient populations is only useful if performed lon-gitudinally and under strict measurement conditions.

Esophagectomy with gastric tube reconstruction (esophagectomy) is mainly performed as a potentially curative treatment in patients with cancer of the esophagus or cardia. The intervention results in a variety of post-operative nutrition-related complaints that may influence the patients’ nutritional status. However, few evidence-based guidelines are available to support these patients in the postoperative phase. Chapter 6 presents a pro-spective, longitudinal cohort study of 96 patients.

138

The results showed that in the first year after the surgical intervention, the majority of the patients struggle with persistent nutrition-related complaints. The most frequently experienced complaints were the following: early satiety, postprandial dumping syn-drome, inhibited passage due to high viscosity, reflux and absence of hunger. Directly after the surgical procedure 78% of the patients lost weight, and their mean body weight remained low during the entire postoperative year. No significant association was found between the presence of complaints and body weight.

Chapter 7 is directly related to the previous chapter and focuses on the question whether it is possible to obtain sufficient intake of macronutrients and micronutrients (vitamins, min-erals and trace elements) one year after esophagectomy. Insufficient intake was defined as an intake <90% of the calculated energy requirements and <90% of the Dutch recom-mendations for protein and micronutrients. Our prospective, longitudinal cohort study of 96 patients showed that energy and pro-tein intake remained below recommendations in 24% and 7% of the patients, respectively. Fewer than 10% of the patients had a sufficient intake of all micronutrients. The lowest intake was found for folic acid, vitamin D, copper, calcium and vitamin B1. The number of nutrition-related complaints was not an independent risk factor for the presence of a suboptimal intake of nutrients (adjusted odds ratio = 1.11; 95% CI 0.94 to 1.31). The results of the study formed the basis of the national guideline national ‘esophagectomy with gas-tric tube reconstruction’, by the Chirurgisch Overleg Diëtisten Academische Ziekenhuizen (CHIODAZ).

In Chapter 8 we summarized our main research findings from the previous chapters, dis-cussed its implications for clinical practice and made suggestions for future research.

SamenvattingNederlandse samenvatting

140

Hoofdstuk 1 is de algemene inleiding en presenteert de achtergrond en opzet van het proefschrift. Het geeft een overzicht gegeven van voeding en voedingsbehoeften tijdens ziekte in het algemeen en na grote abdominale chirurgie in het bijzonder. De risico’s van ondervoeding bij preoperatieve patiënten werden beschreven en frequent gebruikte methoden voor het screenen op ondervoeding en het bepalen van de lichaamssamen-stelling werden besproken.

Preoperatieve screening op ondervoeding is in Nederland verplicht. In Hoofdstuk 2 van dit proefschrift werd bij volwassen preoperatieve poliklinische patiënten de diagnostische accuratesse onderzocht van zelf gerapporteerde antropometrische gegevens voor de detectie van ondervoeding. In een dwarsdoorsnede onderzoek bij 488 patiënten, werden de zelf gerapporteerde gegevens over gewicht en lengte vergeleken met zowel de geme-ten gegevens als met drie veelvuldig gebruikte screeningsinstrumenten voor onder-voeding (MNA, MUST en SNAQ). De interbeoordelaarsovereenkomsten tussen de zelf gerapporteerde gegevens en de klinische evaluaties met betrekking tot de antropome-trische indicatoren was hoog (intraclass correlatie coëfficiënten range: 0.97 tot 0.99). De diagnostische accuratesse van zelf gerapporteerde gegevens om ondervoeding vast te stellen was hoog: zelfrapportage had een sensitiviteit van 0.97 (95% betrouwbaarheidsin-terval [BI] 0.83 tot 0.99) en een specificiteit van 0.98 (95% 0.96 tot 0.99). Vergelijking van de screeningsinstrumenten liet een betere diagnostische accuratesse van de zelf gerappor-teerde gegevens zien in termen van sensitiviteit, specificiteit en voorspellende waarden. Wij concludeerden dat het gebruik van zelf gerapporteerde antropometrische gegevens een eenvoudige en gevoelige methode is voor het screenen op preoperatieve ondervoe-ding bij electieve, chirurgische poliklinische patiënten.

Hoofdstuk 3 presenteert een studie waar bij volwassen preoperatieve poliklinische pati-enten werd onderzocht of handknijpkracht gemeten met een dynamometer kan worden gebruikt als screeningsmethode voor ondervoeding. Er zijn verschillende op de hand-knijpkrachtmetingen gebaseerde algoritmes ontwikkeld om ondervoeding vast te stel-len. De vier meest toegepaste zijn de algoritmes van Álvares-da-Silva, Klidjian, Matos, and Webb. In een dwarsdoorsnede onderzoek bij 504 patiënten werden de resultaten van de vier algoritmes vergeleken met de referentiestandaard van ondervoeding (gedefinieerd als ≥ 5% onvrijwillig gewichtsverlies binnen 1 maand en/of ≥ 10% onvrijwillig gewichts-verlies binnen 6 maanden en / of body mass index <18.5). Hoewel het algoritme van Klidjian in vergelijking met de andere algoritmes de hoogste sensitiviteit liet zien (0.79; 95% BI 0.62 tot 0.90), ontbrak bij alle algoritmes voldoende diagnostische accuratesse om handknijpkracht te introduceren als een systematische screeningsmethode voor onder-voeding bij individuele preoperatieve patiënten.

Omdat een verstoorde lichaamssamenstelling een belangrijke rol speelt bij het ontstaan van postoperatieve complicaties zoals infectie, lekkage van de anastomose, abcessen,

141

Ned

erla

ndse

sam

enva

ttin

g

heroperatie, verlengde opnameduur, heropname en mortaliteit, is het van belang de preoperatieve lichaamssamenstelling van de patiënt te evalueren. Bio-elektrische impe-dantie analyse (BIA) is een veel gebruikte, eenvoudige en niet invasieve methode om de lichaamssamenstelling te meten. Hoofdstuk 4 beschrijft de betrouwbaarheid tussen twee verschillende BIA apparaten met betrekking tot hun schattingen van de vetvrije massa en vetmassa. In een prospectief, observationeel onderzoek bij 123 volwassen, electieve chirurgische patiënten die grote abdominale chirurgie moeten ondergaan, werden de meetovereenkomsten tussen een single-frequentie bio-elektrische impedantie analyse apparaat (SF-BIA) en een bio-impedantie spectroscopie apparaat (BIS) geanalyseerd met behulp van intraclass correlatie coëfficiënten. De overeenkomst was ‘bijna perfect’ voor de vetmassa in kg (0.86, 95 % BI 0.80 tot 0.90) en ‘substantieel’ voor de vetvrije massa in kg (0.78, 95% CI 0.70 tot 0.84 ). De gemiddelde ver-schillen tussen de SF-BIA en BIS metingen waren echter aanzienlijk. De overeenkomsten tussen beide apparaten voor de dichotome variabelen, berekend met Cohen’s Kappa, was ‘substantieel’ voor vetmassa index (0.67, 95 % BI 0.51 tot 0.83), en ‘gering’ voor de vetvrije massa index (0.19, 95 % BI 0.01 tot 0.37). Het BIS apparaat classificeerde meer patiënten met een afwijkende lichaamssamenstel-ling. Een lage vetvrije massa index werd aangetoond bij 16% van de patiënten gemeten met het SF-BIA apparaat versus 47% gemeten met het BIS apparaat; een hoge vetmassa index werd gevonden bij 71% van de patiënten met het SF-BIA apparaat versus 81% met het BIS apparaat. We concludeerden daarom dat de twee verschillende BIA apparaten niet onderling uitwisselbaar zijn.

Twijfels over de betrouwbaarheid van BIA metingen leidde tot een systematische review over dit onderwerp. Het doel van de review, beschreven in Hoofdstuk 5, was om de variabi-liteit van de statistische formules toegepast in de BIA metingen na te gaan en de validiteit van BIA schattingen bij volwassen chirurgische en oncologische patiënten te evalueren. Elf studies (zes chirurgie, vijf oncologie) voldeden aan de selectie criteria. Om de variabiliteit tussen de formules te illustreren, werden vaste normale referentie-waarden in nieuw ontwikkelde en reeds bestaande formules ingevoerd en de resultaten in figuren gevisualiseerd. Aanzienlijke uitkomstverschillen tussen de formules werden gevonden voor zowel de totale hoeveelheid lichaamswater (nieuw ontwikkelde formules: tot 5 liter, bestaande formules: tot 20 liter of kilogram) als de vetvrije massa (meer dan 25 kg). De validiteit van BIA metingen werd onderzocht door de metingen te vergelijken met metingen op basis van de juiste referentie methoden. BIA onderschat vooral de totale hoeveelheid lichaamswater (relatieve verschil varieerde van -18,8% tot + 7,2%) en de vet-vrije massa (relatieve verschillen varieerden van -15,2% tot +3,8%). Ook BIA schattingen van het vetmassa compartiment lieten een grote variatie zien (relatieve verschil varieerde van -15,7% tot 43,1%). Uit de resultaten van deze systematische review blijkt dat een incidentele BIA meting van

142

een lichaam compartiment onvoldoende betrouwbaar en valide is. Voortzetting van BIA metingen in patiënten populaties is alleen dan zinvol als het longitudinaal en onder strikte meetcondities wordt uitgevoerd. De ingreep resulteert postoperatief in diverse aan voed-ing gerelateerde klachten die de voedingstoestand van de patiënt kunnen beïnvloeden. Op dit moment zijn er nauwelijks evidence-based richtlijnen die richting geven aan de gewenste postoperatieve zorg voor deze patiënten.

Hoofdstuk 6 presenteert een prospectieve, longitudinale cohortstudie bij 96 patiënten. De resultaten toonden aan dat in het eerste jaar na de chirurgische ingreep, de meerder-heid van de patiënten worstelt met persisterende aan voeding gerelateerde klachten. De meest frequent ervaren klachten waren: snelle verzadiging, postprandiaal dumping-syndroom, passageklachten over de overgangsnaad, reflux en geen hongergevoel. Al vlak na de operatie verloor 78% van de patiënten lichaamsgewicht, en gedurende het gehele postoperatieve jaar bleef het gemiddelde gewicht van de patiënten verlaagd. Er kon geen significant verband worden aangetoond tussen de aanwezigheid van klachten en verlies van lichaamsgewicht.

Het in Hoofdstuk 7 beschreven onderzoek is een voortzetting van het vorige hoofdstuk, en richtte zich op de vraag in hoeverre de patiënten één jaar na esophagectomie voldoende inname hadden van macro- en micronutriënten (vitamines, mineralen en spoorelemen-ten). Onvoldoende inname werd gedefinieerd als een inname <90% van de berekende energiebehoefte en <90% van de Nederlanders aanbevelingen voor eiwitten en micronu-triënten. Onze prospectieve, longitudinale cohortstudie met 96 patiënten toonde aan dat de energie- en eiwitinname bij respectievelijk 24% en 7% van de patiënten onder de aan-bevelingen bleef. Bij minder dan 10% van de patiënten was sprake van voldoende inname van alle micronutriënten. De intakes van foliumzuur, vitamine D, koper, calcium en vita-mine B1 waren het laagst. Het aantal aan voeding gerelateerde klachten bleek geen onaf-hankelijke risicofactor te zijn voor een suboptimale inname van voedingsstoffen (geadjus-teerde odds ratio = 1.11, 95% BI 0.94 tot 1.31). De resultaten van deze studie vormden de basis voor de Nederlandse richtlijn ‘slokdarmresectie met buismaag reconstructie’ van het Chirurgisch Overleg Diëtisten Academische Ziekenhuizen (CHIODAZ).

In Hoofdstuk 8 hebben we onze belangrijkste onderzoeksbevindingen uit de vooraf-gaande hoofdstukken samengevat, hebben we de implicaties hiervan voor de klinische praktijk besproken en suggesties gedaan voor toekomstig onderzoek.

Addendum

Publications, Dankwoord, Curriculum Vitae

145

Pub

licat

ions

Publications

Haverkort EB, Binnekade JM, de Haan RJ, Busch OR, van Berge Henegouwen MI, Gouma DJ. Suboptimal intake of nutrients after esophagectomy with gastric tube reconstruction. J Acad Nutr Diet. 2012 Jul; 112(7):1080-7.

Haverkort EB, Binnekade JM, de Haan RJ, van Bokhorst-de van der Schueren MA. Handgrip strength by dynamometry does not identify malnutrition in individual preoperative out-patients. Clin Nutr. 2012 Oct;31(5):647-51.

Haverkort EB, de Haan RJ, Binnekade JM, van Bokhorst-de van der Schueren MA. Self-reporting of height and weight: valid and reliable identification of malnutrition in preop-erative patients. Am J Surg. 2012 Jun; 203(6):700-7.

Haverkort EB, Binnekade JM, Busch OR, van Berge Henegouwen MI, de Haan RJ, Gouma DJ. Presence and persistence of nutrition-related symptoms during the first year following esophagectomy with gastric tube reconstruction in clinically disease-free patients. World J Surg. 2010 Dec; 34(12):2844-52.

Van der Gaag NA, Verhaar AC, Haverkort EB, Busch OR, van Gulik TM, Gouma DJ. Chylous ascites after pancreaticoduodenectomy: introduction of a grading system. J Am Coll Surg. 2008 Nov; 207(5):751-7.

Assies J, Haverkort EB, Lieverse R, Vreken P. Effect of dehydroepiandrosterone (DHEA) supplementation on fatty acid and hormone levels in patients with X-linked adrenoleuko-dystrophy. Adv Exp Med Biol. 2003; 544:243-4.

Assies J, Haverkort EB, Lieverse R, Vreken P. Effect of dehydroepiandrosterone supplemen-tation on fatty acid and hormone levels in patients with X-linked adrenoleucodystrophy. Clin Endocrinol (Oxf ). 2003 Oct;59(4):459-66.

Van Geel BM, Assies J, Haverkort EB, Koelman JH, Verbeeten B Jr, Wanders RJ, Barth PG. Progression of abnormalities in adrenomyeloneuropathy and neurologically asymp-tomatic X-linked adrenoleukodystrophy despite treatment with “Lorenzo’s oil”. J Neurol Neurosurg Psychiatry. 1999 Sep;67(3):290-9.

Bruno MJ, Haverkort EB, Tijssen GP, Tytgat GN, van Leeuwen DJ. Placebo controlled trial of enteric coated pancreatin microsphere treatment in patients with unresectable cancer of the pancreatic head region. Gut. 1998 Jan;42(1):92-6.

146

Bruno MJ, Haverkort EB, Tytgat GN, van Leeuwen DJ. Maldigestion associated with exo-crine pancreatic insufficiency: implications of gastrointestinal physiology and proper-ties of enzyme preparations for a cause-related and patient-tailored treatment. Am J Gastroenterol. 1995 Sep;90(9):1383-93.

Van Geel BM, Assies J, Haverkort EB, Barth PG, Wanders RJ, Schutgens RB, Keyser A, Zwetsloot CP. Delay in diagnosis of X-linked adrenoleukodystrophy. Clin Neurol Neurosurg. 1993 Jun;95(2):115-20

Guidelines

van den Berg A, Boelens PG, Dejong CHC, Haverkort EB, Houdijk AJP, Joosten KFM, van Kempen AAMW, Kerkkamp HEM, Kranzlin ASC, van Leeuwen PAM, Madern GC, Oostenbroek RJ, Sauerwein HP, Scheepers HCJ, Severijnen RSVM, Smets MJW, van de Steeg M, Steenhagen E, Strack van Schijndel RJM, Taminiau JAJM, Tepaske R, van Zelm RT. CBO Richtlijn Perioperatief voedingsbeleid. Nederlandse Vereniging voor Anesthesiologie, Nederlandse Vereniging voor Heelkunde en Kwaliteitsinstituut voor de Gezondheidszorg CBO © 2007.

Aleman BMP, Aparicio Pages MN, Bergman JJGM, Boot H, ten Brink SAJM, Copper MP, van Dekken H, Doorn-op den Akker MM, Fockens P, van der Gaast A, Giesbers-van Dinther WM, Goedhart C, Groot D, Haverkort EB, van Heerden-van Putten MSC, Hennipman A, Hurenkamp GJB, Hussain SM, Immerzeel JJFM, Jansen RLH, Knijnenburg J, van der Kolk BM, van Lanschot JJB, Meerwaldt JH, Mulder AH, Mulder NH, Nio CY, Poldermans D, Remie ME, Richel DJ, Rosenbrand CJGM, Roukema JA, Siersema PD, Sloof GW, Smits ME, Stassen LPS, Steyerberg EW, Tilanus HW, van Vliet EJH, Wijrdeman HK, Winkel Y. CBO Richt-lijn Diagnostiek en behandeling oesofaguscarcinoom. Nederlands Genootschap van Maag-Darm-Leverartsen en Kwaliteitsinstituut voor de Gezondheidszorg CBO © 2005. ISBN 90-8523-092-6.

Chapters in books

Haverkort EB, Siersema PD. Voeding bij slokdarmaandoeningen. In: Informatorium voor Voeding en Diëtetiek, april 2006.

Barth PG, Martinez M, Apkarian P, Haverkort EB, Wanders RJA, Schutgens RBH. Disorders of Peroxisome Biogenesis: Classification and Treatment. In: Functions and biogenesis of peroxisomes in relation to human disease (pages 201-226). Koninklijke Nederlandse Akademie van Wetenschappen. Verhandelingen, Afd. Natuurkunde, Tweede Reeks, deel 95, 1995. ISBN 0-444-85799-0.

147

Pub

licat

ions

Haverkort EB, Binnekade JM, de Vos R. De keuze voor vroege enterale voeding bij inten-sive care patiënten. In: Intensive Care, Capita Selecta 2004 (pages 181-191). Venticare, Utrecht.

Popular

Haverkort L, Binnekade JM, Vos de R. Review: De keuze voor vroege enterale voeding bij intensive care patiënten. State of the art. Studie verpleging verzorging ZonMw pro-gramma tussen weten en doen. Elsevier gezondheidszorg en LEW 2003; 5: 252-263.

Binnekade JM, Haverkort L, Mathus-Vliegen L. Sondevoeding op de intensive care afde-ling. Kritiek 2003; 21 (3): 3-13.

148

Dankwoord

Velen hebben bijgedragen aan de totstandkoming van dit proefschrift. Allereerst een woord van dank aan alle patiënten die in dit proefschrift beschreven worden in het bijzon-der de patiënten die deelnamen aan het onderzoek naar aan voeding gerelateerde klach-ten na slokdarmresectie met buismaagreconstructie. Uw inzet en betrokkenheid waren bewonderenswaardig.

Een speciaal woord van dank aan mijn promotores Prof. dr. D.J. Gouma en Prof. dr. R.J. de Haan. Professor Gouma, het belang van optimaal voeden van de chirurgische patiënt is onder uw leiding een belangrijk onderdeel geworden van de behandeling op de afdeling Chirurgie. Veel dank voor al uw hulp en adviezen, het in mij gestelde vertrouwen en uw niet afla-tende interesse en begrip voor mijn persoonlijke situatie. Professor de Haan, u stelde mij als eerste in de gelegenheid een promotietraject te starten. Een kans waarvoor ik u oprecht wil danken. Dank ook voor uw intensieve hulp en sturing, uw niet aflatende scherpe en kritische blik gedurende het gehele promotie traject en uw vermogen om weerbarstige onderzoeksgegevens te vertalen in wetenschappelijke arti-kelen met een praktische invalshoek. Zonder uw inzet was het mij niet gelukt de promotie af te ronden.

Ook aan mijn co-promotor, Dr. J.M. Binnekade een persoonlijk woord van dank. Beste Jan, als jij mij niet had geïntroduceerd bij Professor de Haan zou dit proefschrift er waarschijn-lijk nooit zijn gekomen. Ik ben je veel dank verschuldigd voor je persoonlijke, kritische en inhoudelijke adviezen, je hulp bij de statische berekeningen en het maken van de mooie figuren. Een bijzondere samenwerking die na de promotie hopelijk zal worden voortgezet.

Mw. E.E. Oosterheert - van Wijnen. Beste Els, tot juli 2013 was jij Hoofd van de Afdeling Diëtetiek van het AMC. Jij bood mij de gelegenheid de Uva-master opleiding Evidence Based Practice te volgen en gaf mij daarna de tijd om promotie onderzoek te doen. Voor onze kleine afdeling een enorm risico. Dank voor het gestelde vertrouwen en het creëren van alle kansen en mogelijkheden.

Mw. Dr. M.A.E. de van der Schueren. Lieve Marian, onze persoonlijke en zakelijke wegen kruisen elkaar al meer dan 25 jaar en dat zal na promotie zeer zeker zo blijven. Dank voor alle persoonlijke en inhoudelijke gesprekken en je kritische adviezen tijdens het promotie traject. Jouw hulp heeft het afronden van de promotie absoluut versneld.

Professor dr. J.J.G.H.M. Bergman, Professor dr. M.A. Cuesta, Professor dr. J.H.G. Klinkenbijl, prof. dr. van Laarhoven, Professor dr. E.M.H. Mathus – Vliegen en Professor dr. H. Obertop wil ik hartelijk danken voor hun bereidheid zitting te nemen in de promotiecommissie.

149

Dan

kwoo

rd

Mw. E Steenhagen en Mw. A. Droop. Lieve Elles en Anneke, wat ben ik blij met de Nederlandse traditie om paranimfen mee te mogen nemen naar een promotie. Dank voor jullie geweldige, intensieve, lieve en gezellige hulp bij de voorbereiding van de promotie! Na de promotie blijven we gelukkig met elkaar in contact via het CHIODAZ (Chirurgisch Overleg Diëtisten Academische Ziekenhuizen).

Dr. M.I. van Berge Henegouwen. Professor Dr. O.R.C. Busch, Dr. C.P. Earthman en Dr. P.L.M. Reijven. Uw kritisch commentaar, opmerkingen en aanvullingen als medeauteur waren van onschatbare waarde. Veel dank.

Ook alle studenten die in de afgelopen jaren hebben meegewerkt aan de uitvoer van de diverse projecten en onderzoeken wil ik hartelijk danken: Marloes Bakker, Marieke Berkhout, Nienke Blom, Kristine Kuiper, Marcella Martin, Leonie Roeleveld, Sinja Roetert Steenbruggen, Marije Schrier, Lucie Venrooij, Lieve van der Woude en Danielle Wensveen. Zonder jullie hulp was het niet gelukt in zo’n korte tijd zoveel gegevens te verzamelen. Ik wens jullie allen een mooie toekomst.

Alle (oud) collega’s van de afdeling Diëtetiek van het AMC in het bijzonder Joyce Haver; jul-lie kennis en kunde heeft mij de afgelopen jaren bijgeschoold en scherp gehouden. Dank voor jullie input.

Alle (oud) collega’s van de afdeling Voeding en Diëtetiek van het VU Medisch Centrum wil ik danken voor hun warme en gastvrije onthaal tijdens het project naar ondervoeding op de polikliniek. Dit heeft zeker bijgedragen aan het kunnen volbrengen van de enorme klus! Een bijzonder woord van dank aan Dr. Floor Neelemaat voor alle praktische adviezen en hulp bij de afronding van de promotie.

Mr. C. Frink and Mr. E. Hull. Dear Charles and Ed, thank you very much for all the improve-ments and advices with regard to the grammar and structure of the manuscripts. I have experienced that writing a scientific manuscript in English is a hard job! You both helped me tremendously.

De afdelingen Hematologie van het UMCU en het AMC, in het bijzonder Dr. N.W.C.J. van de Donk, Drs. K P.M. van Galen en Dr. A P. Kater. Zonder uw oprechte zorg en hulp had ik deze bijzondere dag nooit samen met Henk kunnen beleven. We hopen u nog lang te mogen consulteren.

Mw. J. Heemskerk, lieve Jetty. Een mens mag zich gelukkig prijzen met een vriendin voor wie elke verandering en aanpassing in de Nederlandse taal een feest en een uitdaging is! Ondanks je zwakke gezondheid zag je kans de Nederlandse samenvatting te bekijken. Dank voor je adviezen.

150

Mw. L. van Nieukerken, lieve Leonie. We leerden elkaar kennen tijdens een Diëtetiek bij-scholing en dat bleek de opmaat voor een jarenlange trouwe vriendschap. Privé en zake-lijk hebben we al vele hoogte- en dieptepunten met elkaar gedeeld. Ik vind het geweldig dat je Toscane boven IJsland hebt verkozen om bij de promotie aanwezig te kunnen zijn!

Lieve Rob, Eline en Fleur wat er in jullie leven ook mag voorvallen, onthoud dat je veel kunt overwinnen met vitamine L; Liefde.

Lieve Jan, mijn broer. Jouw ziekte en overlijden bepaalden mijn keuze om Diëtetiek te gaan studeren en in de gezondheidszorg te gaan werken. Nu ben ik ruim twee maal zou oud als jij bent geworden, maar je bent nog altijd van invloed op de keuzes die ik in mijn leven maak.

Lieve Gerard en Diny, mijn liefvolle ouders. Ik mag me gelukkig prijzen dat ik mocht opgroeien in zo’n liefdevol, warm en stabiel gezin. Jullie fysieke aanwezigheid zal ik op mijn promotie intens missen, maar in mijn hart zijn jullie er beiden bij.

Lieve Henk, veilige haven – thuis. Ik heb bewondering voor je doorzettingsvermogen en je positiviteit zonder de werkelijkheid te negeren. Ik hoop met jou het leven nog lang te mogen delen in de lijn van Cervantes; Wie het geluk niet kan genieten als het komt, moet niet klagen als het hem verlaat. Ons leven is goed en mooi.

151

Cur

ricu

lum

Vit

ae

Curriculum Vitae

Vanaf haar afronding van de opleiding Voeding en Diëtetiek aan de Hogeschool van Amsterdam in 1986 is Elizabeth Haverkort werkzaam als diëtist in het Academisch Medisch Centrum (AMC). Zij was jarenlang betrokken bij de (experimentele) dieetbehandelingen van peroxisomale en mitochondriale ziekten zoals X gebonden Adrenoleukodystrophie, ziekte van Refsum, Zellweger syndroom en de ziekte van Barth. Tijdens het multicenter onderzoek naar de effectiviteit van dieetbehandeling van patiënten met X gebonden Adrenoleukodystrophie was zij van 1990 tot en met 1995 aangesteld als landelijk coördi-nator Diëtetiek voor de betrokken academische ziekenhuizen. Vanaf 1989 is zij werkzaam voor de afdeling Chirurgie en begeleid zij patiënten die majeure operaties in het maag-darm kanaal ondergaan. Vanuit de werkzaamheden op de afdeling Chirurgie is zij lid van het CHIODAZ (Chirurgisch Overleg Diëtisten Academische Ziekenhuizen) en was zij betrokken bij de totstandkoming van de landelijke CBO richt-lijnen Diagnostiek en behandeling oesofaguscarcinoom en Perioperatief voedingsbeleid. In 2005 behoorde zij tot het eerste cohort dat de Universitaire Master-opleiding Evidence Based Practice (EBP) bij de Universiteit van Amsterdam afrondde. Dit was de opmaat voor het zelfstandig opzetten en uitvoeren van diverse klinische onderzoeken. Op dit moment werkt zij als diëtist-onderzoeker in het AMC, treedt zij op als gastdocent bij de master opleiding Nutrition and Health aan de Faculteit Gezondheidswetenschappen van de Vrije Universiteit te Amsterdam en is als thesis begeleider verbonden aan de EBP opleiding. Daarnaast biedt zij diëtetische ondersteuning aan (oncologische) patiënten die via de poli’s oncologie, radiotherapie en GIOCA (Gastro-Intestinal Oncology Center Amsterdam) naar haar worden doorverwezen.