Clinical Endocrinology (1998) 49, 597–602

597q 1998 Blackwell Science Ltd

Growth hormone treatment improves body fluiddistribution in patients undergoing electiveabdominal surgery *

J. Møller*, M. B. Jensen, E. Frandsen†, N. Møller*,P. Kissmeyer and S. LaurbergDepartment of Surgery L, Aarhus University Hospital,Aarhus C, *Medical Department M, Aarhus UniversityHospital, Aarhus C, and †Department of ClinicalPhysiology, Københavns Amtssygehus Glostrup,Denmark

(Received 8 January 1998; returned for revision 27February1998; finally revised 24 March 1998; accepted 28 April1998)

Summary

OBJECTIVE To investigate the possible beneficialeffects of growth hormone (GH) in catabolic patientswe examined the impact of GH on body fluid distribu-tion in patients with ulcerative colitis undergoingelective abdominal surgery.DESIGN AND MEASUREMENTS Twenty-four patients(14 female, 10 male) aged 19–47 years were in adouble-blinded study randomly assigned to receiveeither placebo ( n ¼ 12) or GH (n ¼ 12) 6IU s.c. twicedaily from 2 days before until 7 days after ileo-anal Jpouch surgery. Extracellular and plasma volume(ECV, PV) were determined using 82Br and 125I albu-min dilution at day ¹2 and at day 7, and body compo-sition was estimated by dual X-ray absorptiometryand bioimpedance. Changes in body weight andfluid balance were recorded and hence intracellularvolume was assessed.RESULTS During placebo treatment body weightdecreased 4·3 6 0·6 kg; during GH treatment bodyweight was constant ( P <0·01). There was a positivefluid balance in the GH-treated patients compared tothe placebo group (GH: 3·6 6 0·7 l; plc: -0·7 6 1·2 l,P <0·01). ECV increased 2·12 6 0·70 l during GH andwas unaffected during placebo ( P ¼ 0·02). PV wasunchanged by GH and decreased 0·39 6 0·08 lduring placebo administration ( P ¼ 0·03). Intracellu-lar volume (ICV) decreased less during GH than

during placebo (GH: ¹1·426 0·45; plc: ¹3·706 0·76;P ¼ 0·02). Bioimpedance remained constant duringGH administration and increased 60 6 9 ohm in theplacebo-treated group ( P <0·05). Plasma renin andaldosterone remained unchanged in both studygroups.CONCLUSION Body weight, plasma volume and intra-cellular volume is preserved during GH treatment incatabolic patients and ECV is increased. From a ther-apeutic point of view these effects may be desirableunder conditions of surgical stress.

The anabolic effects of human GH in severely ill patients havebeen known for many years (Liljedahlet al., 1961). The adventof biosynthetic GH stimulated a number of studies confirmingand expanding this knowledge (Wardet al., 1987; Pontinget al., 1988; Ziegleret al., 1988; Jianget al., 1989; Hammarq-vist et al., 1992; Voermanet al., 1995; Vara Thorbecket al.,1996). These studies have focused mainly on the protein-sparing effect of GH in terms of nitrogen metabolism andretention. The positive impact of GH in catabolic patientscould, however, also involve sodium and water-retaining effectsof GH in humans (Mølleret al., 1991). The importance ofmaintaining normal hydration in catabolic patients has beenemphasized recently (Finnet al., 1996). The intracellular stateof hydration seems to be especially important in order to pro-mote recovery, but some studies suggest additionally that volumeexpansion, i.e. enlargement of plasma volume and extracellularvolume, may improve cardiovascular stability (Garrisonet al.,1996; Sinclairet al., 1997) and stabilize gut perfusion (Mythen& Webb, 1995) during surgery. In addition animal studiessuggest that GH may improve intestinal healing after surgery(Christensen & Oxlund, 1994). The impact of GH on fluiddistribution in catabolic patients has been examined by Gatzenet al.(1992) and GH was, in contrast to studies in normal subjects,surprisingly reported to decrease extracellular volume. Withthis background, we examined the impact of growth hormoneon fluid homeostasis in a homogeneous group of patients withinflammatory bowel disease undergoing elective abdominalsurgery. Our aim was to assess whether GH treatment had thepotential to improve the state of hydration in this group ofpatients and, if so, to dissect the role of changes in extracellularvolume, plasma volume and, by inference, intracellular volume.

* See Commentary on page 573.Correspondence: Jens Møller, Medical Department M, AarhusKommunehospital, DK 8000 Aarhus C, Denmark. Fax:þ45 89492010.

Materials and methods

Patients

Twenty-four patients with ulcerative colitis gave informedconsent and took part in the study, which was approved by theDanish National Board of Health and by the local ethicalcommittee. The patients had all been treated previously witheither proctectomy or proctocolectomy. Their illness was in astable phase and they were all in good physical condition at thetime of referral for surgery. At the time of inclusion none of thepatients received medication due to their disease. The surgicalprocedure was an ileo-anal anastomosis (IAA) with establishmentof a J pouch and a concomitant high ileostomy. Each patient wasrandomly assigned to receive either growth hormone (n¼ 12)or placebo (n¼ 12). Nineteen patients (12 males, 7 females)with a median age of 28 years (range 22–49 years) completedthe study. Three patients receiving GH and two receivingplacebo dropped out of the study due to surgical complications.

Design

The study was performed in a double-blind randomized placebo-controlled design. The patients received either GH (Norditropin,Novo Nordisk, Denmark) 6 IU or placebo injections twice dailyat 0800 h and 2000 h for 10 days. The injections wereadministered subcutaneously in the thigh by a trained nurse.The treatment was started 2 days prior to elective surgery andcontinued until 7 days after the operation. The patients wereinpatients during the whole study period. The patients receivedonly enteral nutrition and enteral or parenteral fluids to keepthem clinically normally hydrated according to routine proce-dures in the surgical department. All fluids and food administeredwere carefully recorded. During each study period blood sampleswere obtained regularly. All blood samples were drawn after atleast 30 min rest in the supine position.

Water balance

Liquids were measured directly and water content in food wasobtained from tables (Danish Composition Tables, Levneds-middelstyrelsen, Søborg, Denmark). Fluid output was measuredin urine, ileostomy, drains and vomit. Insensible water loss wasestimated to be 10 ml/kg/day. Water production by combustionwas calculated as 0·033 (ml/kJ)×basal energy expenditure (kJ/24 h)×1·2 (FAO/WHO/UNU, 1985). Indirect calorimetry(Deltatrac Calorimeter, Helsinki, Finland) was performed ondays -2 and 7. The fluid balance from day -2 until day 7 wascalculated as total output subtracted from total input.

Body composition

Body composition was estimated by bioimpedance, by dexa

scan and isotopically at days -2 and 7. Extracellular volume(ECV) and plasma volume (PV) were determined using82Brand125I-albumin, respectively (Binder & Leth, 1970; Danielsenet al., 1986).82Br-Bromide was injected at 0900 h and bloodsamples were obtained 4 h later. PV was determined between1300 h and 1400 h. Lean body mass and fat mass weredetermined by dexa scan. Body weight was measured dailyduring standardized conditions. The change in intracellularwater was calculated in two different ways: by subtractingchanges in ECV, fat mass (estimated by dexa) and protein(calculated from nitrogen balance), from the change in bodyweight and as the change in ECV subtracted from the fluidbalance.

Blood pressure

Blood pressure and heart rate were measured every 10 min from0800 h to 1400 h at days -2 and 7 using a portable automaticmonitor (SpaceLabs Model 90202, Redmond, WA, USA).

Assays

Serum insulin-like growth factor I (IGF-I), serum insulin andserum GH were measured by radioimmunoassay (Laursenet al.,1996). Plasma renin concentrations were determined using anantibody-trapping method (Mølleret al., 1996). Plasma aldo-sterone was measured using a commercial kit (DiagnosticSystems Laboratories, Webster, Texas, USA). N-terminalproANP was measured radioimmunologically using a commer-cial antiserum (Penninsula Laboratories, Belmont, CA, USA).

Statistics

SAS statistical software was employed. Simple unpairedt-testswere used to analyse the difference between after and beforevalues (delta values) in the two treatment groups. Analysis ofvariance (general linear model, GLM) for repeated measure-ments was used when appropriate. Ap-value less than 0·05 was

598 J. Møller et al.

q 1998 Blackwell Science Ltd,Clinical Endocrinology, 49, 597–602

5000

4000

3000

2000

1000

0

–1000

–2000

–3000

ml

GH Placebo

P = 0.02(a)

100

0

–100

–200

–300

–400

–500 GH Placebo

P = 0.03(b)

Fig. 1 (a) Changes in extracellular volume (ECV) in the two groups.(b) Changes in plasma volume (PV) in the two groups.

considered significant. All results are expressed as mean6 SEM.

Results

Body composition

A significant increase in ECV was observed during GHadministration compared to the placebo-treated group (placebo:16·576 0·90–16·796 0·96 l; GH: 16·416 0·87–18·546 0·95 l;P<0·05) (Fig. 1). During placebo treatment plasma volumewas significantly reduced, whereas GH prevented this reduction(placebo: 3·106 0·14–2·716 0·14; GH: 2·966 0·14–2·896 0·14;P<0·05) (Fig. 1). There was a significant weightloss (kg) during placebo administration and constant weightduring GH administration (placebo: 72·06 5·7–68·46 5·0;GH: 67·16 3·7–66·46 3·8;P<0·05) (Fig. 2c). Bioimpedance(ohm) remained constant during GH administration andincreased significantly during placebo administration (placebo:4936 28–5536 42; GH: 5426 23–5546 28; P< 0·05). Simi-larly, lean body mass (kg) estimated by dexa was unchanged inthe GH-treated patients and decreased significantly in theplacebo group (placebo 51·96 4·4–48·76 3·6; GH: 47·362·2–47·66 2·3;P<0·01). There was a significant positive fluidbalance in the GH-treated patients compared to the placebogroup (plc:¹0·76 1·2 l, GH: 3·66 0·7 l) (P< 0·01) (Fig. 2a).ICV calculated on the basis of changes in body weightdecreased significantly less in the GH-treated patients than inthe placebo group (plc:¹3·76 0·8 l; GH:¹1·46 0·4 l;P<0·05)(Fig. 2d). Similar estimates of ICV were obtained from

calculations of fluid balance (plc:¹0·96 1·2; GH: 1·56 0·8;P¼ 0·13) (Fig. 2b).

Renin, aldosterone and N-terminal proANP

Plasma renin (mU/l) increased during GH exposure and duringplacebo administration (placebo: 30·26 7·4–57·86 21·0; GH:33·76 6·7–42·26 8·8; NS). An insignificant decrease inplasma aldosterone (pmol/l) was observed during both studyperiods (placebo: 5446 104–3566 167; GH: 6136 130–3376 117). Plasma N-ANP (mg/l) remained unaffectedduring both treatment regimens (placebo: 6206 118–5496 111; GH: 8366 201–6796 93; ns).

Blood pressure, heart rate and energy expenditure

Systolic and diastolic blood pressure (mmHg) increased sig-nificantly during GH administration as well as during placeboadministration (placebo: 1216 3/716 2–1256 4/766 3; GH:1156 3/656 2–1206 2/696 2; P¼ 0·04/P¼ 0·03). Heart ratewas increased significantly during GH compared to placebo(placebo: 706 3–756 3; GH: 716 4–896 4; P<0·01). Similarincrements were seen in energy expenditure (placebo:16476 88–16506 89, GH: 15426 75–16986 83, P<0·01).

IGF-I

GH treatment initially increased mean serum IGF-I (mg/l) buton the second postoperative day a fall was observed; similarly,a significant decrease in IGF-I was seen at day 2 after the

Effects of GH on body fluid distribution in catabolic patients 599

q 1998 Blackwell Science Ltd,Clinical Endocrinology, 49, 597–602

5000

4000

3000

2000

1000

0

–1000

–2000

–3000

ml

GH Placebo

P < 0.01(a)

1000

0

–1000

–2000

–3000

–4000

–5000GH Placebo

P < 0.01(c)

5000

4000

3000

2000

1000

0

–1000

–2000

–3000

ml

GH Placebo

P = 0.13(b)

1000

0

–1000

–2000

–3000

–4000

–5000 GH Placebo

P = 0.02(d)

gm

l

Fig. 2 (a) Fluid balance in the two treatmentgroups. From 2 days before until 7 days aftersurgery all fluid output and calculatedinsensible water loss was subtracted from totalinput and the calculated water production bycombustion. (b) Change in intracellularvolume calculated as fluid balance—change inextracellular volume. (c) Changes in totalbody weight in the two groups. (d) Change inintracellular weight calculated as change intotal body weight—change in ECV—changein fat mass—change in protein.

operation in the placebo-treated group (placebo: day -2,2766 21, day 0, 2726 27, day 2, 1416 17, day 7, 1756 17;GH: day -2, 2586 34, day 0, 5136 63, day 2, 2446 37, day 7,4826 78; P<0·01) (Fig. 3). Serum IGFBP-3 (data not shown)and serum insulin (GH: 75·76 1; plc: 59·86 12·3 pmol/l)yielded patterns almost identical to IGF-I.

Discussion

We have examined the possible beneficial effects of GH onbody fluid distribution in a homogeneous group of catabolicpatients using a double-blinded placebo-controlled study. Wefound that GH preserved total body weight, an overall effectwhich was due to expansion of ECV and preservation of ICVand PV.

The anabolic impact of growth hormone in terms of nitrogenretention has been examined in several trials with catabolicpatients (Liljedahlet al., 1961; Wardet al., 1987; Pontinget al.,1988; Ziegler et al., 1988; Jianget al., 1989; Hammarqvistet al., 1992; Voermanet al., 1995; Vara Thorbecket al., 1996).In these trials the catabolic states of the patients in general havebeen heterogeneous with respect to severity and duration ofdisease, and only some studies have examined the patientsduring standardized controlled conditions (Ziegleret al., 1988;Jiang et al., 1989; Hammarqvistet al., 1992; Vara Thorbecket al., 1996). The studies have concentrated on nitrogenretention and metabolic parameters, whereas indices of fluiddistribution have been studied less extensively. In this study theimpact of GH on body fluid distribution in patients in goodphysical condition exposed to comparable surgical stress wasexamined, since the importance of these parameters in catabolicstates of disease is undisputed (Finnet al., 1996) and

assessment of agents potentially capable of preserving orimproving these components may therefore be clinicallyrewarding.

The effects of GH on body fluid distribution in catabolicpatients are difficult to examine due to heterogenity in terms ofduration and intensity of the catabolic state, and only fewstudies have addressed the issue (Gatzenet al., 1992;Schambelanet al., 1996). In the study by Gatzenet al. (1992)19 patients with burns, multiple injuries or surgical complica-tions received GH or placebo for 2 weeks, 3–4 weeks afterhospital admission. In line with the present data ICV decreasedduring placebo. Surprisingly, Gatzenet al. reported a decreasein ECV in the patients treated with GH and an increased ECV inthe placebo-treated patients. The discrepancy between thesefindings and ours might relate to the fact that the patients in thestudy by Gatzenet al. were a heterogeneous group and werestudied at different time points during their catabolic phase. In awell-designed study by Schambelanet al. (1996) 178 catabolicHIV-positive patients received either GH or placebo. Asignificant increase in extracellular, as well as intracellular,water was observed together with an increase in work capacity.We made similar observations regarding fluid distributionsuggesting that a significant proportion of the anabolic impactof GH could be explained in terms of restoration or enlargementof extracellular and intracellular fluid compartments. Theimportance of intracellular volume in the catabolic state iswell established (Haussingeret al., 1993), whereas ECV hasbeen considered mainly responsible for maintaining vascularvolume. We have estimated ICV in two different ways andobtained almost identical results. ICV calculated on the basis ofbody weight was significantly better preserved in the GH-treated patients compared to the placebo group. This method isbased on body weight, fat mass (estimated by dexa scan) andprotein loss (estimated by nitrogen balance), three parameterswhich are easily determined. Large standard deviations wereseen when ICV was calculated from changes in fluid balance,probably due to lack of precise estimates of insensible waterloss. We did not calculate total body water on the basis of dexascan since this procedure includes an algorithm which isdependent on a constant water fraction of lean body mass, acondition which clearly is not present during surgical stress andGH administration. The present data show that GH has thepotential to influence intracellular volume and hence it could bespeculated that GH exerts part of its anabolic action throughpreservation of intracellular hydration, as hypothesized byHaussingeret al. (1993).

Conflicting results regarding perioperative fluid homeostasisare available. In two studies (Reid, 1969; Kragelund, 1970)ECV determined with82Br increased a few hours post-operatively compared to preoperative values. In another studyECV decreased 1 day after major gastric surgery (Shizgalet al.,

600 J. Møller et al.

q 1998 Blackwell Science Ltd,Clinical Endocrinology, 49, 597–602

400

300

200

100

IGF

-1(µ

g/l

)

Time (days)

–2 0 2 7

Fig. 3 Insulin-like growth factor I (IGF-I) in the two groups duringthe study period.B, GH; X, placebo.

1977). The data are, however, blurred by differences in fluidadministration. It is beyond doubt that profound changes inbody fluid composition during major gastrointestinal surgeryoccur due to evaporation, blood and fluid loss and fluidredistribution (Gold, 1992). The increased ECV and stable PVduring GH administration observed in this study underlines theimportance of GH as a regulator of fluid homeostasis. In concertwith the finding of improved outcome of hip surgery aftervascular volume optimization (Sinclairet al., 1997) ourfindings of maintenance of plasma volume in GH-treatedcatabolic patients suggest that GH could be of clinical value tostabilize fluid distribution in catabolic patients.

A potential drawback of the present study is that fluiddistribution was calculated 7 days after surgery. This designwas chosen because it has been shown previously that mostpatients at this time have reached a steady state for fluiddistribution (Reid, 1969). In this way, acute scatter due topatients being more or less severely affected by the operationwas minimized.

The patients received enteral fluids postoperatively accord-ing to standardized routine procedures at the surgical depart-ment. The treatment was entirely managed by ‘blinded’anaesthetists and surgeons without any knowledge of whetherpatients received GH or placebo. This procedure was chosen inorder to make the study realistic and to avoid dehydration oroverhydration as a consequence of the protocol. Equal amountsof fluids were administrated in the two groups, indicating thatGH acts by decreasing sodium and water excretion, and not bycentral stimulation of thirst.

The mechanism behind the observed fluid retention remainsunclear. Previous studies have suggested the renin–angioten-sin–aldosterone system to be a main effector of GH-inducedsodium retention (Ho & Weissberger, 1990; Mølleret al.,1997). However, renin was increased equally by GH andplacebo and plasma aldosterone decreased concomitantly inboth groups. Similarly, atrial natriuretic factor decreased in thetwo groups. It is possible that any impact of GH on the renin–angiotensin–aldosterone system was concealed by the ‘noise’introduced by the dramatic changes in body fluid homeostasisduring the perioperative period, or GH could exert itsantinatriuretic action through a direct cellular stimulation(Herlitz et al., 1994). IGF-I, which is a major effector of therenal effects of GH (Hirschberg & Kopple, 1993) couldpossibly mediate the antinatriuretic actions of GH. However, itwas found recently that IGF-I did not change ECV in elderlypatients treated with IGF-I (Thompsonet al., 1995). Anotherexplanation is that the GH-induced hyperinsulinaemia causesfluid retention since insulin is known to increase sodiumreabsorption at the level of the distal tubule (DeFronzo, 1981).

There was a similar increase in systolic and diastolic bloodpressure in both study groups, suggesting that the relatively

high GH dose chosen and the subsequent expansion of fluidcompartments has little effect on cardiovascular homeostasis.There was a significant increase in heart rate in the GH-treatedpatients, possibly relating to increased energy expenditure (DeBoeret al., 1995).

In general GH dosages chosen by other investigators havebeen between 20 and 30 IU daily. Studies have shown,however, that GH exerts its action when administered in muchsmaller doses (Mølleret al., 1990). The impact of GH on fluiddistribution at small dose schedules is not well examinedwhereas 12 IU daily is known to cause fluid retention (Molleret al., 1991). Therefore 12 IU daily was chosen. The dose wasadministered twice daily to obtain a constantly elevated level ofGH in the serum, so that possible direct effects of GH not wereconcealed. Five patients (3 GH, 2 plc) dropped out of the studydue to surgical complications. No side effects that could bedirectly related to GH were observed.

In conclusion we have shown that GH treatment in surgicalpatients increases extracellular volume and preserves intracel-lular volume and plasma volume. These findings may haveimportant clinical implications under conditions where main-tenance of normal body water is desirable.

Acknowledgements

The skillful technical assistance of Joan Hansen is acknowl-edged. Novo Nordisk A/S Gentofte, Denmark generouslysupplied us with growth hormone and performed the blindingprocedure.

References

Binder, C. & Leth, A. (1970) The distribution volume of82Br as ameasurement of the extracellular fluid volume in normal persons.Scandinavian Journal of Clinical Laboratory and Investigation, 25,291–297.

Christensen, H. & Oxlund, H. (1994) Growth hormone increases thecollagen deposition rate and breaking strength of left colonicanastomoses in rats.Surgery, 116,550–556.

Danielsen, H., Pedersen, E.B., Nielsen, A.H., Herlevsen, P., Kornerup,H.J. & Posborg, V. (1986) Expansion of extracellular volume in earlypolycystic kidney disease.Acta Medica Scandinavica, 219,399–405.

De Boer, H., Blok, G.J. & van der Veen, E.A. (1995) Clinical aspects ofgrowth hormone deficiency in adults.Endocrine Review, 16,63–86.

DeFronzo, R.A. (1981) The effect of insulin on renal sodiummetabolism.Diabetologia, 21, 165–171.

FAO/WHO/UNU Expert consultation report (1985)WHO TechnicalReport Series, 724, pp. 1–208.

Finn, P.J., Plank, L.D., Clark, M.A., Connolly, A.B. & Hill, G.L. (1996)Progressive cellular dehydration and proteolysis in critically illpatients.Lancet, 347,654–656.

Garrison, R.N., Wilson, M.A., Matheson, P.J. & Spain, D.A. (1996)Preoperative saline loading improves outcome after elective,noncardiac surgical procedures.American Surgery, 62, 223–231.

Effects of GH on body fluid distribution in catabolic patients 601

q 1998 Blackwell Science Ltd,Clinical Endocrinology, 49, 597–602

Gatzen, C., Scheltinga, M.R., Kimbrough, T.D., Jacobs, D.O. &Wilmore, D.W. (1992) Growth hormone attenuates the abnormaldistribution of body water in critically ill patients.Surgery, 112,181–187.

Gold, M.S. (1992) Perioperative fluid management.Critical CareClinics, 8, 409–421.

Hammarqvist, F., Stromberg, C., von der Decken, A., Vinnars, E. &Wernerman, J. (1992) Biosynthetic human growth hormonepreserves both muscle protein synthesis and the decrease inmuscle-free glutamine, and improves whole-body nitrogen economyafter operation.Annals of Surgery, 216,184–191.

Haussinger, D., Roth, E., Lang, F. & Gerok, W. (1993) Cellularhydration state: an important determinant of protein catabolism inhealth and disease.Lancet, 341,1330–1332.

Herlitz, H., Jonsson, O. & Bengtsson, B.A. (1994) Effect ofrecombinant human growth hormone on cellular sodium metabolism.Clinical Science (Colch), 86, 233–237.

Hirschberg, R. & Kopple, J.D. (1993) The effects of growth hormoneand insulin-like growth factor I on renal glomerular and tubularfunction. In Growth Hormone and Insulin-like Factor I(eds A.Flyvbjerg, H. Ørskov & K.G.M.M. Alberti), pp. 229–254. JohnWiley and Sons, London.

Ho, K.Y. & Weissberger, A.J. (1990) The antinatriuretic action ofbiosynthetic human growth hormone in man involves activation ofthe renin–angiotensin system.Metabolism, 39, 133–137.

Jiang, Z.M., He, G.Z., Zhang, S.Y., Wang, X.R., Yang, N.F., Zhu, Y. &Wilmore, D.W. (1989) Low-dose growth hormone and hypocaloricnutrition attenuate the protein–catabolic response after majoroperation.Annals of Surgery, 210,513–524.

Kragelund, E. (1970) Changes of the apparent HOH, Br, I humanalbumin and Cr red blood cell dilution volumes before, during andafter operation in human subjects.Annals of Surgery, 172,116–124.

Laursen, T., Møller, J., Jørgensen, J.O.L., Ørskov, H. & Christiansen,J.S. (1996) Bioavailability and bioactivity of intravenous versussubcutaneous infusion of growth hormone (GH) in GH deficientpatients.Clinical Endocrinology (Oxf), 45, 333–339.

Liljedahl, S.O., Gemzell, C.A., Plantin, L.O. & Birke, G. (1961) Effectsof human growth hormone in patients with severe burns.ActaChirugica Scandinavica, 122,1–14.

Møller, J., Frandsen, E., Fisker, S., Jørgensen, J.O.L. & Christiansen,J.S. (1996) Decreased plasma and extracellular volume in GHdeficient adults and the acute and prolonged effects of GHadministration. A controlled experimetal study.Clinical Endocrinol-ogy (Oxf), 44, 533–539.

Møller, J., Jørgensen, J.O., Møller, N., Hansen, K.W., Pedersen, E.B. &Christiansen, J.S. (1991) Expansion of extracellular volume andsuppression of atrial natriuretic peptide after growth hormoneadministration in normal man.Journal of Clinical and Endocrino-logical Metabolism, 72, 768–772.

Møller, N., Jørgensen, J.O., Schmitz, O., Møller, J., Christiansen, J.,Alberti, K.G. & Orskov, H. (1990) Effects of a growth hormone pulseon total and forearm substrate fluxes in humans.American Journal ofPhysiology, 258,E86–E91.

Møller, J., Møller, N., Frandsen, E., Wolthers, T., Jørgensen, J.O. &Christiansen, J.S. (1997) Blockade of the renin–angiotensin–aldosterone system prevents growth hormone-induced fluid retentionin humans.American Journal of Physiology, 272,E803–E808.

Mythen, M.G. & Webb, A.R. (1995) Perioperative plasma volumeexpansion reduces the incidence of gut mucosal hypoperfusionduring cardiac surgery.Archives of Surgery, 130,423–429.

Ponting, G.A., Halliday, D., Teale, J.D. & Sim, A.J. (1988) Post-operative positive nitrogen balance with intravenous hyponutritionand growth hormone.Lancet, 1, 438–440.

Reid, D.J. (1969) The body fluid compartments during surgery.Annalsof the Royal College of Surgeons of England, 44, 88–100.

Schambelan, M., Mulligan, K., Grunfeld, C., Daar, E.S., LaMarca, A.,Kotler, D.P., Wang, J., Bozzette, S.A. & Breitmeyer, J.B. (1996)Recombinant human growth hormone in patients with HIV-associated wasting. A randomized, placebo-controlled trial. SerostimStudy Group [see comments].Annals of Internal Medicine, 125,873–882.

Shizgal, H.M., Solomon, S. & Gutelius, J.R. (1977) Body waterdistribution after operation.Surgery, Gynecology and Obstetrics,144,35–41.

Sinclair, S., James, S. & Singer, M. (1997) Intraoperative intravascularvolume optimisation and length of hospital stay after repair ofproximal femoral fracture: randomised controlled trial.BritishMedical Journal, 315,909–912.

Thompson, J.L., Butterfield, G.E., Marcus, R., Hintz, R., Van Loan, M.,Ghiron, L. & Hoffman, A.R. (1995) The effects of recombinanthuman insulin-like growth factor-I and growth hormone on bodycomposition in elderly women.Journal of Clinical EndocrinologicalMetabolism, 80, 1845–1852.

Vara Thorbeck, R., Ruiz Requena, E. & Guerrero Fernandez, J.A.(1996) Effects of human growth hormone on the catabolic state aftersurgical trauma.Hormone Research, 45, 55–60.

Voerman, B.J., Strack van Schijndel, R.J., Groeneveld, A.B., De Boer,H., Nauta, J.P. & Thijs, L.G. (1995) Effects of human growthhormone in critically ill nonseptic patients: results from aprospective, randomized, placebo-controlled trial.Critical CareMedicine, 23, 665–673.

Ward, H.C., Halliday, D. & Sim, A.J. (1987) Protein and energymetabolism with biosynthetic human growth hormone after gastro-intestinal surgery.Annals of Surgery, 206,56–61.

Ziegler, T.R., Young, L.S., Manson, J.M. & Wilmore, D.W. (1988)Metabolic effects of recombinant human growth hormone inpatients receiving parenteral nutrition.Annals of Surgery, 208,6–16.

602 J. Møller et al.

q 1998 Blackwell Science Ltd,Clinical Endocrinology, 49, 597–602