1

Life-long weight gain and metabolic diseasesRetrospective primary care study on the weight gain differences of elderly patients with diabetes and hypertension

Imre RURIK University of Debrecen, Medical and Health Science Center

Department of Family and Occupational Medicine

Hungary

The Future of Primary Care in Europe IV

Göteborg, September 3- 4 2012

2

BACKGROUND• Metabolic syndrome (MetSy) is one of the “hot topics”

in medical research • Diabetes and hypertension often related to significant

weight gain • Becoming overweight or obese takes many years,

perhaps decades• Genetic factors, lifestyle and social elements could

also be important, but mainly nutritional habits alone or combined with sedentary lifestyle are the real reasons for growing fat

• A medical setting is required to measure MetSy parameters: (plasma glucose level, lipids, BP etc.)

• Body weight and height are usually checked by the individuals themselves

3

AIM and OBJECTIVE

• The aim of this retrospective study is to analyse and to compare the life-long data on weight and BMI of patients with diabetes and hypertension and those without these pathologic conditions.

• Hypothesis: the possible relation between weight gain in the youth and different incidence of hypertension and diabetes in the later age periods.

4

METHODStudy design: RetrospectiveSetting: Primary care practices between May and August, 2011. Selection of participants:Elderly people (between 60 and 70y)

consecutively selected.

Data sources: questionnaire.

Ethics:Regional Ethical Committee Form of informed consent.

Exclusion criteria:•Serious impairment of visual, acoustic or mental capacity•Refusal to participate, •Unreliability: Subjects were excluded if the difference between the self-reported and the measured data exceeded 3 kg (108ps)

Quantitative variables:divided into three groups with diabetes (DM), or with hypertension (Hyp) neither diabetes nor hypertension (None), as the control group. Patients with both illnesses were counted among diabetics.

5

METHOD The QuestionnaireDear Patient !Please contribute to our international scientific evaluation and provide some information on you filling the boxes of the questionnaire.

Please give data about your body weight in [kg] and height in [cm] at your different age periods, based on your memory, private records ore most preferably medical records, hospital reports, discharge letter etc.

Your date will be handled confidentially, known only by our staff, but because of health regulation we need your signature on a different sheet, as an informed consent.

If you are not sure one of the data, you can leave this box empty. Your Family Physician

Year of born: 19_ _ your gender: Male/ Female Recent body weight according to your measurement

kg Recent body height according to your measurement

cm

Body weight at: 20 years

Body height at

30 years 40 years 50 years 60 years 70 years

your highest body weight

kg

it was measured at the age of _ _ _ years

By FEMALEs Only number of deliveries: last delivery at the age of: year last regular menstruation : at the age of: year Recent body weight measured by the practice staff

kg

cm

Doctors records: : Practice ID: diagnosed diabetes at the age of: year, none hypertonia at the age of: year, none

6

RESULTS Descriptive• Refusal: 5 (reasons were not questioned) • Population: 759 persons (337 men, 422 women) • Differences between self-reported and

professionally measured body weight +2.2 Kg men +1.8 Kg women • 14 men and 21 women both hypertensive and

diabetic • mean age of diagnosing: hypertension men: 55.4 (±15.9) (median: 59) y women: 53.8 (±15.7) (median: 56) y diabetes men: 51.3 (±17.1) (median:53) y women: 54.4 (± 13.1) (median: 50) y

7

RESULTS 1. Distribution of patients regarding morbidities (diabetes-DM, Hypertension-Hyp, control group-without these conditions-None) and BMI groups

Morbidity DM Hyp None men women men women men women n:759

81 162 94 337 BMI groups [kg/m2] 97 234 91 422 Underweight (< 18.5)

1 3 2 4 10

Normal (18.5 - 24.9)

20 25 61 80 30 26 242

Overweight (25-29.9)

29 35 67 121 51 47 350

Obese (30 <)

32 36 34 30 11 14 157

8

RESULTS 2. MENmeans SD of body weight and BMI

decades (Dec)

recently (Act)

maximal (Max)

in the last decade prior to the diagnoses

(p=) values within decades

changes between means of decades (Δ).

MEN Body weight [kg] BMI [kg/m2]

n=337 DM n=81

Hyp n=162 None n=94

DM Hyp None

Dec mean mean mean mean mean mean

[year] ±SD Δ ±SD Δ ±SD Δ p ±SD Δ ±SD Δ ±SD Δ p

20 65.810.4

63.29.8

64.212.1

0.047 22.84.8

21.43.1

20.91.8

0.048

30 69.210.7

3.4 64.010.1

0.8 64.912.1

0.7 0.012 25.24.4

2.4 22.93.2

1.5 22.31.9

1.4 0.032

40 71.8

12.92.6 66.6

9.8

2.6 67.412.0

2.5 0.025 26.94.6

1.7 24.43.1

1.5 23.22.2

0.9 0.054

50 76.1

13.74.3 69.3

9.8

2.7 70.412.8

3.0 0.034 28.93.8

2.0 26.12.7

1.7 26.22.3

3.0 0.033

60 78.8

13.72.7 72.2

10.9

2.9 72.915.3

2.5 0.011 29.33.9

0.4 27.53.9

1.4 27.92.9

1.7 0.088

Max 89.9

13.624.1 86.2

16.7

23.0 83.010.5

18.8 0.042 30.94.1

8.1 29.75.3

8.3 28.43.7

7.5 0.087

Act 88.8

16.323.0 85.4

12.9

22.2 79.89.2

15.6 0.021 29.14.0

6.3 28.63.9

7.2 27.24.2

6.3 0.071

Priorto dg

10.0 6.5 0.005 2.4 1.8 0.039

9

RESULTS 3. WOMENmeans ±SD of body weight and BMI decades (Dec)

recently (Act) maximal (Max)

in the last decade prior to the diagnoses(p=) values within decades

changes between means of decades (Δ).

WOMEN Body weight [kg] BMI [kg/m2]

n=422 DM n=97 Hyp n=234 None n=91 DM Hyp None

Dec mean mean mean mean mean mean

[year] ±SD Δ ±SD Δ ±SD Δ p ±SD Δ ±SD Δ ±SD Δ p

20 57.89.9

55.310.9

55.810.3

0.342 21.43.7

20.72.8

21.3 4.9

0.068

30 59.210.1

1.4 56.511.3

1.2 60.410.6

4.6 0.694 22.63.4

1.2 22.43.9

1.7 23.0 4.9

1.7 0.089

40 64.312.1

5.1 60.412.8

3.9 62.88.5

2.4 0.915 25.23.1

2.6 23.94.4

1.5 23.93.9

0.9 0.334

50 71.714.4

8.4 66.912.7

6.5 66.98.4

4.1 0.087 28.24.6

3.0 26.45.4

2.5 25.3 3.6

1.4 0.033

60 74.213.7

2.5 69.912.0

3.0 68.38.4

1.4 0.036 29.23.2

1.0 27.64.9

1.2 24.9 3.6

-0.4 0.088

Max 87.88.9

30.0 78.215.2

22.9 75.110.8

19.3 0.023 32.73.9

11.3 29.76.0

2.1 28.4 3.6

7.1 0.041

Act 78.712.2

20.9 76.215.5

20.9 72.49.6

16.6 0.652 29.94.6

8.4 28.15.2

7.4 28.7 4.8

7.4 0.066

PrDg 12.3 9.5 0.025 2.5 1.8 0.042

10

RESULTS 4. changes by gender /co-morbidity IR and HR

Hypertension Diabetes

Decade

range (%) IR Adjusted HR Adjusted HR (95% CI)

20-30y (95% CI) p IR p

q 1-3 66- 1.3 0.4q 4 101- 1.5 1.13 (0.80-1.57) 0.492 0.7 2.05 (1.28-3.31) 0.003q 5 106-171 1.4 1.30 (0.95-1.79) 0.104 0.5 1.38 (0.82-2.32) 0.219Male* 1.2 1.00 0.5 1.00Female 1.4 1.29 (1.04-1.60) 0.018 0.4 0.69 (0.49-0.98) 0.038Co-morb* No 1.2 1.00 0.2 1.00 Yes 1.5 1.49 (1.18-1.88) 0.001 0.5 1.72 (1.07-2.75) 0.023 20-40yq 1 81- 1.3 0.3q 2 101- 1.3 0.91 (0.66-1.24) 0.548 0.4 1.03 (0.58-1.80) 0.931q 3 106- 1.1 0.72 (0.52-0.99) 0.087 0.4 1.12 (0.64-1.98) 0.690q 4 110- 1.3 0.93 (0.96-1.51) 0.677 0.4 1.18 (0.65-2.14) 0.583q 5 117-192 1.7 1.65 (1.20-2.26) 0.002 0.7 2.32 (1.35-3.97) 0.002Female 1.4 1.20 (0.96-1.51) 0.107 0.4 0.66 (0.45-0.95) 0.025Co-morb Yes 1.5 1.32 (10.05-1.68) 0.019 0.2 0.45 (0.28-0.72) 0.118

20-50yq 1 77- 1.3 0.4q 2 104- 1.2 0.97 (0.70-1.34) 0.841 0.2 0.59 (0.31-1.14) 0.117q 3 111- 1.3 1.04 (0.75-1.44) 0.818 0.5 1.63 (0.94-2.83) 0.084q 4 117- 1.2 0.92 (0.66-1.28) 0.624 0.4 1.20 (0.67-2.14) 0.540q 5 130-218 1.6 1.51 (1.09-2.08) 0.012 0.7 2.38 (1.42-4.01) 0.001Female 1.4 1.31 (1.05-1.64) 0.017 0.64 (0.45-0.93) 0.017Co-morb Yes 1.5 1.38 (1.09-1.75)) 0.008 1.63 (1.00-2.65) 0.05020-60y q 1 77- 1.4 0.3q 2 106- 1.2 0.75 (0.54-1.06) 0.101 0.3 0.98 (0.49-1.94) 0.946q 3 115- 1.3 0.89 (0.64-1.24) 0.477 0.4 1.66 (0.87-3.14) 0.122q 4 123- 1.2 0.72 (0.52-1.01) 0.056 0.5 1.95 (1.07-5.35) 0.029q 5 136-209 1.5 1.19 (0.85-1.67) 0.308 0.7 2.97 (1.64-5.35) 0.000Female 1.4 1.29 (1.03-1.61) 0.029 0.66 (0.46-0.96) 0.028Co-morb Yes 1.5 1.39 (1.09-1.78) 0.008 1.62 (1.00-2.63) 0.05220-70yq 1 91- 1.5 0.3q 2 116- 1.5 0.70 (0.27-1.80) 0.455 0.3 0.94 (0.13-6.78) 0.947q 3 128- 1.6 0.37 (0.12-1.14) 0.084 0.7 1.97 (0.13-7.23) 0.429q 4 139- 1.5 0.56 (0.21-1.46) 0.234 0.3 0.98 (0.13-7.23) 0.985q 5 156-191 1.6 1.02 (0.39-2.68) 0.968 0.2 0.52 (0.46-5.90) 0.600Co-morb Yes 2.91 (1.30-6.49) 0.009 0.63 (0.17-2.34) 0.485

Range- upper and lower values of quintiles with equally distributed data (N=127-141 at age 20y),

calculated from weight at respective decade/ weight at 20y

IR- incidence rate of the respective morbidity (expressed in cases per 100 persons-years) (relevant IR for Male* is the same for all decades)

HR- hazard ratio of developing morbidity (diabetes or hypertension) adjusted for gender, age, and co-morbidity

11

RESULTS 5. Weight changes between decades and hazard ratios

Dec Hypertension Diabetes20-30y Adjusted HR p Adjusted HR p

(95% CI) (95% CI)10% 1.31 (1.08-1.58) 0.006 1.24 (0.92-1.66) 0.16230% 2.23 (1.26-3.97)   1.89 (0.78-4.59)  Male* 1.00  1.00Female 1.28 (1.03-1.59) 0.024 0.69 (0.49-0.98 0.036Co-morNo* 1.00 1.00Yes 1.51 (1.20-1.89) 0.001 1.75 (1.10-2.80) 0.01930-40y      10% 1.16 (1.03-1.30) 0.011 1.35 (1.13-1.63) 0.00130% 1.55 (1.10-2.18)   2.48 (1.43-4.30)Female 1.32 (1.06-1.64) 0.013 0.67 (0.47-0.97) 0.033Co-morb 1.40 (1.11-1.77) 0.004 1.57 (0.98-2.52) 0.06140-50y        10% 1.09 (0.98-1.22) 0.125 1.25 (1.09-1.43) 0.00230% 1.31 (0.93-1.84)   1.93 (1.28-2.92)  Female 1.34 (1.07-1.67) 0.010 0.70 (0.49-1.01) 0.056Co-morb 1.42 (1.12-1.79) 0.003 1.66 (1.02-2.68) 0.03950-60y      10% 1.02 (0.91-1.14) 0.791 1.08 (0.90-1.30) 0.39830% 1.05 (0.75-1.47)   1.27 (0.73-2.12)  Female 1.34 (1.07-1.68) 0.010 0.79 (0.55-1.14) 0.203Co-morb 1.47 (1.16-1.86) 0.002 1.82 (1.11-2.97) 0.01760-70y        10% 0.85 (0.62-1.17) 0.317 0.92 (0.48-1.77) 0.81130% 0.62 (0.24-1.59)   0.79 (0.11-5.50)  Female 1.08 (0.55-2.14) 0.811  0.48 (0.13-1.63) 0.240 Co-morb 2.03 (1.00-4.12)  0.051    

Percentage of weight change: weight at examined decade / weight at 20years.

HR- hazard ratio of developing morbidity (diabetes or hypertension) adjusted for gender, age, and co-morbidity Co-morbidity- No: diabetes or hypertension only (*relevant for all decades)Yes: both are present

Cox regression method was used to determine (HR) and 95% CI

Schoenfeld residual test was used to check the proportional hazards assumption.

12

Main findings

• Steady weight or slow increases in body weight have been seen more often among persons without diabetes.

• Diabetic women and men had higher body weights in their youth, and they gained significantly more weight in their third and fourth decades.

• Weight gain by diabetics in the decades prior to the diagnosis was significantly higher in both genders,

• Rapid weight gain before 40 years of age was registered more frequently among diabetics.

• The similarities of weight gain between the control and hypertensive groups may be due to the high prevalence of hypertension among the elderly.

13

Study limitation

• Reliability of data based on memory and personal records. (The accuracy of data and longevity of follow up need some

compromise.)• Overlap between the different groups with both illnesses • Higher standard deviations in some categories• Therapy and interventions started at a younger age • Genetic elements and other relations remain undiscovered• Diabetic patients without hypertension were preferred for

inclusion, therefore they were over-represented• Study population could not be considered representative (Although both urban and rural inhabitants were involved. Frequent

visitors to primary care surgeries were over-represented; they surely did not belong to the “healthy elderly” population.)

14

CONCLUSIONS

• Stability in body weight or limited and slow continuous weight gain was related to lower hazard ratios for some components of MetSy, likely diabetes.

• More reliable data could be retrieved only from cohort follow-up epidemiological evaluations within a larger population, in the frame of international cooperation, comparing ethnically diverse populations, as was recommended earlier in our pilot study.Implication for primary care

• Family physicians should not just be inactive observers of the weight gain of their patients, they must act earlier.

• They have to focus to the younger persons, who rarely visit medical settings, providing advice for body weight maintenance.

15

Why primary care?

16

Accepted for publication: Kovács E, Jancsó Z, Móczár Cs, Szigethy E, Frese T, Rurik I.

Life-longweight change can predict metabolic diseases. Retrospective primary care study on the weight gain differences

between elderly patients with diabetes and hypertension

Experimental and Clinical Endocrinology & Diabetes, 2012

Thanks for your attention!