Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
International Journal of Epidemiology O International Epldemtok>gical Association 1996 Vol. 25, No. 2 Printed In Great Britain Characteristics of Non-InsulinDependent Diabetes Mellitus in Elderly Men: Effect Modification by Family History JOLANDA M A BOER,*** EDITH J M FESKENS* AND DAAN KROMHOUT+ Boer J M A (Department of Chronic Disease and Environmental Epidemiology, National Institute of Public Health and Environmental Protection, PO Box 1, NL-3720 BA Bllthoven, The Netherlands), Feskens E J M and Kromhout D. Characteristics of non-lnsulin-dependent diabetes mellitus in elderly men: Effect modification by family history. International Journal of Epidemiology 1996; 25: 394-402. Background. Heredity, obesity and fat distribution may interact with each other in their association with diabetes risk. Therefore we tned to elucidate the role of familial diabetes as effect modifier In the association of obesity, glucose metabolism and lipoprotelns with non-insulin-dependent diabetes mellitus. Methods. A cross-sectional study was carried out among 468 elderly men. Within strata of family history, men with diabetes and normal glucose tolerance were compared with respect to anthropometry, characteristics of glucose metabolism and serum lipids. Results. Of the participants, 14.5% were diabetic. In diabetic men a family history of diabetes occurred more often (22.1%) than in men with normal (6.8%, P < 0.001) or impaired glucose tolerance (8.5%). In diabetic men with a family history, the ratio of fasting insulin to glucose and the ratio of areas under the insulin and glucose curves during oral glucose tolerance testing were lower compared to men with normal glucose tolerance. In men without a family history, these differences were smaller (Interaction P = 0.06). In diabetic men without a family history, fasting Insulin levels were markedly elevated (P < 0.001), whereas in men with a family history there was only a slight elevation. The presence of a family history resulted in more severe deteriorations in lipids, especially in fasting trigrycerides (interaction P = 0.075). No interaction between indices of obesity and a family history was observed. Conclusions. Our findings suggest that elderty diabetic men with a family history of diabetes represent a different subgroup than elderly men without such a history, characterized by larger deteriorations in indices for beta-cell function and higher triglyceride levels. Keywords: diabetes mellitus, non-insulin-dependent, family history, glucose tolerance, Insulin, lipoproteins, obesity It is well established that non-insulin-dependent (Type 2) diabetes mellitus (NIDDM) is a heterogeneous syndrome with both environmental and genetic factors contributing to its aetiology.' Whether insulin resistance or a defect in insulin secretion from the pancreatic beta-cell is the primary defect in the development of NIDDM may also depend on genetic background, but conflicting results have been reported.2 In genetically susceptible individuals a defect in insulin-secretion,3 insulin resistance4-5, or both 6 ' 7 were mentioned as the earliest lesion in the development of NIDDM. Several studies suggest that, besides heredity, obesity and fat distribution are strongly related to diabetes risk. 8 "" Recently, some studies reported that these factors interact with each other, and suggest that the effect of obesity is less in subjects with a family history of diabetes. In subjects with a family history of diabetes there seems to be no difference in obesity level and fat distribution between subjects with normal glucose tolerance and impaired glucose tolerance.12-13 Osei et al. found indices of obesity could not predict insulin responses in relatives of diabetic subjects.14 • Department of Chronic Disease and Environmental Epidemiology, National Institute of Public Health and Environmental Protection, PO Box 1, NL-3720 BA Billhoven, The Netherlands. • • Department of Epidemiology and Public Health, Wageningen Agricultural University, Wageningen, The Netherlands. f Division of Public Health Research, National Institute of Public Health and Environmental Protection, Bilthoven, The Netherlands. Lipid metabolism is probably involved in the aetiology of NIDDM.2 Plasma insulin levels and insulin resistance are correlated with plasma lipoprotein levels, e.g. elevated triglycerides and reduced high-densitylipoprotein (HDL) cholesterol levels. 13 It is not clear to what extent these associations depend on genetic susceptibility and are modified by a positive family history of diabetes. 1416 394 FAMILIAL AND NON-FAMIUAL DIABETES MELLITUS Therefore, the aim of the present study was to determine whether a family history of diabetes modifies the association of obesity, glucose metabolism and plasma lipoproteins, with NIDDM. This issue was studied in a population-based study of elderly men with a relatively high prevalence of diabetes mellitus. SUBJECTS AND METHODS Subjects The Zutphen Elderly Study is a longitudinal study on risk factors for chronic diseases.17 It represents a continuation of the Zutphen Study, the Dutch contribution to the Seven Countries Study.18 In 1985, 1266 men born between 1900 and 1920 and living in Zutphen for at least 5 years were invited for the examinations, of which 939 men agreed to participate. This group formed the cohort of the Zutphen Elderly Study. In 1990, 718 of these men were still alive and invited for the followup survey, of which 560 attended. For the analyses presented in this paper known diabetic men using insulin (n = 8) were excluded because of suspicion of insulindependent (Type 1) diabetes mellitus (IDDM) or Late Autoimmune Diabetes Mellitus in Adults (LADA).19 For 468 men (aged 69-90 years) data on family history and glucose tolerance were available and used in analyses. Study Protocol An anthropometric examination took place with subjects dressed in underwear only. Height was measured to the nearest 0.5 cm with a microtoise. Body weight was recorded to the nearest 0.5 kg. Body mass index (BMI) was defined as weight/height2 (kgAn2). Triceps and subscapular skinfold thickness were measured twice to the nearest mm, at the right side of the body, using a Harpenden calliper. Triceps skinfold was measured midway between the acromion and the olecranon, whereas subscapular skinfold was taken just below the tip of the scapula. The ratio of subscapular to triceps skinfold thickness (skinfold ratio) was used as an index of body fat distribution. A 75 g oral glucose tolerance test (OGTT) was carried out in the morning, after an overnight fast. Known diabetic subjects treated with oral hypoglycaemic agents (n = 23) were excluded from the test for ethical reasons. Before the test, a fasting blood sample was taken and thereafter two more samples 1 and 2 hours after the glucose load. Samples were collected in tubes with sodium fluoride. Plasma glucose was determined using the hexokinase method.20 Serum insulin was measured using a radioimmuneassay (Pharmacia Diagnostics, Uppsala, Sweden), and 395 within- and between-run coefficients of variation ranged from 6 to 7%. Fasting serum C-peptide levels were determined using a I25 I RIA Kit (Incstar Corp., Minnesota, USA). The within-run coefficient of variation was 6.5% and the between-run coefficient of variation was 14%. Glucose and insulin areas under the OGTT-curve (AUC) were calculated using the trapezoidal method. Subjects were classified according to WHO criteria as subjects with normal glucose tolerance (fasting glucose <7.8 mmol/1 and 2-h glucose <8.7 mmol/1), impaired glucose tolerance (IGT, fasting glucose <7.8 mmol/1 and 2-h glucose 8.7-11.1 mmol/1), or diabetes mellitus (fasting glucose 3*7.8 mmol/1 and/or 2-h glucose 2=11.1 mmol/1).21 The latter group also included diabetic subjects who were excluded from the OGTT. Triglycerides were determined enzymatically using a test kit (Boehringer, Mannheim)22 in serum collected in the fasting state. The HDL-lipoproteins were isolated according to the method of Warnick et al.23 Non-fasting total serum cholesterol and HDL-cholesterol were determined enzymatically using a CHOD-PAP monotestkit (Boehringer, Mannheim) as described elsewhere.24'25 Low-density-lipoprotein (LDL) cholesterol levels were calculated using the Friedewald equation.26 In the 1985 survey participants were asked whether or not a first degree relative (i.e. father, mother, sibling or child) was diabetic. If this was the case, a family history of diabetes was considered to be present. Statistical Analyses Analyses were carried out on ranked values because other transformations did not improve normality of the data. Differences between the three groups (normal glucose tolerance, impaired glucose tolerance and diabetes mellitus) were tested by analyses of variance. Subjects with normal glucose tolerance were compared to diabetic subjects using the Wilcoxon rank-sum test for independent samples, separately for subjects with and subjects without a family history of diabetes. Subjects with impaired glucose tolerance (n = 47) were excluded from these analyses because of small numbers. Adjusted means of the ranked values were determined by analysis of covariance, and differences between men with normal glucose tolerance and diabetic men were calculated. To transform the differences back to original units, adjusted means were substituted in regression lines relating normal to the ranked values. Interactions of family history of diabetes with other determinants for diabetes in the association with diabetes prevalence were calculated using logistic regression. All analyses were carried out with the Statistical Analysis System (SAS version 6.07, SAS Institute, Cary, NC, USA). 396 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY TABLE 1 Selected characteristics of elderly men with normal glucose tolerance, impaired glucose tolerance and diabetes mellilus: The Zutphen Elderly Study Normal glucose tolerance n = 353 Age (years) Body mass index (kg/m2) Triceps skinfold (mm) Subscapular skinfold (mm) Skinfold ratio Family history of diabetes (%) 75.8 25.4 12.1 16.8 1.44 6.8 Impaired glucose tolerance n-47 ± 4.4 ± 3.0 ±4.4 ±5.8 ±0.37 77.8 26.6 13 8 19.7 1.49 8.5 ±5.1* ± 4.0* ±5.3* ±6.8* ±0.38 Diabetes mellitus n = 68 75.4 ± 4.4f 26.1 ±3.0 12.2 ±4.2 19.4 ± 6.6' 1.64 ±0.49' 22.1« Data are presented as mean ± SD. * P < 0.05, * P < 0.01, • P < 0.001 versus men with normal glucose tolerance. ' P < 0.01 versus men with impaired glucose tolerance. P-values < 0.05 were considered statistically significant. Interaction terms were considered to be significant at P < 0.1, as recommended by Rothman.27 RESULTS According to WHO criteria 353 men (75.4%) had normal glucose tolerance, while 47 men (10.0%) showed impaired glucose tolerance and 68 men (14.5%) had NIDDM. Of the diabetic men, 57% were newly diagnosed by the OGTT. Men with impaired glucose tolerance were older than both men with normal glucose tolerance and diabetes (Table 1). Both diabetic men and men with impaired glucose tolerance were more obese than subjects with normal glucose tolerance, as demonstrated by differences in BMI and subscapular skinfold thickness. Men with diabetes had a more centralized body fat distribution compared to both other groups, but the difference only reached significance compared to men with normal glucose tolerance. Adjustment for age did not alter these results. A family history of diabetes occurred three times more often in diabetic men than in men with normal glucose tolerance (odds ratio [OR] 3.9, 95% confidence interval [CI] : 1.9-7.9), and additional analyses showed that this difference was independent of age, BMI, and skinfold ratio. The occurrence of a family history was comparable in known (24.1%) and newly diagnosed diabetic men (20.5%). The presence of a family history of diabetes in men with impaired glucose tolerance was comparable to those with normal glucose tolerance. The duration from time of clinical diagnosis to time of study did not differ significantly between diabetic men with (4.6 ± 7.8 years) and without a family history (3.4 ± 6.4 years). Further results concerning separate analyses for men with and men without a family history of diabetes are shown in Tables 2, 3 and 4. In men without a family history, obesity was weakly associated with diabetes prevalence. For subscapular skinfold thickness and the skinfold ratio this association was statistically significant (P < 0.05), but not for BMI or triceps skinfold (Table 2). In men with a family history of diabetes, no significant differences in BMI, triceps and subscapular measures were found between diabetic and non-diabetic subjects (P > 0.05). The difference in skinfold ratio was larger than in men without a family history (P < 0.05). Interaction terms between family history of diabetes and indices of obesity and body fat distribution were not statistically significant (P > 0.1). In both diabetic men with and without a family history of diabetes, fasting glucose and the area under the glucose curve were elevated when compared to nondiabetic men (Table 3). In contrast, fasting insulin and C-peptide levels were elevated, suggesting insulin resistance, in diabetic men without a family history, but not in diabetic men with a family history of diabetes (Table 3). Interaction terms were, however, not significant (P > 0.1). The adjusted ratio of fasting insulin to fasting glucose as well as the adjusted ratio AUC insulin to AUC glucose during the OGTT were reduced in diabetic men without a family history (-0.25 and -18.8, respectively). Among diabetic men with a family history, the reduction of the ratio of fasting insulin to fasting glucose was stronger (-0.56) and the reduction in the ratio of insulin to glucose during the OGTT was larger (-26.6, interaction P = 0.06). As shown in Figures 1 and 2 glucose curves during OGTT according to diabetes status were similar (Figure 1), while insulin curves were different for men with and without a family history (Figure 2). The larger impairment in insulin secretion relative to the glucose levels suggests that beta-cell function was more impaired in diabetic men 397 FAMILIAL AND NON-FAMIUAL DIABETES MELLITUS TABLE 2 Selected characteristics of elderly men with normal glucose tolerance and diabetes mellitus according to family history of diabetes: The Zutphen Elderly Study Normal glucose tolerance Diabetes mellitus Difference crude adjusted* Without a family history Age (years) Body mass index (kg/m2) Triceps skinfold (mm) Subscapular skinfold (mm) Skinfold ratio n = 329 75.7 ± 4.3 25.4 ± 3.0 12.1 ±4.5 16.9 ± 5.9 1.45 ±0.38 n = 53 75.5 ±4.7 26.1 ± 3 . 0 15.5 ±4.3 19.5 ±6.1 1.60 ±0.39 -02 0.7 3.4 2.6' 0.15* — 0.8 0.4 2.4* 0.13' With a family history Age (years) Body mass index (kg/m2) Triceps skinfold (mm) Subscapular skinfold (mm) Skinfold ratio n = 24 77.2 ± 5.6 25.3 ± 2.6 12.3 ±4.4 16.3 ±4.6 1.36 ±0.31 n = 15 75.1 ±3.5 25.8 ± 3.2 11.3 ±3.7 19.4 ± 8 3 1.80 ±0.74 -2.1 05 -1.0 3.1 0.44* 0.5 -0.9 2.1 0 26* Data we presented as mean ± SD. * Adjusted for age, *P < 0.05, '/> < 0.01. No significant interactions with family history were found. TABLE 3 Characteristics of glucose metabolism in elderly men with normal glucose tolerance and diabetes mellitus according to family history of diabetes: The Zutphen Elderly Study Normal glucose tolerance Difference Diabetes mellitus crude adjusted* Without a family history Fasting glucose (mmol/1) Fasting insulin (pmol/1) Fasting C-peptide (nmol/1) Insulin/glucose C-peptide/insulin (x 100) AUC glucose (mol/l.min)1 AUC insulin (nmol/1.min)" AUC insulin/AUC glucose1 n = 329 5.6 ± 0.5 64.3 ± 28.6 0.69 ± 0.27 1.58 ±0.66 3.43 ± 1.15 0.84 ±0.16 37.33 ± 16.87 44.51 ± 19.27 n = 53 8.7 ± 3.2 91.8 ±57.8 0 92 ± 0.53 1.61 ± 1.15 3.27 ± 1.02 1.55 ±0.28 38.55 ± 2 4 91 26.15 ± 17.90 3.1' 27.5' 0.23' 0.03 -0.16 0.71' 1.22 18.36' 1.3' 13.1' 0.101 -0.25' -0.18 0.45' -3.23 -18.80' With a family history Fasting glucose (mmol/1) Fasting insulin (pmol/1) Fasting C-peptide (nmol/1) Insulin/glucose C-peptide/insulin AUC glucose (mol/l.min)" AUC insulin (nmol/1.min)1 AUC insulin/AUC glucose1 n = 24 5.6 ± 0.6 62.6 ± 18.3 0.71 ±0.18 1.58 ±0.47 361 ± 1.16 0.87 ±0.16 36.84± 11.10 43.23 ± 12.75 n - 15 9.1 ± 2 . 5 71.8 ±34.3 0.81 ±0.24 1.18 ±0.59 3.81 ± 1.20 1.56 ±0.32 31.37 ± 11.22 20.56 ± 7.60 3.5' 9.2 0.10 -0.40* 0.20 0.69' -5.47 22.67' 1.7' 1.7 0.06 -0.56* 0.00 0.44' -8.25 -26.59' Data are presented as mean ± SD. • Adjusted for age, BMI and skinfold ratio. 1 n •= 36 in diabetic men without and n = 9 in diabetic men with a family history. AUC: Area under the oral glucose tolerance test curve. • P < 0.05, *P < 0.01, '/> < 0.001. Interaction with family history is significant (P = 0.061) for AUC insulin/AUC glucose. 398 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY 20 r 600r 500 15 400 o o a. 10 300 200 100 60 120 Time (min) 60 120 Time (min) FIGURE 1 Glucose levels (mean ± SE) during oral glucose tolerance test (OGTT) among men with normal glucose tolerance and diabetes according to family history of diabetes (The Zutphen Elderly Study) FIGURE 2 Insulin levels (mean ± SE) during oral glucose tolerance test (OGTT) among men with normal glucose tolerance and diabetes according to family history of diabetes (The Zutphen Elderly Study) with a family history than in diabetic men without a family history. The ratio of C-peptide to insulin, an index of hepatic insulin clearance, was normal in diabetic men with a family history, but reduced, albeit not significantly, in diabetic men without a family history. In both diabetic men with and without a family history fasting triglyceride levels were elevated and HDLcholesterol levels were reduced (Table 4). Adjustment for age, obesity level and body fat distribution did not alter these results. The deteriorations seemed more serious in diabetic men with a family history. This was confirmed by the interaction of family history of diabetes in the association between serum triglyceride levels and diabetes prevalence (P = 0.075). For the other lipid variables no interaction was observed (P>0.\). in individuals without a family history impaired hepatic insulin clearance and insulin resistance seemed to contribute to glucose intolerance. In diabetic men with a family history, the disturbances in lipoprotein profile, as often seen in diabetic individuals,15 were more severe than in diabetic men without a family history. This effect was especially clear for triglycerides. Associations of indices of obesity and body fat distribution did not differ according to family history of diabetes. The men who participated in this study were 70 years and over. The prevalence of diabetes is therefore relatively high. We cannot exclude selective loss of subjects, due to death or non-cooperation because of health problems. However, associations of family history, obesity and lipoprotein levels with diabetes prevalence were as expected, l ' 2-8 ~"' 15 ' 16 suggesting that the influence on the results was probably small. If diabetic patients with a less severe state of the disease have survived, this would have resulted in a dilution of the observed differences in glucose and lipid metabolism. A family history of diabetes occurred three times more often in diabetic men than in non-diabetic men. This suggests the presence of genetic factors in the development of NIDDM, as presented earlier. ll - 28 ' 29 DISCUSSION The results of the present study show that a family history of diabetes interacts with characteristics of glucose metabolism and triglycerides in explaining its association with NIDDM. In diabetic individuals with a family history of diabetes there were indications for impaired pancreatic beta-cell function, while 399 FAMILIAL AND NON-FAMIUAL DIABETES MELUTUS TABLE 4 Upids and lipoproteins in elderly men with normal glucose tolerance and diabetes mellitus according to family history of diabetes: The Zutphen Elderly Study Normal glucose tolerance Diabetes mellitus Difference crude adjusted Without • family history Triglycerides (mmol/1) Total cholesterol (mmol/1) LDL-cholesterol (mmol/1) HDL-cbolesterol (mmol/1) n = 329 1.39 ±0.75 6.09 ± 1.07 4.28 ± 0.97 1.17 ±0.30 n = 53 1.70 ±0.79 6.11 ± 1.20 4.30 ± 1.08 1.04 ±0.25 0.31* 0.02 0.02 -0.13 f 0.23* 0.01 0.01 -0.09* With a family history Triglycerides (mmol/1) Total cholesterol (mmol/1) LDL-cholesterol (mmol/1) HDL-cholesterol (mmol/1) n = 24 1.19 ±0.43 6.09 ± 0.96 4.31 ±0.94 1.24 ±0.30 n = 15 1.86 ±0.72 6.52 ± 1.09 4.59 ± 1.07 1.08 ±0.21 0 67 f 0.43 0 28 -0.16 0.60* 0.08 0 02 -0.18 Data are presented as mean ± SD. • Adjusted for age, BMI and sldnfold ratio * P < 0.05,' P < 0.01. Interaction with family history is significant (/> = 0.075) for triglycerides. It is likely that men have been misclassified with respect to family history. Since, as is shown in this study, 57% of diabetic patients are not diagnosed, it is likely that subjects did not know whether or not a relative was diabetic. Some additional, but probably small, misclassification might have occurred from the lag time between the interview where family history was recorded (1985) and the determination of the diabetes status in 1990. If a better index for genetic susceptibility was used, reducing misclassification of family history, the association with diabetes prevalence would probably have been stronger. The relationship between diabetes prevalence and family history of diabetes might proceed from the fact that individuals with diabetic relatives are diagnosed more often than subjects without diabetic relatives. In our population, however, the percentage of men with a family history was comparable in known and newly diagnosed diabetic men. This suggests a true association between family history and non-insulin-dependent diabetes in the present study. Type of family history (maternal, paternal or from a sib) might influence obesity level and diabetes prevalence. Two studies showed that maternal family history is more strongly related to both.30-31 In the present study, however, no differences in obesity level, fat distribution, nor in insulin or glucose levels were found according to type of family history. Therefore no discrimination was made between different types of family history of diabetes. In our study, both diabetic men with and without a family history had a higher BMI than men with normal glucose tolerance, but these differences were not statistically significant. However, diabetic men had significantly higher values for subscapular skinfold thickness and skinfold ratio. These findings have been shown before,32 and confirm that body fat distribution is associated with glucose tolerance33"36 and might play a role in the aetiology of NIDDM.32 The differences in skinfolds between diabetic men and men with normal glucose tolerance were similar for the groups with and without a family history. In a study on 157 Japanese American men Fujimoto e/a/. 12 did not find differences in overweight and fat distribution between diabetic and non-diabetic subjects among individuals with a family history of diabetes, whereas large differences in the group without a family history were observed. In our study such clear interaction between obesity and family history was not seen. This may be caused by a different genetic background (Caucasians) or an older age. Diabetic subjects are known to lose weight, either by the disease itself or by therapy, and this might have influenced our results. The body fat index studied might also affect the outcome. In a study on female relatives of black NIDDM patients Osei et al.u were not able to show an association between BMI and glucose and insulin levels, whereas they observed a positive association with subscapular skinfold thickness, both in the relatives and in a control group. These findings indicate that effect modification by family history may depend on the body fat index studied. 400 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY Our data suggest that, although measured indirectly, beta-cell function was markedly impaired in diabetic men with a family history, and less in men without a family history of diabetes. Assuming the two-step mechanism for the development of NIDDM, with insulin resistance leading to beta-cell exhaustion and glucose toxicity,1 it may be argued that the diabetic men with a positive family history in our study represent a subgroup of diabetic individuals with a more severe state of the disease. This seems, however, an unlikely explanation since the proportion of men with newly diagnosed diabetes and duration from time of clinical diagnosis was similar in the group with and without a family history. Furthermore, adjustment for the time since diagnosis did not alter the results. O'Rahilly et al} demonstrated that impaired betacell function was the primary defect in the development of familial diabetes. Others, however, suggest that insulin resistance is the primary defect in subjects with a family history.14'37 Although our study is crosssectional in design, and employed a standard OGTT, instead of more sophisticated euglycaemic clamps, 38 the Findings seem to support the hypothesis that reduced beta-cell function is an important feature of familial diabetes, whereas insulin resistance, indicated by elevated fasting insulin levels,39 seems to be more important in diabetic individuals without a family history. This is also suggested by a study from Lemieux et a/.13 which compared 39 non-diabetic men with and without a family history, and observed lower fasting insulin levels in those with a family history. In our study, however, the suggestion of the impact of betacell function in familial diabetes mostly results from findings in diabetic men: those with a family history have a more reduced insulin response. This might be due to the fact that our study population consisted of older subjects. The men with a true genetic susceptibility may already have developed diabetes during the course of their life, and therefore differences in normoglycaemic men with and men without a family history were small. We observed that triglyceride levels were higher in men with diabetes than in men with normal glucose tolerance. The increase was most pronounced in diabetic men with a family history of diabetes. Haffner et al.16 observed higher lipid levels in relatives of diabetic patients, but Osei et al. could not confirm this finding in a small study among relatives of black NIDDM patients.14 Elevated triglyceride levels may result from an increased production or decreased clearance of VLDL. Insulin stimulates triglyceride production, 40 but this seems no explanation why in this study higher levels of triglycerides are found in diabetic men with a family history, since they produce less insulin, than the men without a family history. Another explanation may be the suppressive effect of insulin on VLDL secretion.41 In diabetic subjects the hepatic sensitivity to the suppressive effect of insulin might be impaired.13142 A third potential explanation may refer to the effect of lipoprotein lipase polymorphism on VLDL-triglyceride levels43 and a subsequently impaired insulin action.44-45 Patients with familial combined hyperlipidaemia show partial lipoprotein lipase deficiency,46 and lipid and glucose metabolism seem to be linked in these patients.47 The difference in triglyceride elevation between diabetic men with and without a family history might therefore be explained by a more severe defect in insulin action in the group with a family history or an additive genetic effect in lipoprotein lipase in the group with a family history of diabetes. In conclusion, our findings suggest that elderly diabetic men who have a family history of diabetes represent a different subgroup in comparison to elderly diabetic men who do not have such a history, characterized by larger deteriorations in beta-cell function and higher triglyceride levels. No interaction between indices of body fatness and a family history of diabetes was found. ACKNOWLEDGEMENTS We are indebted to the fieldwork team in Zutphen, especially to Dr B P M Bloemberg, for co-ordinating the survey, to Dr E B Bosschieter and Mrs A H ThomassenGijsbers; to the Laboratory of Clinical Chemistry and Haematology of the Nieuwe Spitaal hospital in Zutphen; to the Laboratory of the Unit of Teratology, Endocrinology and Perinatal Screening of the RIVM (Dr G Loeber and Mr L Elvers) for the insulin and C-peptide analyses; and to Dr E G Schouten of the Dept of Epidemiology and Public Health, Agricultural University, Wageningen, for his comments. This study is presented in abstract form in TSG 1994;4:17 (in Dutch). REFERENCES 'Zimmet P. Non-insulin-dependent (Type 2) diabetes mellitus: Does it really exist? Diabetic Med 1989; 6: 728-35. 2 Reaven G M. Insulin secretion and insulin action in non-insulindependent diabetes mellitus: which defect is primary? Diabetes Care 1984; 7(Suppl. 1): 17-24. 3 O'Rahilly S P, Nugent Z, Rudenski A S et al. Beta-cell dysfunction, rather than insulin insensitivity, is the primary defect in familial Type 2 diabetes. Lancet 1986; II: 360-63. 4 Haffner S M, Stem M P, Hazuda H P, Mitchell B D, Patterson J K. Increased insulin concentrations in nondiabetic offspring of diabetic parents. N Engl J Med 1988; 319: 1297-301. FAMILIAL AND NON-FAMILIAL DIABETES MELLITUS 5 Martin B C, Warram J H, Krolewski A S, Bergman R N, Soeldner J S, Kahn C R. Role of glucose and insulin resistance in development of type 2 diabetes mellitus: results of a 25-year follow-up study. Lancet 1992; 340: 925-29. 'Eriksson J, Franssila-Kallunki A, Ekstrand A et al. Early metabolic defects in persons at increased risk for non-insulindependent diabetes mellitus. N Engl J Med 1989; 321: 7 Lillioja S, Mott D M, Spraul M et al. Insulin resistance and insulin secretory dysfunction as precursors of non-insulindependent diabetes mellitus. N Engl J Med 1993; 329: 1988-92. 'Colditz G A, Willet W C, Stampfer M J et al. Weight as a nsk factor for clinical diabetes in women. Am J Epidemiol 1990; 132: 501-13. 9 Knowler W C, Petutt D J, Savage P J, Bennett P H. Diabetes incidence in Pima indians; contributions of obesity and parental diabetes. Am J Epidemiol 1981; 113: 144-56. 10 Kuzuya T, Matsuda A. Family histories of diabetes among Japanese patients with Type 1 (insulin dependent) and Type 2 (non-insulin-dependent) diabetes. Diabetologia 1982; 22: 372-74. " Morris R D, Rimm D L, Hartz A J, Kalkhoff R K, Rimm A A. Obesity and heredity in the etiology of non-insulindependent diabetes mellitus in 32,662 adult white women. Am J Epidemiol 1989; 130: 112-21. 12 l3 Fujimoto W Y, Leonetti D L, Newell-Morris L, Shuman W P, Wahl P W. Relationship of absence or presence of a family history of diabetes to body weight and body fat distribution in Type 2 diabetes. Int J Obes 1991; 15: 111-20. Lemieux S, Despres J-P, Nadeau A, Prud'homme D, Tremblay A, Bouchard C. Heterogeneous glycaemic and insulinaemic responses to oral glucose in non-diabetic men: interactions between duration of obesity, body fat distribution and family history of diabetes mellitus. Diabetologia 1992; 35: 653-59. 14 Osei K, Cottrell D A, Bossetti B. Relationships of obesity indices to serum insulin and lipoproteins in relatives of black patients with non-insulin-dependent diabetes mellitus (NIDDM). Int J Obes 1991; 15:441-51. 13 Howard B V. Lipoprotein metabolism in diabetes mellitus. J Ltpid Res 1987; 28: 613-28. 16 Haffner S M, Stem M P, Hazuda H P, Mitchell B D, Patterson J K, Ferrannini E. Parental history of diabetes is associated with increased cardiovascular risk factors. Arteriosclerosis 1989; 9: 928-33. l7 Hertog M G L, Feskens E J M, Hollman P C H, Katan M B, Kromhout D. Dietary flavonoids and the risk of coronary heart disease (The Zutphen Elderly Study). Lancet 1993; 342: 1007-11. "Keys A, Aravanis C, Blackburn H. Epidemiological studies related to coronary heart disease: characteristics of men aged 40-59 in seven countries. Ada Med Scan 1966; 460 (Suppl.): 1-392. "Zimmet P Z, Tuomi T, Mackay I R et al. Latent Autoimmune Diabetes Mellitus in Adults (LADA): the role of antibodies to glutamic acid decarboxylase in diagnosis and prediction of insulin dependency. Diabetic Med 1993; 11: 299-303. ^Richterich R, Dauwalder H. Zur bestimmung der plasmaglucosekonzentration mit der hexokinase-glucose-6-phosphatdehydrogenase-methode. Schweiz Med Wochenschr 1971; 101:615-18. 401 21 World Health Organization. Diabetes Mellitus. Report of a WHO Study Group. WHO Technical Report Series 727. WHO, Geneva, 1985. 22 Sullivan D R, Kruijswijk Z , W e s t C E , K o h l m e i e r M , Katan M B . Determination of serum triglycendes by an enzymatic method not affected by free glycerol. Clin Chem 1985; 31: 1227-28. 23 Wamick G R, Benderson J, Albers J J. Dextran sulfate-Mg2+ precipitation procedure for quantitation of high density lipoprotein cholesterol. Clin Chem 1982; 28: 1379-88. 24 Siedel J, Schlumberger H, Klose S, Ziegenhorn J, Wahlefeld A W. Improved reagent for the enzymatic determination of serum cholesterol. J Clin Chem Clin Biochem 1981; 19: 838-39. "StShler F, Gruber W, Stinshoff K, Raschlau P. Eine pnuisgerechte enzymatische cholesterin-bestimmung. Med Lab 1977; 30: 29-37. 26 Friedewald W T, Levy R I, Frednckson D S. Estimation of the concentration of low-density lipoprotein cholesterol in plasma without use of the preparative ultracentrifugation. Clin Chem 1972; 18: 499-502. "Rothman K J, Greenland S, Walker A M. Concepts of interaction. Am J Epidemiol 1980; 112: 467-70. ^Elbein S C, Maxwell T M, Schumacher C. Insulin and glucose levels and prevalence of glucose intolerance in pedigrees with multiple diabetic siblings. Diabetes 1991; 40: 1024-32. 29 Stem M P, Gonzalez C, Mitchell B D, Villalpando E, Haffner S M, Hazuda H P. Genetic and environmental determinants of Type II diabetes in Mexico City and San Antonio. Diabetes 1989; 41: 484-92. ^Alcolado J C, Alcolado R. Importance of maternal history of non-insulin-dependent diabetic patients. Br Med J 1991; 302: 1178-80. 31 Lapidus L, Lissner L, Bengtsson C, Smith U. Family history of diabetes in relation to different types of obesity and change of obesity during 12-yr period. Diabetes Care 1992; 15: 1455-58. 32 Feskens E J M , K r o m h o u t D . C a r d i o v a s c u l a r risk factors a n d the 25-year diabetes incidence in middle-aged men. The Zutphen Study. Am J Epidemiol 1989; 130: 1101-08. 33 Feskens E J M , K r o m h o u t D . Effects of body fatness and its development over a ten-year period on glucose tolerance in euglycaemic men: the Zutphen Study. Int J Epidemiol 1989; 18: 368-73. ^Freedman D S, Srinivasan S R, Harsha D W, Webber L S, Berenson L S. Relation of body fat distribution to hyperinsulinemia in children and adolescents: The Bogalusa Heart Study. Am J Clin Nutr 1987; 46: 403-10. 33 Harlan L C, Harlan W R, Landis J R, Goldstein N G. Factors associated with glucose tolerance in U.S. adults. Am J Epidemiol 1987; 126: 674-84. 36 Newell-Morris L L, Treder R P, Shuman W P, Fujimoto W Y. Fatness, fat distribution, and glucose tolerance in secondgeneration Japanese American (Nisei) men. Am J Clin Nutr 1989; 50: 9-18. 37 Groop L C, Bonadonna R C, DelPrato S et al. Glucose and free fatty acid metabolism in non-insulin-dependent diabetes mellitus. Evidence for multiple sites of insulin resistance. J Clin Invest 1989; 84: 205-13. 38 Groop L C, Widen E, Ferrannini E. Insulin resistance and insulin deficiency in the pathogenesis of Type 2 (non-insulindependent) diabetes mellitus: errors of metabolism or of methods? Diabetologia 1993; 36: 1326-31. 402 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY " H o l l e n b e c k C B, Chen N, Chen Y D 1, Reaven G M. Relationship between the plasma insulin response to oral glucose and insulin-stimulated glucose utilisation in normal subjects. Diabetes 1984; 33: 460-63. 40 Reaven G M, Chen Y-D. Role of insulin in regulation of lipoprotein metabolism in diabetes. Diabetes Metab Rev 1988; 4: 639-52. 41 Byrne C D, Wang T W, Hales C N. Control of Hep G2-cell triacylglycerol and apolipoprotein B synthesis and secretion by polyunsaturated non-esterified fatty acids and insulin. BiochemJ 1992; 288: 101-07. 4J Gibbons G F. Assembly and secretion of hepatic very-lowdensity lipoprotein. Biochem J 1990; 268: 1-13. 43 Peacock R E, Hamsten A, Nilsson-Ehle P, Humphries S E. Associations between lipoprotein lipase gene polymorphism and plasma correlations of lipids, lipoproteins and lipase activities in young myocardial infarction survivors and agematched healthy individuals from Sweden. Atherosclerosis 1992; 97: 171-85. " B i e g e r W P, Michel G, Barwich D, Biehl K, Wirth A. Diminished insulin receptors on monocytes and erythrocytes in hypertriglyceridemia. Metabolism 1984; 3 3 : 982-87. 43 Yki-JBrvinen H, Taskinen M-R. Interrelationships among insulin's antilipolytic and glucoregulatory effects and plasma triglycerides in non-diabetic and diabetic patients with endogenous hypertriglyceridemia. Diabetes 1988; 37: 1271-78. 46 Babirak S P, Iverius P H, Fujimoto W Y, Brunzell J D. Detection and characterization of the heterozygote state for lipoprotein lipase deficiency. Arteriosclerosis 1989; 9: 326-34. 47 Cabezas M C, de Bruin T W A, de Valk H W, Shoulders C C. Jansen H, Erkelens W. Impaired fatty acid metabolism in familial combined hyperlipidaemia: a mechanism associating hepatic apolipoprotein B overproduction and insulin resistance. J Clin Invest 1993; 92: 160-68. (Revised version received September 1995)