Judith K. Patterson

Incidence of Type II Diabetes in Mexican Americans Predicted by Fasting Insulin and Glucose Levels, Obesity, and Body-Fat Distribution

Few data exist on predictors of non-insulin-dependent (type II) diabetes mellitus. We examined body mass index (BMI), ratio of subscapular-to-triceps skin fold (centrality index), and fasting glucose and insulin concentrations as predictors of decompensation to type II diabetes in Mexican Americans, a population at high risk for this disorder. Twenty-eight of 474 initially nondiabetic Mexican Americans developed type II diabetes after 8 yr of follow-up. Converters to diabetes were older and had higher BMIs, centrality indices, and fasting glucose and insulin concentrations than nonconverters. Subjects in the highest quartile of the insulin distribution had 6.6 times the risk of developing type II diabetes as subjects in the remaining three quartiles combined (95% confidence interval [CI] = 3.14–13.7). In multivariate analysis, fasting glucose (odds ratio [OR] = 5.80, 95% CI = 2.57–13.1) and insulin (OR = 3.12, 95% CI = 1.36–7.14) remained significantly related to conversion to diabetes. However, BMI and centrality index, which were significantly related to conversion in the univariate analysis, were no longer significant in the multivariate analysis once glucose and insulin concentrations were taken into consideration, suggesting that the effect of these variables may be mediated by insulin resistance. Nearly half of the incident cases developed in a subset of the population who were simultaneously in the highest quartile of both fasting insulin and glucose concentrations (population-attributable risk 44.2%). Our results support the insulin resistance/pancreatic exhaustion theory of type II diabetes.

Do Upper-Body and Centralized Adiposity Measure Different Aspects of Regional Body-Fat Distribution? Relationship to Non-Insulin-Dependent Diabetes Mellitus, Lipids, and Lipoproteins

Both central and upper-body adiposity are associated with high rates of type II non-insulin-dependent diabetes mellitus (NIDDM), high triglyceride levels, and low high-density lipoprotein (HDL) cholesterol levels. Previous data have also suggested that central and upperbody adiposity are relatively uncorrelated and hence may measure different aspects of regional body fat distribution. We assessed body mass index (BMI), the ratio of subscapular-to-triceps skinfold (STR), the ratio of waist-to-hip circumference (WHR), lipids, lipoproteins, and glucose tolerance in 738 Mexican Americans (ages 25-64 yr), who participated in the San Antonio Heart Study, a population-based study of diabetes and cardiovascular risk factors. NIDDM was diagnosed according to National Diabetes Data Group criteria. In general, STR and WHR were associated with high NIDDM rates, low HDL cholesterol levels, and high triglyceride levels, although WHR was somewhat more predictive of these than STR. In females, BMI, WHR, and STR all made independent contributions to prediction of NIDDM and HDL cholesterol; in males, WHR and STR both made independent contributions to prediction of triglyceride levels. This suggests that both indices may measure different aspects of body-fat distribution. Investigators should consider measuring both of these indicators of body-fat distribution in studies of diabetes and other cardiovascular risk factors, although if only a single measure is feasible, WHR appears to be preferable.

Hyperinsulinemia in a Population at High Risk for Non-Insulin-Dependent Diabetes Mellitus

The prevalence of non-insulin-dependent diabetes mellitus (NIDDM) is higher in Mexican Americans than in non-Hispanic white Americans, even after adjustment for the formers greater overall and more centralized adiposity. We postulated that this excess risk of NIDDM could be due to resistance to insulin. We performed oral glucose-tolerance tests with measurements of serum insulin concentrations in 225 Mexican Americans and 180 non-Hispanic whites without diabetes as part of the San Antonio Heart Study, a population-based study of risk factors for diabetes. Changes in serum insulin concentrations in response to the glucose challenge were quantified by the area under the serum insulin curve. Overall adiposity was characterized by body-mass index, and regional body-fat distribution by the ratio of subscapular to triceps skinfolds and the ratio of waist to hip circumference. After adjustment for these indicators of adiposity and also for differences in glucose tolerance, Mexican Americans were found to have significantly greater areas under the serum insulin curve than non-Hispanic whites. These data suggest that, like other populations at high risk for NIDDM such as Pima Indians and Micronesians, Mexican Americans have more hyperinsulinemia than can be accounted for by their adiposity.

Hyperinsulinemia, upper body adiposity, and cardiovascular risk factors in non-diabetics

Previous studies have suggested that hyperinsulinemia and upper body adiposity are each separately associated with elevated BP and triglyceride (TG) levels, and with lower high density lipoprotein (HDL) cholesterol levels. The joint effect of hyperinsulinemia and upper body adiposity on lipids, lipoproteins, and BP, however, has not been previously studied. We hypothesized that the effect of body fat distribution on cardiovascular risk factors might be mediated through hyperinsulinemia. We measured BP, lipids and lipoproteins, HDL subfractions, and insulin and glucose concentrations as part of the San Antonio Heart Study, a population-based study of diabetes and cardiovascular risk factors. Insulinemia and glycemia were assessed as the sum of the fasting, half-hour, one-hour, and two-hour insulin and glucose levels, respectively, measured during a standardized oral glucose tolerance test. Individuals who had diabetes according to National Diabetes Data Group criteria were excluded from the analyses. In univariate analyses, both hyperinsulinemia and waist-to-hip ratio (WHR), a measure of upper body adiposity, were positively associated with TG and negatively associated with total HDL and HDL2 cholesterol levels. However, when the effects of glycemia and insulinemia were controlled for by analysis of variance, WHR was no longer significantly related to TG levels. By contrast, WHR continued to be inversely related to total HDL and HDL2 cholesterol even after adjustment for glycemia and insulinemia. Hyperinsulinemia was only weakly related to HDL cholesterol. These results suggest that insulinemia and glycemia might mediate the effects of upper body adiposity on TG, although not on HDL and HDL2 cholesterol. Hyperinsulinemia was also positively associated with diastolic and systolic BP in men.

Increased Insulin Concentrations in Nondiabetic Offspring of Diabetic Parents

Abstract Insulin resistance is thought by many to be the primary defect that results in non-insulin-dependent diabetes mellitus (NIDDM). An implication of this theory is that prediabetic persons have higher serum insulin levels than normal subjects. We assessed serum insulin concentrations in a cohort of 1497 nondiabetic Mexican Americans, a population at high risk for NIDDM, according to whether their parents or siblings had diabetes. It was assumed that prediabetic persons would be more likely to have strong family histories of diabetes. We found a stepwise increase in fasting insulin levels in nondiabetics with neither, one, or both parents with diabetes (69.8, 77.8, and 94.6 pmol per liter, respectively; P = 0.002). Similar results were observed for insulin sum (the total of insulin concentrations in the fasting state and at 30, 60, and 120 minutes after a 75-g oral glucose load). The differences in insulin sums according to family history remained statistically significant in analyses of covariance, ...

Microalbuminuria. Potential marker for increased cardiovascular risk factors in nondiabetic subjects

Microalbuminuria is associated with progression to renal disease in insulin-dependent diabetes and with increased mortality in noninsulin-dependent diabetes. In contrast, few studies have addressed the effect of microalbuminuria on cardiovascular risk in nondiabetics. We, therefore, determined the level of microalbuminuria in 316 nondiabetic subjects from the San Antonio Heart Study, a population-based study of diabetes and cardiovascular risk factors. Microalbuminuria (greater than or equal to 30 mg/l) was found in 42 of these 316 subjects (13%). Subjects with microalbuminuria had significantly higher blood pressure, triglyceride concentration, sum of insulin concentrations during a glucose tolerance test, and prevalence of hypertension and of self-reported myocardial infarction than subjects without microalbuminuria. When subjects with hypertension were excluded (n = 27), normotensive subjects with microalbuminuria (n = 31) still had significantly higher triglyceride concentrations and insulin sum than normotensive subjects without microalbuminuria (n = 258), suggesting that an increased atherogenic risk factor pattern exists even in normotensive subjects with microalbuminuria. Microalbuminuria may be a marker for cardiovascular risk, although it is not certain whether microalbuminuria causes these metabolic changes or results from some metabolic disturbance such as insulin resistance.

Diabetic Retinopathy in Mexican Americans and Non-Hispanic Whites

Mexican Americans (MAs) have a threefold greater prevalence of non-insulin-dependent diabetes mellitus (NIDDM) than non-Hispanic Whites (NHWs). Because MA diabetic subjects have greater hyperglycemia and an earlier age of onset than NHW diabetic subjects, we postulated that diabetic MAs might also have more severe diabetic retinopathy. Stereoscopic retinal photographs of the seven standard fields of each eye were taken in 257 MAs and 56 NHWs with NIDDM. The photographs were read by the University of Wisconsin Fundus Photographic Reading Center and graded with standardized criteria. The MAs had a nonsignificantly increased risk of retinopathy relative to the NHWs [odds ratio (OR) = 1.71; 95% confidence interval (CI) = (0.93, 3.17)]. The risk of severe retinopathy (proliferative or preproliferative) relative to background or no retinopathy was significantly greater in MAs than in NHWs [OR = 2.37; 95% CI = (1.04, 5.39)]. After control by logistic regression for duration of disease, severity of hyperglycemia, age, and systolic blood pressure, MAs still had an increased risk of severe retinopathy relative to NHWs [OR = 3.18; 95% CI = (1.32, 7.66)]. Severe retinopathy was related to duration of disease, hyperglycemia, and insulin therapy in both ethnic groups. Previously diagnosed MA diabetic subjects also had an increased prevalence of any retinopathy [OR = 2.39; 95% CI = (1.63, 3.50)] and severe retinopathy [OR = 3.21 ; 95% CI = (2.24, 4.59)] relative to previously diagnosed White diabetic subjects (n = 896) from Wisconsin. The combination of an increased prevalence of NIDDM in MAs plus an increased severity of retinopathy in those MAs who have diabetes suggests that a major public health effort should be made to screen this ethnic group for retinopathy.

Increased Incidence of Type II Diabetes Mellitus in Mexican Americans

Objective To determine whether Mexican Americans have an increased incidence of non-insulin-dependent (type II) diabetes mellitus relative to non-Hispanic whites. Currently, no study has reported on the incidence of this disorder in Mexican Americans. Research Design and Methods We determined the 8-yr incidence of type II diabetes in 617 Mexican Americans and 306 non-Hispanic whites who participated in the San Antonio Heart Study, a population-based study of diabetes and cardiovascular disease. Forty Mexican Americans (6.5%) and 6 non-Hispanic whites (2%) developed type II diabetes, as defined by World Health Organization criteria. The age-adjusted ethnic odds ratio (OR; Mexican Americans/non-Hispanic whites) for diabetes incidence was 8.13 (95% confidence interval [C1] 1.10–59.9) in men and 3.62 (95% CI 1.37–9.55) in women. We adjusted for age, sex, ethnicity, body mass index, and level of educational attainment with multiple logistic regression analyses. Results Mexican Americans continued to show a statistically significant increase in diabetes incidence (OR 2.72, 95% CI 1.02–7.28). Obesity and age were also positively related to diabetes incidence in this analysis (P < 0.001). In addition, subjects with at least some college education had a lower incidence of diabetes than those with less than a high school education (OR 0.51, 95% CI 0.26–0.99). Conclusions The incidence of type II diabetes in Mexican Americans is greater than in non-Hispanic whites, a difference that is not explained by ethnic differences in obesity, age, or level of educational attainment.

Myocardial infarction in Mexican-Americans and non-Hispanic whites. The San Antonio Heart Study.

Mexican-American men experience reduced cardiovascular mortality compared with non-Hispanic white men. There is no corresponding ethnic difference in cardiovascular mortality in women. The difference in men could result either from a lower incidence of cardiovascular disease or a lower case fatality rate among Mexican-Americans. Although the incidence of cardiovascular disease in Mexican-Americans is unknown, we have collected data on prevalence of myocardial infarction in 5,148 individuals examined in the San Antonio Heart Study, a population-based survey of cardiovascular disease conducted between 1979 and 1988 in Mexican-Americans and non-Hispanic whites aged 25-64 years. Myocardial infarction was assessed by Minnesota-coded electrocardiograms and by a self-reported history of a physician-diagnosed heart attack. For both end points, the age-adjusted prevalence of myocardial infarction was lower in Mexican-American men than in non-Hispanic white men. After adjustment for age and diabetes status (present/absent), the odds of a myocardial infarction, as defined by either criterion, was approximately one third lower in Mexican-American men than in non-Hispanic white men (p = 0.06). In women, the prevalence of both myocardial infarction end points was slightly higher in Mexican-Americans than in non-Hispanic whites, although neither of these differences was significant. Although the ethnic differences in prevalence in this study were not statistically significant, their pattern parallels the pattern in the mortality due to cardiovascular diseases. Therefore, the results support the hypothesis that the reduced cardiovascular mortality rate observed in Mexican-American men reflects a lower incidence of myocardial infarction rather than a reduced case fatality rate because the latter would result in a higher prevalence.

Proteinuria in Mexican Americans and Non-Hispanic Whites With NIDDM

Mexican Americans have a threefold greater prevalence of non-insulin-dependent diabetes mellitus (NIDDM) than non-Hispanic Whites as found in the San Antonio Heart Study, a population-based study of diabetes and cardiovascular disease. In addition, Mexican-American d a etic subjects have higher levels of glycemia than non-Hispanic White diabetic subjects. We therefore hypothesized that the prevalence of clinical proteinuria would be greater among Mexican-American diabetic subjects (n = 317) than among non-Hispanic White diabetic subjects (n = 67). Clinical proteinuria, defined as ≥1 + on the Ames Albustix test, was 2.82 times more prevalent in Mexican-American diabetic subjects compared with non-Hispanic White diabetic subjects adjusting for age and duration (95% confidence interval [Cl] = 1.05, 7.55; P = .039). After controlling for other possible confounding variables (i.e., glycemia, systolic b ood pressure, smoking, and insulin use), the excess of proteinuria in Mexican-American diabetic subjects was only slightly attenuated, although the statistical significance became borderline (odds ratio [OR] = 2.59, 95% Cl = 0.91, 7.32; P = .072). The prevalence of microalbuminuria (>30 mg/L) was also significantly higher in Mexican-American diabetic subjects than in non-Hispanic White diabetic subjects (OR = 3.54, 95% Cl = 1.28, 9.81; P = .015). We also compared previously diagnosed Mexican-American diabetic subjects (n = 243) from San Antonio with previously diagnosed non-Hispanic White diabetic subjects in Wisconsin (n = 476). After controlling for age and duration of diabetes, Mexican-American diabetic subjects had a significantly higher prevalence of clinical proteinuria than the Wisconsin diabetic subjects (Mantel-Haenszel OR = 1.58, 95% Cl = 1.05, 2.40; P = .017). Our data indicate that Mexican-American diabetic ubjects have a higher prevalence of clinical proteinuria and microalbuminuria than non-Hispanic White diabetic subjects.