Search

Background: Cardiovascular disease (CVD) in insulin dependent diabetes mellitus (IDDM) has been linked to renal disease. However, little is known concerning international variation in the correlations with hyperglycaemia and standard CVD risk factors., Methods: A cross-sectional comparison was made of prevalence rates and risk factor associations in two large studies of IDDM subjects: the Pittsburgh Epidemiology of Diabetes Complications Study (EDC) and the EURODIAB IDDM Complications Study from 31 centres in Europe. Subgroups of each were chosen to be comparable by age and duration of diabetes. The EDC population comprises 286 men (mean duration 20.1 years) and 281 women (mean duration 19.9 years); EURODIAB 608 men (mean duration 18.1 years) and 607 women (mean duration 18.9 years). The mean age of both populations was 28 years. Cardiovascular disease was defined by a past medical history of myocardial infarction, angina, and/or the Minnesota ECG codes (1.1-1.3, 4.1-4.3, 5.1-5.3, 7.1)., Results: Overall prevalence of CVD was similar in the two populations (i.e. men 8.6% versus 8.0%, women 7.4% versus 8.5%, EURODIAB versus EDC respectively), although EDC women had a higher prevalence of angina (3.9% versus 0.5%, P < 0.001). Multivariate modelling suggests that glycaemic control (HbA1c) is not related to CVD in men. Age and high density lipoprotein cholesterol predict CVD in EURODIAB, while triglycerides and hypertension predict CVD in EDC. For women in both populations, age and hypertension (or renal disease) are independent predictors. HbA1c is also an independent predictor-inversely in EURODIAB women (P < 0.008) and positively in EDC women (P = 0.03). Renal disease was more strongly linked to CVD in EDC than in EURODIAB., Conclusions: Despite a similar prevalence of CVD, risk factor associations appear to differ in the two study populations. Glycaemic control (HbA1c) does not show a consistent or strong relationship to CVD.

Urinary samples were concentrated rapidly and efficiently and were used to develop several protein assays that may be of value in monitoring individuals with progressive renal disorders. Transforming growth factor-beta1 (TGF-11) and retinol binding protein (RBP) were measured with modification of commercially available methods used to assay serum specimens; type 3 collagen (T3C) was measured with a new immunonephelometric assay. The precision characteristics of these assays are comparable with those reported for microalbuminuria. The clinical utility of measuring a panel of these markers was demonstrated in urine samples from 16 control subjects and from 46 individuals with insulin-dependent diabetes mellitus (IDDM) with various albumin excretion rates (AERs). TGF-beta1 and T3C were used as markers of cytokine expression and of the renal fibrogenic process, whereas RBP excretion served as a marker of tubular injury or dysfunction. Compared with controls, T3C excretion was significantly increased in 18 normoalbuminuric and further increased in 13 microalbuminuric (AER 20 < or = 200 microg/min) IDDM subjects. RBP excretion was increased in macroalbuminuric IDDM subjects (AER >200 microg/min, overt nephropathy). Significant correlations were also found between AER and RBP in all but macroalbuminuric individuals, whereas TGF-beta1 correlated with T3C excretion in controls and in normoalbuminuric diabetic subjects. Urinary RBP but not AER was an excellent predictor of diabetic nephropathy as defined by serum creatinine (P = 0.0001). This underscores the importance of an early tubulopathy in the subsequent development of glomerulopathy and overt nephropathy. The data suggest that longitudinal monitoring of a panel of urinary markers such as that used in the current study may better define their relevance in progressive glomerulosclerosis and may also provide greater insight into the mechanisms underlying such process.

The authors used 4-year incidence data from the Pittsburgh Epidemiology of Diabetes Complications (EDC) Study to investigate the wider applicability of recent research findings that demonstrate an association between glycemic control and insulin-dependent diabetes mellitus (IDDM) complications. EDC subjects participated in clinical examination at baseline (1986-1988) and were followed up every 2 years. Results demonstrated that, during the first 4 years of follow-up, subjects who were in "poor" control (glycosylated hemoglobin (GHb) > or = 11%) at baseline were significantly (p < 0.001) more likely to develop microalbuminuria, proliferative retinopathy, and distal symmetrical polyneuropathy (DSP), compared with subjects who were in "fair" control (GHb < 11%). Subjects who were in poor control were somewhat more likely to develop overt nephropathy (p = 0.08) and renal failure (p = 0.085) during follow-up; however, no associations were observed with either coronary heart disease or lower extremity arterial disease (LEAD). These results confirm the strong association between prior glycemic control and the onset of microalbuminuria, proliferative retinopathy, and DSP observed in the Diabetes Control and Complications Trial study. However, the results of the study suggest weaker associations for the later stages of renal disease, and little relation was seen between glycemic control and LEAD or coronary disease. Other risk factors may be more important for the development of the later complications of IDDM. Further follow-up is necessary in order to rule out type II error.

The beneficial effect of physical activity in the general population is well known, but, to the authors' knowledge, has not been reported for persons with insulin-dependent diabetes mellitus. In a cohort of 548 diabetes patients followed as part of the Pittsburgh Insulin-dependent Diabetes Mellitus Morbidity and Mortality Study, physical activity was ascertained by survey in 1981, and mortality was ascertained through January 1, 1988. Cases were also compared with non-diabetic sibling controls. Activity level among cases varied inversely with the occurrence of diabetic complications. Overall activity level was inversely related to mortality risk. Sedentary males (< 1,000 kcal/week) were three times more likely to die than active males (> 2,000 kcal/week). A similar, but statistically nonsignificant, relation was seen in females. Cox proportional hazards analysis controlling for potential confounders (age, body mass index, insulin dose, reported diabetes complications, cigarette smoking, and current alcohol drinking) similarly revealed that activity level was inversely associated with mortality risk. Comparison of cases with non-diabetic sibling controls identified similar activity levels for the two groups. The results suggest that activity is not detrimental with regard to mortality, and may in fact provide a beneficial effect in terms of longevity in diabetes patients.

The risk for insulin-dependent diabetes mellitus (IDDM) associated with genetic susceptibility markers at the human leukocyte antigen (HLA) DQA1 and DQB1 loci was evaluated among individuals with and those without islet cell antibodies. A total of 108 antibody-positive parents and siblings of IDDM patients from the Pittsburgh registry were identified among 1,592 who were screened. HLA-DQ molecular typing was performed on 79 of these individuals and on 78 antibody-negative relatives. There were similar proportions of homozygotes for both of the diabetogenic alleles DQA1 arginine-52 (R/R) and DQB1 non-aspartate-57 (nD/nD) among the antibody-positive and antibody-negative relatives (19.0 and 15.4%, respectively). However, subsequent development of IDDM was restricted to individuals who were both antibody positive and carried the potential to make at least one diabetogenic DQ heterodimer. A dose-response effect was observed among the antibody-positive relatives, in which two of 18 capable of generating one diabetogenic heterodimer and six of 29 generating two heterodimers became insulin requiring. Nine of 15 who were homozygous for both R/R and nD/nD, coding exclusively for diabetogenic variants, became diabetic over the course of the follow-up. With a multivariate model, the relative risk for IDDM among those with islet cell antibodies who were also R/R and nD/nD was estimated to be 229.3 compared with those lacking both, after age and sex were controlled for. The data suggest that while autoimmunity, indicated by the presence of cytoplasmic islet cell antibodies may be relatively common, it progresses only in those with variant HLA-DQ molecules.

Body mass and body fat distribution are important considerations in the study of hypertension. However, few studies have investigated the relationships with regards to race differences in elevated arterial pressure. A population-based sample of black and white adults was assessed by interview and physical measurement. The prevalence of hypertension (defined as 140/90 mmHg and/or medically treated) was disproportionately higher among blacks than whites. In addition, blacks had a higher prevalence of the more severe hypertension (160/95 mmHg) and hypertension with higher prevalence at earlier ages than whites. Black females had a significantly higher distribution of body mass index (BMI) than white females, while no difference was found in the distributions of males. White males had a higher distribution of waist to hip ratio (WHR) than black males, while black females had the higher values compared to white females. The prevalence of hypertension increased with BMI and WHR. Blacks maintained higher rates of hypertension after controlling for BMI and WHR, however, the margin of difference diminished when BMI and WHR was considered together. The black-white difference in hypertension was not completely explained by BMI and WHR. In addition, the strength of the association of hypertension and body size was different for blacks and whites which suggests possible differences in the mechanisms regulating blood pressure.

The relation between serum albumin levels and subsequent incidence of myocardial infarction and coronary heart disease deaths was evaluated using stored serum from the Multiple Risk Factor Intervention Trial (MRFIT). There were 91 coronary heart disease deaths, 113 myocardial infarction patients, and 405 controls matched to cases within 5 years of age, treatment group, and clinic site. There was a highly significant inverse relation between serum albumin level and risk of coronary heart disease. Individuals with a baseline level of serum albumin greater than or equal to 4.7 g/dl had an odds ratio of 0.45 as compared with individuals with a baseline level of serum albumin less than 4.4 g/dl. The relation persisted after adjusting for other cardiovascular risk factors (blood pressure, smoking, and serum cholesterol). The association was stronger for coronary heart disease deaths than for surviving myocardial infarction patients, and for cigarette smokers as compared with cigarette nonsmokers. The deaths studied occurred in the time period at least 6 years after the sera had been obtained and up to 10.5 years of follow-up, and the myocardial infarctions studied occurred within the first 6.5 years of follow-up. There was no consistent relation between time and death due to coronary heart disease or myocardial infarction and albumin levels. Albumin levels are related to the acute phase reaction. Lower albumin levels may be a marker of persistent injury to arteries and progression of atherosclerosis and thrombosis. The consistent relation between albumin and coronary heart disease risk requires further evaluation.

The roles of potential risk factors for background and proliferative retinopathy were evaluated in cross-sectional analyses from the Epidemiology of Diabetes Complications Study, Pittsburgh, Pennsylvania. This report presents results from the 657 insulin-dependent diabetic participants seen at the baseline examination (1986-1988). The presence of and severity of retinopathy were judged from stereoscopic photographs of three views of the ocular fundus using the modified Airlie House classification system. Fifty-three percent of the participants had background retinopathy, and 31% had proliferative retinopathy. Logistic regression analyses showed that among participants aged less than 18 years, those with background retinopathy were older and had higher levels of glycosylated hemoglobin compared with those without retinopathy. In the 18-29-year age group, participants with background retinopathy had a longer duration of diabetes, higher low density lipoprotein (LDL) cholesterol levels, and lower high density lipoprotein cholesterol levels and were more likely to have microalbuminuria compared with those without retinopathy. Participants aged 18-29 years with proliferative retinopathy had a longer duration of diabetes, higher diastolic blood pressure, and higher fibrinogen and LDL cholesterol levels than those with background retinopathy. In the age group greater than or equal to 30 years, diabetes duration, diastolic blood pressure, and fibrinogen, LDL cholesterol, and triglyceride levels were increased in participants with proliferative retinopathy versus those with background retinopathy. In a multivariate model of proliferative retinopathy, controlling for concurrent renal disease weakened the influence of blood pressure, fibrinogen, triglycerides, and LDL cholesterol and improved the overall fit of the model. These results suggest that diabetic nephropathy may contribute to the development of proliferative (but not background) retinopathy by increasing blood pressure and fibrinogen, by altering the lipoprotein profile, and possibly through other mechanisms.

Proportional hazards models measuring the effect of age at onset of insulin-dependent diabetes mellitus on mortality risk are presented. The study population consisted of 924 insulin-dependent diabetic patients who were seen within 1 year of diagnosis at Children's Hospital of Pittsburgh between 1950 and 1981 and were more than 20 years old at follow-up. Age at diabetes onset was categorized as prepubertal or pubertal, defined by age. Individuals with pubertal onset of diabetes had a significantly higher risk of mortality compared with those with prepubertal onset by diabetes duration but not by attained age. It is proposed that age at onset is an independent determinant of mortality in diabetic individuals and may represent either heterogeneity within insulin-dependent diabetes mellitus with respect to long-term prognosis or an interactive effect of diabetes duration and puberty on prognosis.

The incidence of insulin dependent diabetes mellitus, as determined from the Pittsburgh Insulin Dependent Diabetes Mellitus Registry, was examined in relationship to social class. The registry consists of all Allegheny County, Pennsylvania, cases of insulin dependent diabetes diagnosed under age 20 years in 1965-1976. Analyses of the registry data revealed little relationship between incidence, age at onset and social class. The lack of a strong relationship to social class indicates that the epidemiology of insulin dependent diabetes is not similar to those of polio or Hodgkin's disease, in which social class is strongly related to incidence and age at onset.

Multivariate models for the risk of insulin-dependent diabetes mellitus for siblings of patients with insulin-dependent diabetes were developed using logistic regression analysis. The individuals studied in this report are full siblings of the insulin-dependent diabetic patients diagnosed at Children's Hospital of Pittsburgh between 1964 and 1982. For all siblings considered together, the sharing of two (but not one) HLA haplotypes, the presence of insulin-dependent diabetes in a parent, and being relatively young at the time the proband in the family was diagnosed were significantly associated with increased risk. On the other hand, B7 + siblings had a significantly decreased risk compared to B7--siblings, indicating the presence of an HLA-linked protective gene for the development of the disease. There was a significant interaction between sharing two HLA haplotypes and maternal age at the time of birth; for non-HLA-identical siblings, risk increased with increased maternal age, but maternal age had little or no effect on the risk for HLA-identical siblings. When non-HLA-identical siblings (0 or 1 HLA haplotypes shared) were analyzed separately, only the presence of insulin-dependent diabetes in one of the parents and increased maternal age at the time of birth of the sibling were found to be significantly associated with increased risk. Both of the totally non-HLA-identical diabetic siblings (neither HLA haplotype shared) in this data set had a parent with insulin-dependent diabetes, indicating that at least one HLA haplotype must be inherited in common with an affected family member for diabetes to develop. For HLA-identical siblings, the significant variables were the age of the sibling when the proband was diagnosed, the possession of B7, and maternal age at the time of birth of the sibling, and all three were negatively associated with risk.

Though atherosclerosis may have its origins in childhood, intervention studies on coronary heart disease risk factors have usually begun in older adults. Whether young adults exhibit similar relationships of lifestyle to coronary heart disease risk factors, and therefore might be suitable candidates for appropriate intervention, is poorly understood. The Beaver County Lipid Study is a 9-year follow-up study of a free-living young adult population (mean age 22 years) who were initially screened for hypercholesterolemia as seventh grade school children. This report focuses upon cross-sectional correlates of lipid and lipoprotein concentrations in 561 individuals (males n = 262; females n = 299) in 1981-1982. Body mass index was positively related to low density lipoprotein (LDL) cholesterol and triglycerides in men (r = 0.21; p less than 0.001 and r = 0.41; p less than 0.001) and women (r = 0.16; p less than 0.001 and r = 0.20; p less than 0.001). Cigarette smoking was inversely associated with high density lipoprotein (HDL) cholesterol in men (r = -0.11; p less than 0.001) and women (r = -0.20; p less than 0.001) but positively related to triglycerides in both sexes (r = 0.10; p = 0.05 for men and r = 0.19; p less than 0.01 for women). Alcohol consumption was positively related to HDL cholesterol and triglycerides only among men (r = 0.19; p less than 0.001 and r = 0.12; p less than 0.05, respectively). Educational achievement was also positively related to HDL cholesterol in men (p less than 0.01) and women (p less than 0.001). Multivariate analyses indicate that the sex difference in LDL cholesterol was largely eliminated by controlling for body mass index while significant sex differences in both HDL cholesterol and triglycerides remained after controlling for covariates. Results suggest that the known associations in older adults of body mass index and health-related behavior with lipoproteins are well established by young adulthood. Early intervention particularly for obesity may help ameliorate some of the male excess in cardiovascular disease risk.

Plasma insulin concentrations have been shown to be predictive of future cardiovascular disease in men. Though many clinical studies have documented correlations between plasma insulin and triglyceride concentrations, few epidemiologic studies have reported insulin-lipid correlations. In this report, the authors present correlations obtained from 323 non-diabetic first degree relatives of insulin dependent diabetic patients who underwent 4 hour oral glucose tolerance tests at Children's Hospital, Pittsburgh, Pennsylvania, in February 1980-December 1981 as part of an epidemiologic study of insulin dependent diabetes mellitus. Significant positive correlations were seen between insulin (measured as fasting insulin and the 3 hour area under the insulin curve during the oral glucose tolerance test) and the atherogenic lipids, total and low density lipoprotein cholesterol and triglycerides ranging from r = +0.14 (p less than 0.01) to r = + 0.35 (p less than 0.001). An inverse correlation with high density lipoprotein cholesterol was also noted r = -0.27 (p less than 0.001). A computed score of insulin activity, which the authors call the insulin-glucose sensitivity index, shows equally strong correlations but of reverse sign. In multivariate analyses, these insulin measures and age largely account for the associations of sex and obesity (measured as body mass index) with the atherogenic lipids, though this was only partly true for high density lipoprotein cholesterol. The biologic plausibility of these findings and their relevance to the development of atherosclerosis are discussed.

The clustering of premature mortality was investigated in 1,761 insulin-dependent diabetics and their family members from the Children's Hospital of Pittsburgh Insulin-Dependent Diabetes Mellitus Registry from 1950-1981. At follow-up, 5% of the mothers and 13% of the fathers were deceased. Life table analyses revealed that fathers of deceased diabetics were significantly more likely to die prematurely than fathers of living diabetics (18% vs. 8% at age 55 years; p = 0.02). A father-diabetic son concordance of mortality appeared to be responsible for this effect. A similar overall trend was observed for maternal mortality, although the difference was not statistically significant. Cause-specific analyses revealed that the increased paternal mortality was primarily the result of cardiovascular disease. Overall mortality rates of parents of deceased diabetics were higher than those of the general population, reaching statistical significance in the age group 35-44 years (p less than 0.05). Mortality among diabetic siblings was also examined. Diabetic siblings of deceased diabetics had a markedly increased risk of dying compared with diabetic siblings of living diabetics (p = 0.001). These findings indicate that premature mortality among both diabetic and nondiabetic relatives of diabetics clusters in families in which there is a deceased insulin-dependent diabetic, and suggest that the marked increase in mortality among persons with insulin-dependent diabetes may be partly under familial control.

Epidemiologists studying the risk factors of coronary heart disease, as indicated by a heart attack, often presume that these risk factors combine in a susceptible host to increase the risk of the heart attack. The "biological" basis for the interaction of the key risk factors is often not considered in the analysis. A more rational model for the study of clinical heart disease, stroke, and peripheral vascular disease would be to separate the epidemiology of atherosclerosis from that of the clinical event. In the past, this has been extremely difficult because of the absence of techniques for the measurement of atherosclerosis in vivo, especially in well-defined populations. However, in recent years, greater emphasis on the quantification of atherosclerosis and its relationship to specific risk factors has improved our ability to study the underlying pathology that is atherosclerosis. Atherosclerosis is an example of a common-source epidemic. The environmental agent is the intake of cholesterol and saturated fat. The interaction of specific dietary factors and genetic determinants is the key to the evolution of atherosclerosis. There are marked variations in the evolution of the disease in different vascular beds as well as among individuals exposed to similar environmental factors. In the future, we will probably be able to study atherosclerosis as a continuous variable in populations and to relate the rate of progression, the extent of disease at any point in time, and topical distribution of atherosclerosis within individual vascular beds to specific genetic and environmental determinants.

Racial differences in the incidence and natural history of insulin-dependent diabetes mellitus were evaluated in Allegheny County, Pennsylvania. The yearly incidence rate for whites was approximately 1.5 times that for blacks. The cumulative risk of developing insulin-dependent diabetes prior to age 20 was greater for whites (3.1/1,000, males; 2.8/1,000, females) than for blacks (1.7/1,000, males; 2.0/1,000, females). There was a slight excess risk for white males versus white females (relative risk = 1.09, p = 0.04), while no sex differential in risk was noted for blacks. The temporal trends and seasonal patterns were similar. Blacks more frequently possessed human lymphocyte antigen (HLA)-DR3/X and whites DR4/X; however, the risks associated with these antigens were similar by race. At onset, blacks had evidence of greater immunologic disturbance and less frequently reported infection than whites, although blacks more often had evidence of previous coxsackie viral infections. Evaluation of mortality patterns by race revealed that blacks had a higher mortality rate than whites. By 20 years duration of diabetes, blacks had a 2.4-fold increase in the risk of mortality compared with whites.

The lipid results of a cross-sectional study of cardiovascular disease risk factors in over 600 13-18 year olds drawn from general practices in contrasting areas are reported. Differences were noted between the 3 practices, including a significant variation in the blood group distribution. HDL cholesterol showed a marked variation between the 3 practices which was unlikely to be due solely to seasonal variation. Alcohol consumption though not an univariate correlate of HDL cholesterol was however a strong partial correlate, as were age, height, sexual maturation, adiposity and smoking. Using 6 factors in each sex, 13% of the interchild variance of HDL cholesterol in boys could be explained in a multiple linear regression analysis, and 8% in girls. Including the practice variation as a further independent variable increased the explanation in 18% in boys and 15% in girls. An intraclass correlation (sib-sib) of 0.36 was found for HDL cholesterol, suggesting a strong familial influence. The implications of these results are discussed, along with the difficulties of screening for hyperlipidaemia in this age group.

Higher levels of physical activity have been related to higher concentrations of high density lipoprotein (HDL) cholesterol and lower concentrations of triglycerides. To test the hypothesis that the association between physical activity and the lipoprotein profile is mediated at least in part through increased insulin sensitivity, the authors measured fasting serum levels of HDL cholesterol, triglycerides, insulin, and glucose in 87 men and 83 women (aged 20-24 years) from a population-based survey in Beaver County, Pennsylvania, in 1981-1982. An insulin sensitivity index was calculated as the reciprocal of the insulin and glucose product multiplied by 10,000. Univariate analysis among men indicates that HDL cholesterol was positively related to insulin sensitivity (r = 0.24, p less than 0.05) and to the physical activity score as assessed with Paffenbarger's questionnaire (r = 0.21, p less than 0.05). Insulin sensitivity and physical activity score were positively related (r = 0.14), although not significantly (p = 0.21). Triglycerides were inversely related to both physical activity (r = -0.22, p less than 0.05) and insulin sensitivity (r = -0.19, p = 0.07). No significant findings among women were noted. Multivariate results indicate that the relation between physical activity and the male lipoprotein profile is reduced after controlling for the effects of insulin sensitivity (p greater than 0.10). The authors conclude that in these young men the beneficial effect of physical activity is likely to be partially mediated by increased insulin sensitivity. The lack of findings among women suggests that sex hormones may influence the association between insulin sensitivity and lipoprotein lipids.

The mortality rate from coronary heart disease is much higher among men than women except in diabetes mellitus, which appears to reduce this sex difference. It is hypothesized that the female advantage is due, at least partly, to the more efficient insulin mediated glucose homeostasis in females, an advantage lost in the diabetic state. The authors studied 170 young adult men and women aged 20-24 years from a population-based survey in Beaver County, Pennsylvania, in 1981-1982, in an attempt to further elucidate the sex-specific relationships between fasting serum insulin concentrations and several risk factors. Women who used oral contraceptives and subjects whose fasting serum glucose exceeded 110 mg/dl were excluded. Insulin was related to body mass index in both sexes (r = 0.31; p less than 0.01 for men; r = 0.26, p less than 0.01 for women) and to systolic blood pressure (r = 0.27, p less than 0.01 for men; r = 0.36, p less than 0.001 for women). Insulin was related to diastolic blood pressure in men only (r = 0.31, p less than 0.05). Multivariate analysis revealed fasting serum insulin to be an independent predictor of systolic blood pressure in both sexes and of diastolic blood pressure in men only. Insulin was inversely related to high density lipoprotein cholesterol only among men and this relationship appeared to be largely independent of body mass index and triglycerides. Results indicate that insulin concentration is associated with an adverse coronary heart disease risk factor profile especially among men, consistent with their excess risk of cardiovascular disease.