Cardiovascular Disease, Neuropathy, and Retinopathy

Role of glycemia in cardiovascular disease Jacqueline Dekker (Amsterdam, the Netherlands) discussed the association between hyperglycemia and cardiovascular disease (CVD), particularly addressing A1C measurement in nondiabetic individuals. A1C, she noted, is considered a gold-standard measure of chronic glycemia in diabetic patients. In nondiabetic populations, its implications are somewhat less clear. In a study of 648 apparently healthy individuals, 12% had an elevated total A1 level (measured in that study) not explained by measurement error or glucose intolerance and remaining in the same range over 3.5 years in 90% of those initially with high and in 68% of those initially with low baseline levels (1). There was no correlation with glucose tolerance, with caloric intake, or with physical activity, but A1 level was associated with cigarette use and with clinically overt atherosclerosis, leading Dekker to conclude that “factors unrelated to glucose metabolism are the main determinants of A1” in nondiabetic individuals, perhaps with bearing on risk of what are considered complications of diabetes. A subsequent study of 3,240 nondiabetic individuals showed weak correlation of A1C with fasting and 2-h glucose, cholesterol, and insulin. Weak but significant correlations were also found with erythrocyte characteristics, and cigarette use was associated with higher A1C, again suggesting nonglycemic effects on hemoglobin glycation (2). In the Islington Diabetes Survey, 1,084 participants had glucose tolerance tests and four different A1C assays. The 2-h postload glucose explained 19–41% of the variation in A1C, depending on the assay, and 26–48% of the variation was explained by the fasting glucose, with low and high glycators remaining constant over 4 years (3). If 50– 80% is not explained by glycemia, the implication is, again, that although important in understanding diabetes, A1C must be considered cautiously with regard to implications of mechanistic causes of complications of hyperglycemia. In a study of five individuals with low and seven with high A1C, relative to the 2-h glucose, there was a stronger relationship between glycated albumin and blood glucose and no differences in insulin, urea, free fatty acids, vitamin C, or intraerythrocyte organic phosphate. Higher erythrocyte 2,3 diphosphoglycerate levels and higher intracellular pH were found in the higher glycators (4). Dekker showed further evidence from a study of identical twins that A1C levels are genetically determined even in type 1 diabetes, explaining 62% of the variation in A1C (5). Potential glucose-independent mechanisms include hemoglobin glycation, erythrocyte cell membrane permeability to glucose, erythrocyte survival, and intraerythrocyte pH. In the Diabetes Prevention Program, among individuals with impaired glucose tolerance there were ethnic group differences in A1C not explained by glycemic differences, with blacks having mean A1C levels 0.5% greater than whites without a difference in glucose tolerance (6). The relationship between hyperglycemia and mortality was best shown in the European DECODE study, with a stepwise relationship between 2-h glucose and mortality from low levels to impaired glucose tolerance to type 2 diabetes, with the association less convincing for fasting glucose (7). A meta-analysis of studies measuring fasting and 2-h glucose and A1C suggested that risk was related to all indicators, with a stronger relationship for women than for men; after adjustment for cardiovascular risk factors, the highest glycemia group had 19% greater risk than the lowest (8). In the Rancho Bernardo study, A1C was a better CVD predictor than fasting or 2-h glucose (9); in the Framingham Offspring study CVD risk increased 15% per SD increase in A1C, independent of fasting glucose and 2-h glucose (10); in the Epic-Norfolk study, there was a 1.2to 1.3-fold increase in cardiovascular risk for each 1% increase in A1C (11); and in the Atherosclerosis Risk in Communities study, A1C was associated with both stroke and coronary heart disease risk (12). However, in the Women’s Health Study myocardial infarction, stroke, and revascularization were not associated with A1C after adjustment for risk factors (13). In Dekker’s studies in Hoorn in the Netherlands, baseline examination of 2,848 men and women (aged 50 –75 years) between 1989 and 1990 included glucose tolerance testing, with the correlation coefficients of A1C vs. fasting and 2-h glucose 0.24 and 0.14, respectively, in individuals with normal glucose tolerance, with 16% of this group having A1C 6%. Among 1,675 nondiabetic individuals with A1C measurement, 10 year follow-up showed strong associations of A1C both with fatal and with nonfatal CVD, with 1.7and 3.0-fold increases in risk per 1% increase in A1C among men and women, respectively, while neither fasting nor 2-h glucose showed significant effect when other CVD risk factors were taken into account, leading Dekker to speculate that the continuous relationship between A1C and CVD is only to a limited extent due to glycemia. A1C might rather reflect an individual’s vulnerability to hyperglycemia. She suggested that glucose tolerance testing remains important to determine an individual’s status, with lifestyle intervention particularly effective among individuals with higher diabetes risk (14). She noted further that glucose is a direct measure of diabetes, while A1C is indirect and the time course of the relationship between increase in A1C and glycemia is not known. Furthermore, there is no information as to effectiveness of interventions ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●