Aims/hypothesis: Prevalence and incidence of coronary heart disease (CHD) are increased in patients with Type II (non-insulin-dependent) diabetes mellitus; whether this is entirely due to more extensive coronary atherosclerosis is, however, controversial.
Methods: We analysed the clinical, angiographic and follow-up data of 2253 consecutive patients undergoing coronary angiography over the decade 1983-1992.
Results: Abnormal coronary arteries (> or =50% stenosis) were found more frequently in diabetic than in non-diabetic subjects (85 vs 67%, p < 0.0001), the excess being explained by a higher prevalence of three-vessel disease (36 vs 17%, p < 0.0001). The sum of all angiographically detectable lumen stenoses (atherosclerosis score, ATS) was higher in diabetic than in non-diabetic subjects (352 +/- 232 vs 211 +/- 201 units, p < 0.0001). After adjusting for measured cardiovascular risk factors, diabetes was still associated with an excess ATS (114 units in men and 187 units in women, p < 0.0001 for both, p < 0.03 for the interaction ATS x sex). Within the diabetic group, the only variable that was independently (of sex and age) associated with ATS was serum cholesterol, whereas plasma glucose concentration, disease duration and type of treatment were not correlated with the severity of coronary atherosclerosis. In contrast, clinical grade proteinuria was not associated with a more diffuse coronary atherosclerosis either in diabetic (366 +/- 243 vs 354 +/- 233 units) or non-diabetic subjects (231 +/- 201 vs 207 +/- 197 units). Over a mean follow-up period of 88 months, 19% of diabetic patients compared with 10% of non-diabetic patients died of a cardiac cause (age and sex-adjusted odds ratio OR = 1.34 [1.14-1.57]). In a Cox model adjusting for age, sex and all major risk factors, diabetes was still associated with a significant excess risk of dying of a cardiac cause (OR = 1.37 [1.14-1.60]); this excess was similar to, and independent of, that carried by the presence of prior myocardial infarction in the whole population (OR = 1.42 [1.25-1.62]). Proteinuria was associated with a higher risk of cardiac death, particularly in diabetic patients, independently of coronary atherosclerosis (adjusted OR = 1.46 [1.03-1.99]).
Conclusion/interpretation: In patients undergoing angiography, diabetes, especially in women, is associated with more severe and diffuse coronary atherosclerosis which is not explained by either the traditional risk factors or the presence of proteinuria. On follow-up, these patients experience an excess of cardiac deaths, to which coronary atherosclerosis and proteinuria make independent, quantitative contributions.