DIABETES, MORTALITY AND CORONARY HEART-DISEASE IN THE PROSPECTIVE DUBBO STUDY OF AUSTRALIAN ELDERLY

Background: A prospective study of Australian elderly living in Dubbo has shown that diabetes is a significant predictor of all-causes morta lity and coronary heart disease (CHD). Aim: To examine and contrast cl inical and socio-demographic predictors of these outcomes in those wit h and without diabetes. Methods: The data are derived from a community -based sample of subjects 60 years and older followed over 62 months s ince 1988. Of 1155 men and 1472 women, 9.2% and 6.9% respectively mani fested diabetes at baseline, based on history or fasting hyperglycaemi a. Results: In the presence of diabetes, all-causes mortality was incr eased twofold in both sexes, CHD incidence was increased twofold in me n and threefold in women, stroke incidence was increased twofold in wo men but little changed in men. Proportional hazards models were derive d separately for persons with and without diabetes and risk factors di fferentially predictive in diabetes were sought. Significant predictor s of death in diabetes were old age and current smoking. Those factors differentially predictive were 'being married' (Relative Risk [RR] 1. 60 with diabetes and 0.69 without diabetes) and higher body mass index (BMI) (RR 1.03 with diabetes and 0.79 without diabetes). Significant predictors of CHD in diabetes were old age, prior CHD, severe hyperten sion, low HDL cholesterol and self-rated health. Those factors differe ntially predictive were higher body mass index (RR 1.14 vs 0.83) and p hysical disability (RR 0.69 vs 1.55). Differential predictions with re gard to BMI may relate in part to excess CHD and mortality at low BMI in non-diabetic subjects. Conclusion: The vascular disease burden of d iabetes in the elderly has been confirmed, especially in women. A numb er of conventional risk factors are contributing to this burden and ma y be amenable to treatment.