Sex differentials in predictors of mortality for patients with adult-onset diabetes: a population-based follow-up study in Beer-Sheva, Israel.

OBJECTIVE To test the hypothesis that factors predicting mortality differ between diabetic men and women. RESEARCH DESIGN AND METHODS A total of 498 known patients with diabetes residing in a well-defined geographical area and receiving primary health care in 3 primary care community clinics were interviewed and examined between 1988 and 1990. RESULTS By 31 July 1998, after a mean follow-up period of 7.8 years, 148 patients (68 men and 80 women) had died (29.7%). No statistical differences in survival rate or in the specific causes of death were found between men and women. In the univariate analysis of factors examined at baseline, GHb levels were significantly higher among women who died compared with women who survived, but this was not the case for men. Conversely, a trend of higher triglyceride and uric acid levels was found for men who died compared with men who survived, but this was not the case for women. Multivariate Poisson regression analysis showed significantly higher risk ratios for mortality in men > or = 63 years of age, men with microalbumin excretion > or = 30 mg/l, and men with higher triglyceride levels. In contrast, the analysis in women showed that higher GHb and creatinine levels and a reported history of heart disease were the only factors at the baseline examination significantly and independently associated with an increased risk ratio of mortality. CONCLUSIONS The results suggest the existence of sex-specific interactions with various metabolic factors associated with diabetes that may have a different effect on mortality for each sex.