Long-Term Glycemic Control Relates to Mortality in Type II Diabetes

Abstract

To study the influence of long-term glycemic control on mortality in a cohort of newly detected type II diabetic individuals. A total of 411 newly detected type II diabetic individuals diagnosed between 1972 and 1987 were followed until 31 December 1989. Clinical data concerning fasting blood glucose (FBG) values, body mass index (BMI), type of treatment, and concomitant diseases were collected during 8 randomly selected years. Long-term glycemic control was measured as annual averages of FBG values during these years. Mortality data were obtained from official registers. There were 161 diabetic subjects who died during a mean follow-up time of 7.4 years. In univariate analyses, higher age at diagnosis; higher baseline or average FBG; and the presence of heart disease, cerebrovascular disease, or kidney disease at the beginning or during the course of diabetes were related to higher mortality. Type of diabetes treatment or having a diagnosis of hypertension was not related to mortality. In multiple logistic regression analyses, age at diagnosis and average FBG were independently related to all-cause (P = 0.0002), cardiovascular (P = 0.0006), and ischemic heart disease mortality (P = 0.03). No correlations between glycemic control and noncardiovascular deaths were found. In a Cox's regression analysis, average FBG was significantly related to length of survival when age at diabetes diagnosis was taken into account (P < 0.05). Diabetic subjects with average FBG > or = 7.8 mmol/l had 50% higher mortality compared with diabetic subjects with average FBG < 7.8 mmol/l. Diabetic patients with good long-term glycemic control had a better survival rate than subjects with high average FBG values. The findings should be interpreted cautiously because possible confounding factors such as dyslipoproteinemia and smoking were not studied.