Effect of glycemia on mortality in Pima Indians with type 2 diabetes.

The effect of plasma glucose concentration on overall and cause-specific mortality was examined in 1,745 Pima Indians (725 men, 1,020 women) > or = 15 years old with type 2 diabetes. During a median follow-up of 10.6 years (range 0.1-24.8), 533 subjects (275 men, 258 women) died; 113 of the deaths were attributable to cardiovascular disease, 96 to diabetes-related diseases (diabetic nephropathy for 92 of these), 249 to other natural causes, and 75 to external causes. After adjusting for age, sex, duration of diabetes, and BMI in a generalized additive proportional hazards model, higher baseline 2-h postload plasma glucose concentration predicted deaths from cardiovascular disease (P = 0.007) and diabetes-related diseases (P = 0.003), but not from other natural causes (P = 0.73). An increment of 5.6 mmol/l (100 mg/dl) in the 2-h plasma glucose concentration was associated with 1.2 times (95% CI 1.1-1.4) the death rate from cardiovascular disease, 1.3 times (95% CI 1.1-1.5) the death rate from diabetes-related diseases, and almost no change in the death rate from other natural causes (rate ratio = 1.0; 95% CI 0.94-1.1). In Pima Indians with type 2 diabetes, higher plasma glucose concentration predicts deaths from cardiovascular and diabetes-related diseases but has little or no effect on deaths from other natural or external causes.