Fasting plasma glucose in non-diabetic elderly women predicts increased all-causes mortality and coronary heart disease risk

Aim: To confirm the hypothesis that upper normal plasma glucose levels in non-diabetic subjects are independently predictive of mortality and cardiovascular disease (CVD). Methods: The study reports on 113 months' follow-up in a prospective study of CVD in the Australian elderly, The Dubbo Study. The cohort, first examined in 1988-89, consisted of 2805 men and women 60 years and older. Of the cohort, 2419 (86%) were defined as non-diabetic. The prediction of outcomes by quartile of fasting plasma glucose was examined in a Cox proportional hazards model, after linkage to hospital and death records. Results: All-causes mortality increased progressively across quartile of fasting plasma glucose in both sexes, reaching statistical significance only in women. Coronary heart disease (CHD) incidence increased similarly, the increases being proportionately greater in women. Ischaemic stroke did not show a consistent gradient with fasting plasma glucose. After adjustment for age and other risk factors, all-causes mortality, CHD and ischaemic stroke incidence were not significantly related to plasma glucose in men. In women, all-causes mortality and CHD incidence showed a significant gradient with glucose quartile. Hazard Ratio (95% confidence intervals) for death in glucose Quartile IV(5.3-6.0 mmol/L) was 1.49 (1.03-2.14) and for CHD incidence was 1.52 (1.08-2.15). Subjects in the upper quartiles of fasting plasma glucose showed a clustering of overweight, hypertension, elevated serum triglycerides, reduced high density lipoprotein cholesterol and excess of small dense low density lipoprotein, suggestive of the Insulin Resistance Syndrome. Conclusion: Fasting plasma glucose levels in the upper normal range in non-diabetic elderly subjects appear to be associated with increased all-causes mortality and CHD, especially in women.