Gami and colleagues correctly note that we measured plasma glucose following a 50-g oral glucose load, as described else-where (1). We also evaluated serum cholesterol, blood pressure, cig-arette smoking, weight, height, electrocardiographic data, education, and ethnicity. The primary goal of our analysis was to determine whether traditional CHD risk factors differed between young and middle-aged men in follow-up for long-term mortality. To answer Gami and colleagues' question concerning asymptomatic glycemia and long-term CHD mortality, we included plasma glucose level (in increments of 1.11 mmol/L [20 mg/dL]) in the multivariate models we originally reported (Table). After adjustment for major covari-ates, plasma glucose level and 20-year CHD mortality had a border-line relationship in middle-aged men (relative risk, 1.03 [95% CI, 1.00 to 1.06]) and were not significantly related in men who were 18 to 39 years of age at baseline. However, no significant difference was seen in CHD risk between the two age groups because of overlap-ping confidence intervals (P > 0.2).

A previous report from this cohort (1) examined black and white men who were 35 to 64 years of age at baseline. Elevated postload glucose level was significantly related to total death from cardiovascular disease in 22-year follow-up in this more heteroge-neous, and mostly older, group of men. These results were among those compiled for the meta-regression analysis (2) mentioned by Gami and colleagues.

Long-term prediction of CHD based on single glucose measure-ments in asymptomatic nondiabetic patients has produced inconsis-tent results, both within and between studies (2). However, our primary purpose was to determine and raise awareness of the long-term consequence of major CHD risk factors that are already present in young adulthood. Perhaps the greatest hope for elimination of CHD as an epidemic in western societies is prevention of the risk factors themselves. Data on asymptomatic glycemia and CHD risk are inconsistent. However, prevention of high cholesterol levels, high blood pressure, cigarette smoking, and overweight involves attention to healthy behaviors and to regular exercise—the same measures that can prevent glucose intolerance and diabetes (3). Gami and colleagues are therefore correct in drawing attention to factors associated with asymptomatic glycemia in the long-term prevention of CHD.