The Columbia population study. I. Plasma cholesterol and triglyceride levels

The distributions of plasma cholesterol and triglyceride levels, and the influence of age, sex and race on these lipids were studied in the Columbia Medical Plan, a prepaid comprehensive health care plan. Of 23,600 members, 1,553 were randomly selected; 3,522 household members were also eligible. Of 4,382 participants (86.3% response rate), the plasma lipids were analyzed on 3,559 fasting, nonpregnant subjects. Their mean age was 23 years (range, 6 weeks to 92 years); 87% were white and 13% nonwhite. Logarithmic transformation reduced some but not all of the positive skewness and leptokurtosis in the lipid distributions. The rank order of cholesterol levels in the first two decades was nonwhite females > nonwhite males > white females > white males, while that of triglyceride levels was white females > white males > nonwhite females > nonwhite males. In the second decade the cholesterol level fell initially and then increased, a concave pattern most marked in the male. The triglyceride level began to increase appreciably at about 10 years of age. In adulthood, a crossover occurred, and males developed considerably higher lipid levels than females in both the white (third decade) and the nonwhite (fourth decade) groups; these differences abated in the fifth decade. The 95th percentiles for triglycerides were markedly higher (range, 42 to 268 mg/100 ml) in adult males than in females. Arbitrary 'cutpoints' for hyperlipidemia (95th percentiles) in Columbia were lower for cholesterol (range, 16 to 62 mg/100 ml) but higher for triglyceride (range, 19 to 133 mg/100 ml) than those previously developed at the National Heart, Lung and Blood Institute. We conclude that age, race and sex differences in lipid levels become most marked in adult life and appear to begin during the second decade. Values commonly used for the 95th percentiles should be reevaluated.