Analysis of genetic and environmental sources of variation in serum cholesterol in Tecumseh, Michigan. I. Analysis of the frequency distribution for evidence of a genetic polymorphism

Abstract

Analyses of serum cholesterol measurements on 4,619 males and 4,730 females residing in the community of Tecumseh, Michigan, were conducted to estimate the contribution of sex, age, temporal variation, and bimodality to determining the normal variation among individuals sampled without regard to their health status. Female values had a higher mean (2.8 mg/100 ml greater) but smaller variance than males when adjusted by polynomial regression to a common age. Positive skew in the frequency distribution for both sexes was removed by natural logarithm (ln) transformation. Age variation accounted for 28.5% and 29.4% of the variance in a ln cholesterol measurement of males and females, respectively. Between 7% and 10% of the variance in a ln cholesterol value was estimated to be attributable to differences between age-adjusted replicate measurements of the same individual. The reduction in individual variability by adjustment for these contributions to variance will allow a more precise evaluation of the relative contribution of alternate genetic hypotheses as explanations for normal variation in cholesterol. Assuming bimodality, approximately one in 1,000 males and one in 1,000 females belong to a second mode of hypercholesterolemic individuals. The locus determining familial hypercholesterolemia is not a major source of normal phenotypic variation in the Tecumseh population.