The effects of generation and gender on the joint distributions of lipid and apolipoprotein phenotypes in the population at large

Abstract

The generation and gender effects on the joint distributions of total plasma cholesterol (Total-C), 1n triglycerides (1nTrig), HDL-cholesterol (HDL-C), LDL-cholesterol (LDL-C), apolipoproteins AI (Apo AI), AII (Apo AII), and E (1nApo E) were studied in 184 male grandparents (MGP), 242 female grandparents (FGP), 237 male parents (MP), 235 female parents (FP), 202 male children (MC), and 200 female children (FC). Homogeneity of variance tests revealed that lipid variances were gender and/or generation specific while apolipoprotein variances were homogeneous across strata. In the absence of heterogeneity of variance, significant heterogeneity in LDL:1nTrig and 1nTrig:Apo AII covariation was found between genders in the parental generation. In the presence of heterogeneity of variance, significant heterogeneity of correlation between genders and/or across generations was found for the HDL-C: LDL-C, Total-C:LDL-C, Total-C: 1nTrig, 1nTrig:LDL-C, Total-C: 1nApo E and HDL-C:1nApo E bivariate distributions. Analyses of principal components revealed that the generation and gender specific cohorts have similar eigenvalues but distinct eigenvectors for the first two principal components underlying the seven dimensional lipid and apolipoprotein distribution. We conclude that the amount of variability explained by the first two principal components is the same across cohorts but how the interindividual variability is distributed among the lipid and apolipoprotein traits is generation and gender specific. This study documents the role that variance and covariance might play in determining risk of disease for special subgroups of the population at large. It also demonstrates how variances and covariances between risk factors traits characterize life processes of aging and sexual dimorphism. This study argues that future biometrical genetic and epidemiological studies of coronary artery disease must take into account age and gender effects on interindividual variability and covariability of risk factors.