Background: The purpose of the present study was to study the effects of age, weight change, and covariates on lipid and lipoprotein levels cross-sectionally and prospectively in an elderly population.
Methods and results: A community-based sample of 1041 men and 1303 women aged 50 to 93 years was studied cross-sectionally in 1984 to 1987, with follow-up of 372 men and 545 women 8 years later. In the cross-sectional study, levels of total cholesterol (TC) and LDL cholesterol (LDL-C) decreased and levels of HDL cholesterol (HDLC) increased with age in men (all P < .001) but not in women. In the prospective study, TC, LDL-C, and HDL-C levels all decreased in both men and women, in all age groups (50 to 64 years, 65 to 74 years, and > or = 75 years) and in all weight change groups (> 2.5-kg loss, change within 2.5 kg, and > 2.5-kg gain) and in all waist girth change groups, for an overall decrement of approximately 1% per year. In multiple linear regression models, change in weight was the most important independent and consistent predictor of changes in TC, LDL-C, and HDL-C. Similar results were obtained in analyses excluding subjects taking lipid-lowering drugs or estrogen and in analyses adjusted for changes in cigarette smoking, alcohol intake, physical activity, medication use, and incident myocardial infarction, cancer, or diabetes.
Conclusions: Cross-sectional decrements in TC and LDL-C with age in men are not explained by survivor bias because they are also observed prospectively. Although weight change was the most important explanatory variable, TC, LDL-C, and HDL-C levels also decreased in those who lost or gained weight. Age was not an independent predictor of change. Other prospective studies are recommended to better define the causes and consequences of cholesterol and lipoprotein changes in old age.