To the Editor.
—We are concerned that busy practitioners reading the cholesterol analysis by Dr Krumholz and colleagues1 and the accompanying Editorial2 may be left with the impression that lipoprotein lipids are not risk factors for older people. A comprehensive review of the literature does not support this conclusion, particularly when heterogeneity of the older population is considered.Level of cholesterol in older persons reflects health status as well as diet and genetic factors, such that lower cholesterol in old age is paradoxically associated with poorer health status. Therefore, studies of healthier older persons have tended to support the cholesterol hypothesis, while studies of frailer or more mixed populations have not. When change in cholesterol was considered directly, those with the greatest declines in cholesterol had increased risk of both cardiovascular and all-cause mortality.3 These findings may account for the lack of association or the inverse association
Harris T, Havlik R, Ettinger WH. Cholesterol and Coronary Heart Disease Risk in Elderly Patients. JAMA. 1995;273(17):1329. doi:10.1001/jama.1995.03520410023008