Prediction of cardiovascular disease among the elderly is complicated by the fact that traditional risk factors are not as meaningful in this age group. Basic research has indicated that Lp(a) lipoprotein is found in atherosclerotic plaques and plays a role in thrombotic complications at these plaque sites. Prospective studies of Lp(a) levels as a predictor of cardiovascular disease in middle-aged populations have been inconclusive. Ariyo and colleagues determined the ability of Lp(a) levels to identify elderly persons at risk for cardiovascular complications.

Data for the analysis were drawn from the Cardiovascular Health Study sponsored by the National Heart, Lung, and Blood Institute. This prospective trial assessed 5,888 persons 65 years and older for cardiovascular disease risk factors and followed them to determine the incidence of cardiovascular events. Lp(a) level was included among the laboratory parameters assessed at study entry in the initial cohort of 5,201 mostly white participants, but the measurement had been dropped from the study by the time a later cohort of participants from minority population groups was added. After excluding participants with established heart disease, those taking lipid-lowering medications, and a few persons with severely elevated Lp(a) levels, 3,972 subjects were analyzed for an association between Lp(a) level and cardiovascular disease.

No association was found between Lp(a) levels and cardiovascular complications in women. Among men, development of coronary heart disease was not significantly associated with Lp(a) levels. For stroke, however, men with Lp(a) levels in the highest quintile (8.2 mg per dL or higher) had triple the unadjusted risk of those in the lowest quintile (1.2 mg per dL or lower). Compared with men in the lowest quintile, death from all causes was approximately double in men in the highest quintile of Lp(a) values. Statistical adjustment for age, lipids, smoking, diabetes, and other traditional risk factors did not alter the increased risk for stroke and all-cause mortality.

The authors conclude that elevated Lp(a) lipoprotein level is an independent predictor of stroke and all-cause mortality in elderly men, but not in elderly women.

editor's note: Lp(a) lipoprotein has been bandied about for decades as a possible risk factor for cardiovascular disease, yet a number of large studies have yielded inconsistent results. This analysis appears to offer little that would change these findings. Lp(a) did not predict anything in women, was not studied in minority groups, and did not predict heart disease in either sex. Even the limited association with stroke and all-cause mortality appears to be suspect. There was no stepwise elevation in risk with each ascending quintile of Lp(a); only the highest quintile had a significant risk association. There are other non-traditional risk factors (e.g., Creactive protein) that have a more broad-based applicability (men and women, middle-aged and elderly) than Lp(a). Because there is no treatment that specifically targets Lp(a), control of traditional risk factors remains the focus of therapy even when patients with elevated Lp(a) levels are identified.—B.Z.