Plasma Lp(a), apolipoprotein(a) isoforms and acute myocardial infarction in men and women: a case-control study in the Jerusalem population

The relationship of Lp(a) with manifestations of coronary heart disease (CHD) has not been studied extensively in women. There is little information as to the association of the unique Lp(a) apolipoprotein moiety (apo(a)) with CHD in either men or women. We therefore assessed the association of the apo(a) polymorphism and of Lp(a) with first acute myocardial infarction (MI) in a population-based case-control study in Jewish residents of Jerusalem between the ages of 25 and 64. The patients consisted of 238 men and 47 women hospitalized for a first acute MI in the 4 hospitals of Jerusalem serving the population (70% response rate among all first MI patients). The control subjects comprised 318 men and 159 women sampled from the national population registry and who were free of CHD (75% response). Lp(a) and apo(a) were measured in plasma stored at -20 degrees C for 6-24 months. Among men, plasma Lp(a) concentrations were higher in cases than controls in both univariate and multivariate analyses. The elevated risk was limited to the upper fifth of the Lp(a) distribution (unadjusted odds ratio = 1.65, P0.01 vs. the lower four quintiles, multivariable odds ratio = 1.82, P0.01). Among women, Lp(a) was not elevated in acute MI patients. Apo(a) isoforms with a B, S1 or S2 band (associated with higher Lp(a) values and having lower molecular weights) were more prevalent in female MI cases than controls (unadjusted odds ratio = 2.5, P = 0.016). This association could not be attributed to the higher Lp(a) concentrations associated with these isoforms and was not seen in men. In conclusion, our study points to an association of the apo(a) isoforms with acute MI in women, not evident in this population sample in men. Previously described associations of elevated Lp(a) with acute MI were confirmed in men but not in women. While the role of chance and inadequate statistical power cannot be excluded, the suggestion of a sex difference in the strength of these associations deserves further investigation, as does the question of whether apo(a) phenotype contributes to risk independently of Lp(a) level.