Lp(a) lipoprotein [Lp(a)] was found in previous studies to be independently associated with early atherosclerosis and its sequelae. Lp(a) in vitro bound to glucosaminoglycans and was easily aggregated at physiological Ca2+ concentration, and small Lp(a) aggregates were phagocytosed by macrophages. Lp(a) was also found to be related to carbohydrate metabolism, and increased Lp(a) levels have been described in diabetic patients with clinical complications and were recently found in rheumatoid arthritis patients. In this study of nondiabetic male patients with documented CAD before 50 years of age and controls, a significant correlation was found between Lp(a) and IGF-1 levels. HLA class II DR13 (DR6) was more frequent and DR15 (DR2) was less frequent in patients than in controls. The calculated relative risk for CAD was 4.0 for DR17 (DR3), but the difference was not significant. These differences seem to be related to high Lp(a) levels. It is suggested that phagocytosis of preferably Lp(a) aggregates can induce an immunological tissue response that may contribute in the pathogenesis of Lp(a)-associated diseases and may be more prominent in combination with some inherited HLA class II haplotypes. Probably due to sex hormone effects, the association may be most pronounced in young males and in older females.