Background Background Homocysteine (HCY) has been strongly associated with cardiovascular disease. In contrast to retrospective studies prospective trials have been inconsistent in the power of risk prediction. We aimed at investigating whether gender differences in homocysteine associated cardiovascular risk exist.

Methods and Results We obtained baseline samples of plasma homocysteine from a prospective cohort of 1458 consecutive patients with coronary artery disease (CAD). 468 (28%) were of female gender. Cardiovascular events (non-fatal MI and cardiovascular death) (N=272) were registered during a follow-up of 5.9 ± 1.3 years. HCY concentrations were higher in men in comparison with women (12.6 (10.4/15.4) vs. 11.6 (10.4/14.9)); P=0.03). The event rate was similar in both genders (P_{chi square}: 0.63). In both subgroups patients with future cardiovascular events presented with increased levels of HCY at enrolment (14.7 (11.7/19.4) vs. 12.3 (9.8/15.6); P=0.006 in women; 14.7 (11.9/19.3) vs. 12.8 (10.6/15.6); P<0.0001 in men). Kaplan-Meier curves revealed a stepwise increase of cardiovascular risk in men whereas in the female population a clear risk increase for individuals with HCY concentrations above the median was seen. HCY as a continuous variable was predictive for future cardiovascular events. However, after adjustment for various potential confounders like classical risk factors, number of diseased vessels, cardiac medication and creatinine, homocysteine soon lost its predictive power in women whereas it remained a powerful risk predictor in males (P<0.0001). In a subgroup of 608 patients we measured the anti-oxidative glutathione-peroxidase (GPx)-1. Enzyme activity was significantly higher in women (51.3 ± 11.7 U/gHb vs. 48.6 ± 11.6 U/gHb in men; P=0.01).

After adjustment for GPx-1 homocysteine lost its predictive power in the female subgroup (P=0.076) whereas it remained independently predictive in males (P<0.0001).