were not always consistent across the studies, and negative findings have also been reported (Bower and Stanley, 1992a; Hayes et al., 1996; Scanlon et al., 1998). Because multivitamins were used in all these studies, it is difficult to disentangle the effect of folate from that of other constituents. Also, the presence of unmeasured confounding cannot be excluded.

The link between homocysteine and the risk of vascular disease was derived from the study of homocystinuria. Classical homocystinuria is a rare autosomal recessive disease caused by a deficiency of cystathionine β-synthase and characterized by excessively elevated plasma homocysteine (Mudd et al., 1985). Clinical manifestations include mental retardation, skeletal abnormalities, lens dislocation, and a marked tendency to develop premature and severe atherosclerosis with thromboembolic events. In 1976 a study first showed a significant difference in homocysteine plasma concentration between patients with vascular disease and normal control subjects (Wilcken and Wilcken, 1976).

Since then, many observational and experimental studies have been published on the risk of vascular disease associated with elevated homocysteine levels. In 1995 Boushey and coworkers first published a meta-analysis; this work has recently been updated and includes a total of 20 studies (Beresford and Boushey, 1997). The relative increase in risk of coronary heart disease (CHD) as estimated by the combined odds ratio was 1.6 (95 percent CI, 1.5 to 1.7) for men and 1.5 (95 percent CI, 1.3 to 1.7) for women for each increment of 5 mmol/L in total plasma or serum homocysteine. The combined odds ratio for data from both men and women was 1.8 (95 percent CI, 1.6 to 2.0) for cerebrovascular disease and 2.0 (95 percent CI, 1.5 to 2.6) for peripheral vascular disease. There is evidence that hyperhomocysteinemia is a risk factor for CHD independent of other known risk factors such as smoking, cholesterol, body mass index, age, high blood pressure, and diabetes (Beresford and Boushey, 1997; Graham et al., 1997; Mayer et al., 1996; Verhoef et al., 1996). Similarly, Nygård and colleagues (1997) reported that plasma homocysteine values were a strong predictor of mortality in patients with angiographically confirmed coronary artery disease.

The mechanism by which elevated homocysteine might increase the risk of developing vascular disease is unclear. Several hypotheses have been proposed, and the subject was reviewed by Mayer et al. (1996). Homocysteine can exert a direct toxic effect on endothelial