High levels of homocysteine have been associated with atherosclerosis and cerebrovascular disease. One meta-analysis indicated that even moderately elevated levels of homocysteine are an independent risk factor for cardiovascular disease. Folic acid, pyridoxine, and cobalamin are known to reduce homocysteine levels. Toole and associates, on behalf of the Vitamin Intervention for Stroke Prevention trial, investigated whether adding a multivitamin containing folic acid, pyridoxine, and cobalamin to optimal medical and surgical management of patients with nondisabling cerebral infarction and elevated homocysteine levels could reduce the incidence of recurrent cerebral infarction, as well as coronary heart disease (CHD) and death.

Patients with a presumptive diagnosis of stroke with elevated homocysteine levels were included in this multicenter, randomized, double-blind, controlled trial. Of the 3,680 eligible adults, 1,827 were randomized to a high-dose vitamin group (25 mg pyridoxine, 0.4 mg cobalamin, 2.5 mg folic acid), while 1,853 patients were randomized to a low-dose vitamin group (200 mcg pyridoxine, 6 mcg cobalamin, 20 mcg folic acid). Primary end points included recurrent stroke, increased score on the National Institutes of Health Stroke Scale, and death. Other end points were myocardial infarction requiring hospitalization, coronary revascularization, cardiac resuscitation, and fatal CHD.

The study was terminated after all participants had at least one year of follow-up, because the investigators determined that additional follow-up would be unlikely to show a difference between groups. Mean plasma homocysteine levels were virtually identical in the groups at randomization, and both experienced a subsequent drop in total homocysteine levels; however, the difference was greater in the high-dose vitamin group, ranging from 2.0 μmol per L at one month to 2.3 μmol per L at two years. Among patients in the low-dose vitamin group, 8.1 percent had a recurrent ischemic stroke compared with 8.4 percent of patients in the high-dose group. The risk ratio for ischemic stroke between the two groups was 1.0 at two years. A risk ratio of 0.9 was found between the groups for CHD events and death at two years.

Although no treatment effect was found between the groups, a graded association occurred between baseline homocysteine levels and subsequent homocysteine levels. In the low-dose group, a 3 μmol per L lower total homocysteine level was associated with a 10 percent lower risk of stroke, a 26 percent lower risk of CHD events, and a 16 percent lower risk of death. Decreases in risk in the high-dose group were 2 percent for stroke, 7 percent for CHD events, and 7 percent for death, but these figures were not significant.

In this trial, vitamin therapy had no effect on stroke, CHD events, or death. Homocysteine may be a marker for vascular disease risk rather than its cause. It also is possible that the use of grain fortified with folate in the United States attenuated the difference in the amount of folate the two study groups consumed. Finally, reduced homocysteine levels may be beneficial only in patients with higher homocysteine levels than the patients included in this study.