Design.-Cross-sectional analysis of homocysteine levels and vitamin blood levels and intake in elderly participants in the Framingham Study.

Setting.-Population-based cohort in Framingham, Mass.

Participants.-A total of 1160 adult survivors, aged 67 to 96 years, from the original Framingham Heart Study cohort.

Main Outcome Measures.-Plasma homocysteine concentration correlated with plasma folate, vitamin B-12, pyridoxal-5'-phosphate (PLP), and oral intakes of these vitamins, and the contribution of these vitamins to the prevalence of elevated homocysteine in the population.

Results.-Homocysteine levels were positively correlated with age after controlling for vitamin concentrations. After controlling for age, sex, and levels of other vitamins, homocysteine exhibited a strong inverse association with plasma folate. When subjects were grouped by deciles of plasma folate, mean homocysteine was significantly higher in the lowest two folate deciles (1 5.6 and 1 3.7 mumol/L, respectively) than in the highest decile (11.0 mumol/L). Homocysteine demonstrated weaker, inverse associations with plasma vitamin B-12 and PLP. Similar inverse associations were demonstrated between homocysteine and intakes of folate and vitamin B-6, but not vitamin B-12. Prevalence of high homocysteine (> 14 mumol/L) was 29.3% in this cohort, and was greatest among subjects with low folate status. Inadequate plasma concentrations of one or more B vitamins appear to contribute to 67% of the cases of high homocysteine.

Conclusions.-These results indicate a strong association between homocysteine concentration and folate, vitamin B-12, and vitamin B-6 status, as well as age. It is possible that a substantial majority of the cases of high homocysteine in this older population can be attributed to vitamin status.