Home

There is evidence to suggest that folate, homocysteine, or both affect the (n-3) long chain PUFA composition of tissues; however, this evidence is derived largely from experiments with animals and small observational studies in humans. Results from randomized controlled trials are needed. The objective of this study was to determine whether homocysteine lowering with a B vitamin supplement affects the proportion of (n-3) long-chain PUFA in plasma phosphatidylcholine. We conducted a double-blind, placebo-controlled, randomized clinical trial involving 253 participants, 65 y or older, with plasma homocysteine concentrations of at least 13 µmol/L. Participants in the vitamin group (n = 127) took a daily supplement containing 1000 µg folate. 500 µg vitamin B-12, and 10 mg vitamin B-6 for 2 y. The fatty acid composition of plasma phosphatidylcholine was measured at baseline and at 2 y. Plasma homocysteine concentrations during the course of the study were 4.4 µmol/L lower in the vitamin group than in the placebo group. The proportions of eicosapentaenoic, docosapentaenoic, and docosahexaenoic acids in plasma phosphatidylcholine did not differ between the vitamin and placebo groups at 2 y; the mean differences after adjusting for baseline values and sex were -0.03 (99% CI: -0.22, 0.16), 0.03 (99% CI: -0.03, 0.09), and -0.02 (99% CI: -0.27, 0.24) mol%, respectively. Lowering plasma homocysteine concentrations of older men and women with folate, vitamin B-12, and vitamin B-6 had no effect on the proportion of (n-3) long-chain PUFA in plasma phosphatidylcholine.

Discussion

The results of our study showed that lowering plasma homocysteine concentrations by 4.4 µmol/L for 2 y with high intakes of folate, vitamin B-12, and vitamin B-6 did not alter the (n-3) long chain fatty acid composition of plasma phosphatidylcholine. This finding does not support the hypothesis that homocysteine or high intakes of folate, vitamin B-12, or vitamin B-6 influence the metabolism of (n-3) fatty acids in the body.

Krauss-Etschmann et al. (16) recently reported the outcome of a 2 × 2 factorial, placebo-controlled trial of fish oil and folate supplementation on plasma docosahexaenoic acid in which pregnant women consumed 400 µg of 5-methyltetrahydrofolate daily from wk 22 of gestation to delivery. There was a folate treatment × time interaction (P = 0.047) on the proportion of docosahexaenoic acid but not eicosapentaenoic acid (P = 0.081) in maternal plasma; however, the magnitude of the effect of folate on docosahexaenoic acid was very small. In contrast to the finding in maternal plasma, there was no effect of folate supplementation on docosahexaenoic acid proportions in cord blood plasma (P = 0.095). Krauss-Etschmann et al. (16) suggested that folate supplementation may increase maternal docosahexaenoic acid, a possibility that warrants elucidation in larger trials. It is possible that the difference between these findings and those we report simply reflects the different physiological state of the participants in the 2 studies, pregnant compared with nonpregnant. We also cannot exclude the possibility that our study reflects the absence of long-term effects, whereas the results reported by Kraus-Etschmann reflect transient changes.

ur trial had several design features that made it particularly suited for testing the hypothesis that folate, homocysteine, or both affect docosahexaenoic acid status. The 2-y duration of our trial ensured that plasma vitamin concentrations were maintained at high and steady-state concentrations in plasma (22) and presumably in other tissues, such as liver, for long enough to detect an effect on plasma docosahexaenoic acid if one existed. Participants had high plasma homocysteine concentrations at baseline, 13 µmol/L or higher, and the vitamin supplement lowered homocysteine concentration by 4.4 µmol/L compared with placebo. Despite the long duration of supplementation and large difference in homocysteine concentrations between the treatment and placebo groups, there was no difference in docosahexaenoic acid in plasma phosphatidylcholine at 2 y; even the outer CI of the mean difference (-0.27 mol% and 0.24 mol%) exclude meaningful effects when compared with those caused by small changes in intake of docosahexaenoic acid (23).

The hypothesis that folate deficiency or increased homocysteine concentrations affect (n-3) long chain PUFA composition of tissues originated from results of experiments with animals exposed to folate-deficient diets or extremely high concentrations of homocysteine, physiological conditions and exposures that are quite different to those faced by our participants, who were selected for having a serum homocysteine concentration 13 µmol/L or higher. Adult rats fed a folate-deficient diet for 6 wk had decreased eicosapentaenoic and docosahexaenoic acid composition of plasma and platelet lipids compared with rats fed a folatereplete diet (10). Chick embryos injected with homocysteine (0, 100, 200, and 300 µmol/kg egg) during the first 3 d of embryonic development, had a marked decrease in the percentage of total fatty acids as docosahexaenoic acid in egg phospholipids at 11 d of embryonic development (11). Finally, i.m. injection of 5-methyl tetrahydrofolate (150 µg/d) into rats for 15 d resulted in a higher proportion of docosahexaenoic acid in plasma and erythrocyte phospholipids compared with saline-injected animals (12).

Two mechanisms have been proposed through which folate and homocysteine may exert an effect on docosahexaenoic acid. Durand et al. (10) have suggested that by reducing homocysteine concentrations, folate may reduce the generation of reactive oxygen species and thus spare docosahexaenoic acid, which is a major target for lipid peroxidation. The second mechanism is more complex and is based on the involvement of folate in liver synthesis of phosphatidylcholine through its regulation of 1 -carbon metabolism (13). Synthesis of phosphatidylcholine in the liver and secretion into lipoproteins is a major route for the appearance of docosahexaenoic acid into plasma. Despite the plausibility of the proposed mechanisms, the exact metabolic steps that would account for the preferential increase in docosahexaenoic acid as compared with other (n-6) or (n-3) long chain PUFA have not been elucidated.