Dose-dependent effects of folic acid on blood concentrations of homocysteine: a meta-analysis of the randomized trials

¹ From the Clinical Trial Service Unit, University of Oxford, Oxford, United Kingdom (RC, PS, SL, and RC), and the Medical Statistics Unit, London School of Hygiene and Tropical Medicine, London, United Kingdom (CF). (The names of the authors whose initials are listed here are given in the acknowledgments section at the end of the article.)

² Supported by the British Heart Foundation, the Medical Research Council, and the European Union BIOMED Program (BMH4-98-3549).

³ Reprints not available. Address correspondence to R Clarke, Clinical Trial Service Unit, Radcliffe Infirmary, Oxford, OX2 6HE, United Kingdom. E-mail: robert.clarke{at}ctsu.ox.ac.uk.

See corresponding editorial on page 717.

ABSTRACT  
Background: Dietary supplementation with B vitamins that lower blood homocysteine concentrations is expected to reduce cardiovascular disease risk, but there has been uncertainty about the optimum regimen to use for this purpose.

Objective: The objectives were to ascertain the lowest dose of folic acid associated with the maximum reduction in homocysteine concentrations and to determine the additional relevance of vitamins B-12 and B-6.

Design: A meta-analysis of 25 randomized controlled trials involving individual data on 2596 subjects assessed the effect on plasma homocysteine concentrations of different doses of folic acid and of the addition of vitamins B-12 and B-6.

Results: The proportional reductions in plasma homocysteine concentrations produced by folic acid were greater at higher homocysteine (P < 0.001) and lower folate (P < 0.001) pretreatment concentrations; they were also greater in women than in men (P < 0.001). After standardization for sex and to pretreatment plasma concentrations of 12 µmol homocysteine/L and 12 nmol folate/L, daily doses of 0.2, 0.4, 0.8, 2.0, and 5.0 mg folic acid were associated with reductions in homocysteine of 13% (95% CI: 10%, 16%), 20% (17%, 22%), 23% (21%, 26%), 23% (20%, 26%), and 25% (22%, 28%), respectively. Vitamin B-12 ( Conclusions: Daily doses of 0.8 mg folic acid are typically required to achieve the maximal reduction in plasma homocysteine concentrations produced by folic acid supplementation. Doses of 0.2 and 0.4 mg are associated with 60% and 90%, respectively, of this maximal effect.

Key Words: Homocysteine • folic acid • randomized trial

INTRODUCTION  
Elevated blood concentrations of homocysteine have been suggested as a modifiable risk factor for coronary artery disease, stroke, and dementia (1–3). Dietary intake of folate is a major determinant of blood homocysteine concentrations, and vitamin supplements containing folic acid effectively lower homocysteine concentrations (4, 5). However, the optimum dose of folic acid associated with the maximal reduction in homocysteine concentrations is not reliably known. Several large-scale randomized trials are currently underway to test whether lowering homocysteine concentrations with high doses of folic acid reduces the risk of recurrent cardiovascular disease (CVD) (6). The Homocysteine Lowering Trialists’ Collaboration was established to determine the size of the reduction in homocysteine concentrations achieved with different oral doses of folic acid and with the addition of vitamin B-12 or B-6 (5). In the first cycle of this meta-analysis (or pooled analysis), which involved individual data from 1114 participants in 12 trials, daily supplementation with 0.5–5 mg folic acid was associated with reductions in plasma homocysteine concentrations of 25%, but the effects of lower daily doses of folic acid could not be investigated (5). The current cycle of this collaboration, which involves individual data from 2596 participants in 25 trials (7–26) was initiated to determine both the dose-dependent effects on plasma homocysteine concentrations of lower daily doses of folic acid (which might be of particular relevance to the fortification of foods with folic acid) and any additive effects of vitamins B-12 or B-6.

SUBJECTS AND METHODS  
We aimed to identify all published or unpublished randomized trials that had assessed the effect on plasma homocysteine concentrations of folic acid supplements, with or without the addition of vitamins B-12 or B-6, and we found 27 such trials (7-28). Studies were not eligible if they did not include an untreated control group, assessed treatment only after methionine loading, or treated subjects for <3 wk. Eligible studies were identified through MEDLINE searches (by using search terms for folic acid, vitamin B-12, and vitamin B-6 and by including non-English-language literature), by scanning reference lists, and through personal contact with relevant investigators. A trial that used a sequential design was excluded from this meta-analysis (29). Of the 27 completed trials identified in June 2002, 1 factorial design trial was excluded because of confounding by concomitant treatment with Omapatrilat (Bristol-Myers Squibb, Princeton, NJ) that had an adverse effect on homocysteine concentrations (27), and data from another trial (28) were not available to this collaboration. Most of the trials had a parallel-group design, but 2 trials (9, 19) had a crossover design; we used only data from the first period of these trials to avoid any carryover effects. The allocated treatment was blinded in all trials except 2 that had untreated controls (11, 13).

Information collected
For each participant enrolled in these trials, we sought details on patient age, sex, smoking habits, history of vascular disease, serum creatinine concentration, vitamin use before randomization, randomly allocated treatment regimen (ie, daily dose of folic acid, daily dose of any vitamin B-12 or B-6, and scheduled duration of treatment), and plasma concentrations of homocysteine, folate, and vitamins B-12 and B-6 before and during the scheduled treatment.

Statistical analysis
The doses of folic acid studied were classified into 5 groups (<0.4, 0.4, >0.4–1.0, >1.0 to <5.0, and 5 mg), and the median daily doses in these groups were 0.2, 0.4, 0.8, 2.0, and 5.0 mg, respectively. Proportional reductions in plasma homocysteine concentrations in the treated and the control groups were determined by an extension of the previously adopted analysis of covariance (5). Preliminary analyses had indicated that the effect on plasma homocysteine concentrations of folic acid supplementation varied according to the use of vitamin B-12 treatment, pretreatment concentrations of homocysteine and folate, and sex. Hence, to obtain standardized dose-specific effects for each trial, we used a linear mixed model that allowed the posttreatment homocysteine concentration to depend on these findings (with the extent of the dependencies for the latter 3 factors being allowed to vary between trials). This linear mixed model also included interactions between folic acid supplementation (any dose), pretreatment plasma concentrations of folate and homocysteine, and sex. The interactions with baseline folate and homocysteine concentration were treated as random effects in view of the significant heterogeneity between the results of individual trials. Heterogeneity in trial-specific effects at each dose were assessed by using likelihood ratio tests, and the pooled estimates were obtained by expressing the trial-specific effects as random effects. These models were also extended to investigate possible effect modification by concomitant vitamin B-6 use, age, duration of treatment, and history of CVD and to assess whether any difference between dose-specific effects was dependent on sex. All analyses were carried out with SAS software (version 8.2; SAS Institute Inc, Cary, NC).

RESULTS  
Characteristics of individual trials
Among the 2596 people included in the 25 randomized trials, the mean (±SD) age was 52 ± 19 y, and the age range was 17–92 y. The mean duration of treatment was 8 ± 6 wk, and it varied from 3–24 wk between the different trials (Table 1). Approximately 46% of the participants were men, and 30% had a history of a vascular disease event. The median pretreatment plasma concentration of homocysteine was 10.5 µmol/L (interquartile range: 8.5–13.6 µmol/L), and that of folate was 14.2 nmol/L (interquartile range: 9.8–21.0 nmol/L). All of the trials compared folic acid treatment with the control treatment or folic acid plus vitamin B-12 or vitamin B-6 (or both) tratment with the control treatment; one trial also compared different doses of folic acid (15), and another also compared folic acid treatment with combination treatment (18) (Table 2). Data on serum creatinine concentrations were available for only 1007 individuals in 9 trials, and the median creatinine concentration in these trials was 90 µmol/L (interquartile range: 80–106 µmol/L).

View this table:
TABLE 1. Selected characteristics of trials of folic acid–based vitamin therapy¹

 

View this table:
TABLE 2. Plasma concentrations of homocysteine in individual trials¹

 
Exploration of heterogeneity
As in a previous analysis of a subset of these trials (5), there was strong evidence from linear mixed models that the proportionalreductions in plasma homocysteine were significantly greater at higher pretreatment concentrations of homocysteine, at lower pretreatment concentrations of folate, and with concomitant use of vitamin B-12 (P < 0.001 for all). Furthermore, the current analysis indicated that the proportional reductions in homocysteine concentrations were significantly greater in women than in men (P < 0.0001). The mean age of the participants, the presence of CVD, and the duration of treatment were not significantly associated with the proportional reductions in homocysteine concentrations achieved with folic acid supplementation. The proportional reductions in homocysteine concentrations in the individual trials after standardization to a pretreatment homocysteine concentration of 12 µmol/L and folate concentration of 12 nmol/L are shown in Figure 1; there is an equal proportion of men and women and no concomitant use of vitamin B-12. Despite adjustment for these interactions, there was still significant heterogeneity between the results of individual trials within each of the folic acid dose groups.

View larger version (32K):
FIGURE 1.. Proportional reductions in plasma homocysteine concentrations with folic acid supplements in individual trials (standardized to pretreatment plasma concentrations of 12 µmol homocysteine/L and 12 nmol folate/L, an equal proportion of males and females, and no concomitant use of vitamin B-12). Trials are sorted by increasing daily doses of folic acid. Squares indicate the ratios of posttreatment plasma homocysteine concentrations in subjects allocated folic acid supplements to the concentrations in control subjects: the size of the square is proportional to the number of persons, and the horizontal line through the square indicates the 95% CI.

 
Effect of different folic acid doses on blood homocysteine
The proportional reductions (and 95% CI) in plasma homocysteine concentrations with different doses of folic acid are shown in Figure 2. After standardization to pretreatment concentrations of 12 µmol homocysteine/L and 12 nmol folate/L and to equal proportions of men and women, median daily doses of 0.2, 0.4, 0.8, 2.0, and 5.0 mg folic acid were associated with 13% (95% CI: 10%, 16%), 20% (17%, 22%), 23% (21%, 26%), 23% (20%, 26%), and 25% (22%, 28%) reductions in homocysteine concentrations, respectively. The reduction in the plasma homocysteine concentration associated with a daily folic acid dose of 0.8 mg was significantly greater than that associated with a dose of 0.4 mg (P = 0.002), and the reduction with a dose of 0.4 mg was significantly greater than that associated with a dose of 0.2 mg (P < 0.0001), respectively. However, there were no significant differences between the homocysteine-lowering effects of a daily folic dose of 0.8 mg and the effects of a dose of 2.0 or 5.0 mg.

View larger version (26K):
FIGURE 2.. Proportional reductions in plasma homocysteine concentrations with different doses of folic acid supplements in men and in women separately (standardized to pretreatment plasma concentrations of 12 µmol homocysteine/L and 12 nmol folate/L). Squares indicate the ratios of posttreatment plasma homocysteine concentrations in subjects allocated folic acid supplements to the concentrations in control subjects: the size of the square is proportional to the number of subjects, and the horizontal line through the square indicates the 95% CI. There was no significant sex x dose interaction for the effect of folic acid on plasma homocysteine concentrations.

 
Sex differences in the effects of folic acid
Folic acid supplementation was associated with a significantly (P < 0.001) greater proportional reduction in homocysteine concentrations in women than in men (Figure 2). There was no evidence that the magnitude of the sex difference in the effect of folic acid was dependent on dose, nor was the sex difference explained by differences in the history of CVD between men and women. It was not possible to assess the extent to which the greater proportional and absolute reductions in homocysteine in women than in men could be explained by differences in kidney function, because data on creatinine concentrations were available for only 1007 of the 2591 subjects studied, and more accurate measures of glomerular filtration rate were not available.

Effect of adding vitamin B-12 or vitamin B-6 to folic acid
The addition of vitamin B-12 (median dose: 0.4 mg) to folic acid reduced homocysteine concentrations by 7% (95% CI: 4%, 9%) (P < 0.0001) more than did folic acid alone. After standardization to pretreament concentrations of 12 µmol homocysteine/L and 12 nmol folate/L and to equal proportions of men and women, the addition of vitamin B-12 to a folic acid dose of 5 mg changed the reduction in homocysteine from 25% (22%, 28%) to 30% (27%, 33%) (P < 0.0001). The addition of vitamin B-12 to median folic acid doses of 0.2, 0.4, 0.8, and 2.0 mg/d resulted in homocysteine reductions of 19%, 25%, 29% and 28%, respectively. The data were insufficient for examination of differences in the effect of different doses of vitamin B-12. The addition of vitamin B-6 (mean: 12 mg/d) to folic acid was not associated with any further reduction in plasma homocysteine concentrations.

Effect of pretreatment concentrations of folate and homocysteine
The extent to which the proportional reduction in plasma homocysteine concentrations achieved with folic acid depends on the pretreatment plasma concentrations of both folate and homocysteine is shown in Table 3. The results presented are the predicted reductions with a dose of 0.8 mg folic acid/d and without concomitant vitamin B-12 supplementation in a population made up of equal proportions of men and women.

View this table:
TABLE 3. Predicted proportional reductions in plasma homocystene concentrations with 0.8 mg folic acid/d for individuals at various pretreatment blood concentrations of folate and homocysteine¹

 

DISCUSSION  
This meta-analysis provides reliable evidence of a dose-dependent reduction in plasma homocysteine concentrations with incremental doses of folic acid up to 0.8 mg/d, which was the lowest daily dose of folic acid associated with the maximal reduction in homocysteine concentrations. Daily doses of 0.2 and 0.4 mg folic acid were associated with 60% and 90%, respectively, of the maximal reduction in blood homocysteine concentrations produced by folic acid supplementation. Vitamin B-12 was associated with an additional 7% reduction in plasma homocysteine concentrations. After standardization to pretreatment plasma concentration of 12 µmol homocysteine/L and 12 nmol folate/L, coadministration of folic acid (0.8 mg) and vitamin B-12 lowered homocysteine concentrations by 30%, which is equivalent to an absolute reduction of 3–4 µmol/L in populations with typical homocysteine concentrations of 10–12 µmol/L.

The results of this meta-analysis with respect to the homocysteine-lowering efficacy of daily doses of 0.2 and 0.4 mg folic acid and the additive effects of vitamin B-12 are relevant to the debate on the fortification of foods with folic acid and vitamin B-12. The introduction of mandatory folic acid fortification in North America during 1998 increased the mean plasma folate concentration from 11 to 23 nmol/L and reduced the mean homocysteine concentration in middle-aged persons to 8–10 µmol/L (30). The present meta-analysis also shows that additional supplementation with folic acid is likely to lower homocysteine concentrations by only 15% in such fortified populations, and these predictions have been confirmed in trials conducted after the introduction of mandatory folic acid fortification (31).

There was significant heterogeneity of the results of individual trials within each of the folic acid dose groups in this meta-analysis. Heterogeneity of effects may be caused by differences in sample handling or in differences the assays used to measure homocysteine concentrations, but the approach to analysis (which involved within-trial comparisons) should have corrected for such effects. Some of the heterogeneity may reflect differences in the populations studied, including the introduction of folic acid fortification, the proportion with the MTHFR polymorphism, or the degree of renal function. By standardizing for the pretreatment concentrations of folate and of homocysteine, however, the analysis should have corrected for the effects of differences in the mean pretreatment concentrations of homocysteine and folate (such as those produced by fortification, genetic variants, or differences in renal function), but it was not possible to correct directly for differences in genotype or renal function. Although folic acid is effective in reducing homocysteine concentrations, it rarely succeeds in normalizing homocysteine concentrations in patients with end-stage renal disease.

Large trials are currently underway to determine whether folic acid supplementation can reduce the risk of CVD (6). Almost all of these trials are using doses of folic acid (median: 2 mg; range: 0.5–40 mg) that are much larger than those available in most multivitamin supplements. The current analyses indicate that the proportional reduction in homocysteine following any dose of folic acid is significantly greater in women than in men. If the moderately greater risk of CVD associated with higher homocysteine concentrations is causal, then (other things being equal), folic acid supplementation may produce greater reductions in women.

The present meta-analysis indicates that little further homocysteine reduction is achieved by increasing the dose of folic acid above 0.8 mg/d, but combined administration of folic acid with vitamin B-12 will achieve a greater reduction in plasma homocysteine concentration than does that of folic acid alone.

ACKNOWLEDGMENTS  
The following were members of the Homocysteine Lowering Trialists’ Collaboration: R Clarke, C Frost, P Sherliker, S Lewington, and R Collins (writing committee); L Brattstrom (Brattstrom group); I Brouwer, M van Dusseldorp, and RPM Steegers-Theunissen (Brouwer group); G Cuskelly, M Ward, H McNulty, and J Scott (Cuskelly group); M den Heijer, H Blom, and N van der Put (den Heijer group); CJ Shorah (Lucock group); MR Malinow (Malinow group); M McMahon, J Tobert, and D Kush (McMahon group); E Joosten and R Riezler (Naurath group); K Pietrzik, J Dierkes, and A Bronstrup (Pietrzik group); P Jacques, J Mason, and I Rosenberg (Saltzman group); J Thambyrajah, M Landray, J Townend, and D Wheeler (Thambyrajah group); J Ubbink (Ubbink group); F van Oort, A Melse-Boonstra, and P Verhoef (van Oort group); JV Woodside, J Yarnell, IS Young, and AE Evans (Woodside group); and D Wald, M Law, and N Wald (Wald group).

RC, CF, SL, PS, and RC designed the study. RC, PS, and SL collected and assembled the data. PS and CF analyzed the data, and PS produced the tables and figures. RC, SL, PS, CF, and RC interpreted the data, wrote the manuscript, provided advice or consultation, and gave final approval of the manuscript. None of the authors had any personal or financial conflicts of interest.

REFERENCES  

1. Homocysteine Studies Collaboration. Homocysteine and risk of ischemic heart disease and stroke: a meta-analysis. JAMA 2002;288:2015–22.
2. Wald DS, Law M, Morris JK. Homocysteine and cardiovascular disease: evidence on causality from a meta-analysis. BMJ 2002;325:1202–6.
3. Clarke R, Smith AD, Jobst K, Sutton L, Refsum H, Ueland PM. Plasma homocysteine, folate and vitamin B-12 as risk factors for confirmed Alzheimer’s disease. Arch Neurol 1998;55:1449–55.
4. Selhub J, Jacques PF, Wilson PW, Rush D, Rosenberg IH. Vitamin status and intake as primary determinants of homocysteinemia in an elderly population. JAMA 1993;270:2693–8.
5. Lowering blood homocysteine with folic acid based supplements: meta-analysis of randomised trials. Homocysteine Lowering Trialists’ Collaboration. Br Med J 1998;316:894–8.
6. Clarke R, Armitage J. Vitamin supplements and cardiovascular risk: review of the randomized trials of homocysteine-lowering vitamin supplements. Semin Thromb Haemost 2000;26:341–8.
7. Ubbink JB, van der Merwe A, Vermaak WJH, Delport R. Hyperhomocysteinemia and the response to vitamin supplementation. Clin Invest 1993;71:993–8.
8. Ubbink JB, Vermaak WJH, van der Merwe A, Becker PJ, Delport R, Potgieter HC. Vitamin requirements for the treatment of hyperhomocysteinemia in humans. J Nutr 1994;124:1927–33.
9. Dierkes J. Vitamin requirements for the reduction of homocysteine blood levels in healthy young women . Bonn: University of Bonn, 1995.
10. Saltzman E, Mason JB, Jacques PF, et al. B-vitamin supplementation lowers homocysteine levels in heart disease. Clin Res 1994;42:172A (abstr).
11. Landgren F, Israelsson B, Lindgren A, Hultberg B, Andersson A, Brattström L. Plasma homocysteine in acute myocardial infarcation: homocysteine-lowering effect of folic acid. J Intern Med 1995;237:381–8.
12. Naurath HJ, Joosten E, Reizler R, Stabler SP, Allen RH, Lindenbaum J. Effects of vitamin B12_, folate, and vitamin B6 supplements in elderly people with normal serum vitamin concentrations. Lancet 1995;364:85–9.
13. Cuskelly G, McNulty W, McPartlin J, Strain JJ, Scott JM. Plasma homocysteine response to folate intervention in young women. Ir J Med Sci 1995;164:3 (abstr).
14. Woodside JV, Yarnell JWG, Young IS, McMaster D, Gey F, Patterson CC, Evans AE. The effect of B-vitamins and antioxidant vitamins on hyperhomocysteinemia: a double-blind randomized factorial design controlled trial. Am J Clin Nutr 1998;67:858–66.
15. Malinow MR, Nieto FJ, Kruger WD, et al. The effects of folic acid supplementation on plasma total homocysteine are modulated by multivitamin use and methylenetetrahydrofolate reductase genotypes. Arterioscler Throb Vasc Biol 1997;17:1157–62.
16. Den Heijer M, Brouwer IA, Bos GMJ, et al. Vitamin supplementation reduces blood homocysteine levels: a controlled trial in patients with venous thrombosis and healthy volunteers. Arterioscler Thromb Vasc Biol 1998;18:356–61.
17. Schorah CJ, Devitt H, Lucock M, Dowell AC. The responsiveness of plasma homocysteine to small increases in dietary folic acid: a primary care study. Eur J Clin Nutr 1998;52:407–411.
18. Malinow MR, Duell PB, Hess DL, et al. Reduction of plasma homocyst(e)ine levels by breakfast cereals fortified with folic acid in patients with coronary heart disease. N Engl J Med 1998;338:1009–15.
19. Bronstrup A. Effects of single and combined B-vitamin supplementation on homocysteine concentrations in different population groups. Doctoral dissertation. University of Bonn, Bonn, Germany, 1998.
20. Dierkes J, Kroesen M, Pietrzik K. Folic acid and vitamin B6 supplementation and plasma homocysteine concentrations in healthy young women. Int J Vitam Nutr Res 1998;68:98–103.
21. Brouwer IA, van Dusseldorp M, Thomas CMG, et al. Low-dose folic acid supplementation decreases plasma homocysteine: a randomized trial. Am J Clin Nutr 1999;69:99–104.
22. MacMahon M, Kirkpatrick C, Cumming SCE, et al. A pilot study with Simvastatin and folic acid/vitamin B12 in preparation for the study of the effectiveness of additional reductions in cholesterol and homocysteine (SEARCH). Nutr Metab Cardiovasc Dis 2000;10:195–203.
23. Thambyrajah J, Landray MJ, McGlynn FJ, Jones HJ, Wheeler DC, Townend JN. Does folic acid decrease plasma homocysteine and improve endothelial function in patients with predialysis renal failure? Circulation 2000;102:871–5.
24. Wald DS, Bishop L, Wald NJ, et al. Randomized trial of folic acid supplementation and serum homocysteine levels. Arch Intern Med 2001;161:695–700.
25. Thambyrajah J, Landray MJ, Jones HJ, McGlynn FJ, Wheeler DC, Townend JN. A randomized double blind placebo controlled trial of the effect of homocysteine- lowering therapy with folic acid on endothelial function in patients with coronary artery disease. J Am Coll Cardiol 2001;37:1858–63.
26. van Oort FVA, Boonstra AM, Brouwer I, et al. Folic acid and plasma homocysteine reduction in older adults: a dose-finding study. Am J Clin Nutr 2003;77:1318–23.
27. Neal B, MacMahon S, Ohkubo T, TonkinA, Wilcken D, for the Pacific Study Group. Dose-dependent effects of folic acid on plasma homocysteine in a randomised trial conducted among 723 individuals with coronary heart disease. Eur Heart J 2002;23:1509–15.
28. Rydlewicz A, Simpson JA, Taylor RJ, Bond CM, Golden MHN. The effect of folic acid supplementation on plasma homocysteine in an elderly population. Q J Med 2002;95:27–35.
29. Ward M, McNulty H, McPartlin J, Strain JJ, Weir DG, Scott JM. Plasma homocysteine, a risk factor for cardiovascular disease, is lowered by physiological doses of folic acid. Q J Med 1997;90:519–24.
30. Jacques PF, Selhub J, Bostom AG, Wilson PWF, Rosenberg IH. The effect of folic acid fortification on plasma folate and total homocysteine concentrations. N Engl J Med 1999;340:1449–54.
31. Toole JF, Malinow MR, Chambless LE, et al. Lowering homocysteine in patients with ischemic stroke to prevent recurrent stroke, myocardial infarction, and death: the Vitamin Intervention for Stroke Prevention (VISP) randomized controlled trial. JAMA 2004;291:565–75.