A dose-finding trial of the effect of long-term folic acid intervention: implications for food fortification policy 1 3

Transcription

1 See corresponding editorial on page 1. A dose-finding trial of the effect of long-term folic acid intervention: implications for food fortification policy 1 3 Paula Tighe, Mary Ward, Helene McNulty, Owen Finnegan, Adrian Dunne, JJ Strain, Anne M Molloy, Maresa Duffy, Kristina Pentieva, and John M Scott ABSTRACT Background: The lowest dose of folic acid required to achieve effective reductions in homocysteine is controversial but important for food fortification policy given recent concerns about the potential adverse effects of overexposure to this vitamin. Objective: We compared the effectiveness of 0.2 mg folic acid/d with that of 0.4 and 0.8 mg/d at lowering homocysteine concentrations over a6-moperiod. Design: A randomized dose-finding trial with folic acid was conducted. Of 203 participants screened, 101 patients with ischemic heart disease and 71 healthy volunteers completed the study. Participants were randomly assigned to receive placebo or folic acid at doses of 0.2, 0.4, or 0.8 mg/d for 26 wk; subsamples of patients with ischemic heart disease were also examined at 6 or 12 wk. Results: Participants with higher baseline homocysteine concentrations had the greatest reductions in homocysteine in response to folic acid doses of 0.2 mg (220.6%), 0.4 mg (220.7%), and 0.8 mg (227.8%); in those with lower baseline homocysteine concentrations, the responses were 28.2%, 28.9%, and 28.3%, respectively. No significant differences in homocysteine responses to the different doses were observed. In the patient group sampled at intervals during the intervention, the maximal homocysteine response appeared to be achieved by 6 wk in the 0.8-mg/d group and by 12 wk in the 0.4-mg/d group. However, the homocysteine response was suboptimal in the 0.2-mg/d group at both 6 and 12 wk compared with that at 26 wk. Conclusions: A folic acid dose as low as 0.2 mg/d can, if administered for 6 mo, effectively lower homocysteine concentrations. Higher doses may not be necessary because they result in no further significant lowering, whereas doses even lower than 0.2 mg/d may be effective in the longer term. Previous trials probably overestimated the folic acid dose required because of a treatment duration that was too short. This trial was registered at clinicaltrials.gov as ISRCTN Am J Clin Nutr 2011;93:11 8. Moreover, a recent meta-analysis of randomized trials showed that folic acid significantly reduced the risk of stroke overall by 18%, but to a greater extent by up to 25% in those trials that showed greater homocysteine lowering or in persons with no history of stroke (10). The reported improvement in stroke mortality in North America that relates to the timing of introduction of mandatory folic acid fortification (11) adds further support to the potential benefit of enhancing folate status and/or lowering homocysteine in the primary prevention of stroke. The first meta-analysis to assess the effect of folic acid showed that doses ranging from 0.5 to 5 mg/d could decrease homocysteine by 20 25% (12). Although 2 other studies suggested that folic acid ranging from 0.2 to 0.4 mg/d could effectively decrease homocysteine (13, 14), both of these studies had limitations that affected the interpretation of their findings. One study was conducted exclusively in young women, thereby limiting the application of the results to the general population. In the earlier trial (also by our group), an increasing doses model was used in which participants acted as their own controls; therefore, the findings, although informative, lacked the robustness of those from a randomized controlled trial. Another trial reported that folic acid doses as high as 0.8 mg/d were required for maximal homocysteine lowering in patients with ischemic heart disease (IHD) (15). Likewise, a subsequent meta-analysis of 25 randomized trials involving 2596 subjects concluded that 0.8 mg folic acid/d was required to achieve a maximal reduction in homocysteine (16). A dose of folic acid as high as 0.8 mg/d is not, however, a feasible target for food fortification because achieving this level as a population average will result in some people being exposed to much higher folic acid intakes. Moreover, long-term exposure to folic acid (the synthetic form of the vitamin) is as- INTRODUCTION Evidence from numerous retrospective and prospective studies predicted that lowering homocysteine by 25% would reduce the risk of heart disease by 11 16% and stroke by 19 24% (1, 2). However, the secondary prevention trials in at-risk patients published since have generally failed to show a benefit of homocysteine-lowering therapy on cardiovascular events (3 8). One of these trials did, however, show a clear benefit in reducing the risk of stroke. Although this benefit was largely overlooked in the original report (4), it was recently reported separately (9). 1 From the Northern Ireland Centre for Food and Health, School of Biomedical Sciences, University of Ulster, Coleraine, Northern Ireland (PT, MW, HM, JJS, MD, and KP); the Causeway Hospital, Coleraine, Northern Ireland (OF); the School of Mathematical Sciences, University College, Dublin, Ireland (AD); and the Schools of Clinical Medicine (AM) and Biochemistry & Immunology (JMS), Trinity College, Dublin, Ireland. 2 Supported by The Northern Ireland Chest Heart and Stroke Association. 3 Address correspondence to H McNulty, Northern Ireland Centre for Food and Health, University of Ulster, Cromore Road, Coleraine, BT52 1SA Northern Ireland. h.mcnulty@ulster.ac.uk. Received February 24, Accepted for publication September 24, First published online October 27, 2010; doi: /ajcn Am J Clin Nutr 2011;93:11 8. Printed in USA. Ó 2011 American Society for Nutrition 11

2 12 TIGHE ET AL sociated with safety concerns, including the potential risk that high folic acid intake might mask the anemia of vitamin B-12 deficiency (17) or that it may be associated with an increased risk of cognitive impairment in older persons with a low vitamin B-12 status (18). Furthermore, despite considerable evidence that folate within the dietary range plays a protective role against various cancers (19, 20), high-dose folic acid may promote colorectal tumorigenesis in patients with preexisting lesions (21) or even increase the cancer risk in general (22, 23). Because of these concerns, many governments worldwide have delayed decisions to implement population-wide folic acid fortification policies similar to those introduced.10 y ago in North America. It is important to ensure that the known and potential beneficial effects of folic acid in preventing disease can be achieved without exposing the general population to unnecessarily high doses of folic acid. The aim of this study was to determine the minimum dose of folic acid (within the range of mg/d) required to effectively lower homocysteine over a longer duration of intervention than previously investigated. SUBJECTS AND METHODS Participants Ethical approval of the study was granted by the Research Ethical Committee of the University of Ulster. Patients with IHD, of any age, were recruited from the Cardiac Rehabilitation program, Causeway Hospital, Northern Ireland. Inclusion criteria were as follows: proven myocardial infarction.3 mo previously, IHD on coronary angiography, or a clinical diagnosis of angina confirmed by electrocardiogram. Healthy subjects were recruited from the local community. The exclusion criteria for all subjects were as follows: history of diabetes, hepatic or renal disease, hematologic disorders, use of B vitamin supplements, or use of medication known to interfere with folate metabolism. In addition, the healthy group had no history of cardiovascular disease. All participants gave written informed consent and completed a short medical questionnaire. As part of the ethical considerations, screening blood samples were collected for analysis of vitamin B-12 status before the intervention started, and those subjects found to be deficient (,140 nmol/l) were excluded from the study and referred to their medical practitioner for treatment. Intervention and randomization The study was conducted as a double-blinded, randomized, placebo controlled trial. The sample size was estimated by using data from a similar dose-finding study in IHD patients (15). On the basis of the mean (6SD) homocysteine response to different folic acid treatments compared with that to placebo in this previous study (15), it was estimated that a sample size of 29 subjects per group was needed to detect a difference of 2.7 lmol homocysteine/l with a power of 80% at a = To allow for a 20% dropout rate, we estimated that a sample size of 35 subjects in each folic acid treatment group would be required. Within both the IHD and healthy groups, participants were stratified into tertiles of homocysteine concentration (from the screening blood sample); subjects in each stratum were then randomly assigned to receive placebo or 0.2, 0.4, or 0.8 mg folic acid/d for a total intervention period of 26 wk. Vitamins were specially produced for this study by Boots Contract Manufacturing (Nottingham, United Kingdom) in tablet form. All folic acid tablets, including placebo, were identical in color, size, and shape. To maximize compliance, vitamins were distributed every 3 wk to the participants homes in 7-d pillboxes with instructions to take one tablet each morning. The pillboxes were then collected, and the number of unused tablets was recorded to monitor compliance. Participants were instructed not to consume any B vitamin supplements other than those provided during the study and were advised to maintain their usual diet, including their intake of any folic acid fortified foods, throughout the intervention period. Sample collection and measurements Fasting blood samples were collected from participants in their own homes at screening, week 0, and week 26 and, in addition, in a subset of participants (IHD patients only) at 6 or 12 wk of intervention. Samples collected for plasma homocysteine analysis were wrapped in foil, placed on ice immediately after collection, and centrifuged within 4 h of sampling. Samples collected at screening and before, during, and after intervention were analyzed by using standard laboratory assays for plasma total homocysteine (24) and serum folate (25). Samples collected at screening were analyzed for serum vitamin B-12 (26), vitamin B-6 (plasma pyridoxal phosphate, PLP) (27), and riboflavin status. The status of riboflavin was determined on the basis of the erythrocyte glutathione reductase activation coefficient (EGRac) a functional assay that measures the activity of glutathione reductase before and after in vitro reactivation with its prosthetic group FAD (28). EGRac is calculated as a ratio of FAD-stimulated to -unstimulated enzyme activity, with values 1.3 generally indicative of suboptimal riboflavin status. The MTHFR 677C/T genotype was identified by polymerase chain reaction (PCR) amplification followed by HinF1 restriction digestion (29). For all assays, samples were analyzed blind, and quality control was provided by the repeated analysis of stored batches of pooled washed red blood cells (EGRac), plasma (homocysteine and PLP), and serum (folate and vitamin B-12) covering a wide range of values. The folate content of the intervention tablets was determined 3 times during the intervention period by using a Lactobacillus casei microbiological assay (25) after dissolving the tablets in 1 ml of 0.01 mol NaOH/L and further diluting the solutions with 0.5% sodium ascorbate. The mean (6SD) measured folate contents of the 0.2, 0.4, and 0.8 mg folic acid tablets were , , and mg per tablet, respectively. Folate was undetectable in the placebo tablet. Dietary evaluation and anthropometric measurements Dietary intakes at baseline were assessed by means of a 4-d food diary. Subjects were given detailed advice, both written and oral, on how to complete the food diaries and were asked to do so over 2 weekdays and 2 weekend days. Portion sizes were estimated by using household measurements and were subsequently quantified by using published data (30). The food diaries were then analyzed for daily nutrient intakes by using the dietary analysis program WISP (version 1.28; Tinuviel Software, Warrington, United Kingdom). Weight (kg) and height (m) were measured at the time of the dietary assessments.

3 EFFECTS OF LONG-TERM FOLIC ACID INTERVENTION 13 Statistical analysis Statistical analysis was performed by using the SPSS Statistical Package for the Social Sciences (version 17.0; SPSS UK Ltd, Chersey, United Kingdom). For normalization purposes, variables were log transformed as appropriate before statistical analysis. Differences at baseline between IHD patients and healthy volunteers were assessed by using an independentsamples t test. Categorical data were assessed by using chisquare analysis. Response to intervention was calculated as the posttreatment minus the pretreatment value. The percentage response to the intervention was calculated as the posttreatment minus the pretreatment values and expressed as a percentage of the pretreatment value. Differences in the response to the intervention between treatment groups were examined by using one-factor analysis of covariance (ANCOVA), with age, sex, and baseline homocysteine as covariates. Adjustments for multiple comparisons with the use of a Bonferroni post hoc test were incorporated into the analyses. When examining the response to intervention according to initial homocysteine concentrations, subjects were categorized into low or high homocysteine by using the median baseline homocysteine concentration as a cutoff. Differences at baseline between subjects with low or high initial homocysteine concentrations (matched for age) were assessed by using an independent-samples t test. Univariate analysis was used to examine the determinants of homocysteine concentrations before and after the intervention. P values,0.05 were considered significant. RESULTS Baseline characteristics The response rate for eligible IHD patients initially approached to take part in the intervention was.95%. Details of the patients and healthy volunteers screened, those randomly assigned to treatment (following application of exclusion criteria), and completion rates are provided in Figure 1. Of the 113 patients and 75 healthy volunteers randomly assigned to treatment, 16 failed to complete intervention, which resulted in 101 patients and 71 healthy volunteers completing the study, all of whom were included in the analysis. Characteristics of the patients and healthy volunteers are listed in Table 1. Homocysteine concentrations were higher and concentrations of serum folate, vitamin B-12, and vitamin B-6 were lower in the patients than in the healthy volunteers. No significant differences in riboflavin status or in the frequency of the homozygous mutant (TT) genotype for the MTHFR 677C/T polymorphism were observed between the 2 groups. Reported dietary energy intakes were lower in the patients than in the healthy volunteers (possibly as a result of dietary restriction in the patient group after receiving dietary advice at the time of their diagnosis). No significant differences in dietary intakes of the relevant B vitamins were found, except that vitamin B-6 was higher in the healthy group, and B vitamin intakes in both groups compared favorably with current recommendations. However, when nutrient intakes were expressed per unit energy and reexamined, folate intakes were found to be significantly higher in the patients (P = 0.03), whereas the apparent difference in vitamin B-6 was no longer significant (data not shown). Response to folic acid Mean compliance was estimated (by pill counting) to be 98% in patients and 95% in the healthy volunteers. Serum folate concentrations responded to the intervention in a linear, dose-dependent manner with responses increasing significantly in response to each increment in folic acid dose (Figure 2), FIGURE 1. Flow of participants through each stage of the study. IHD, ischemic heart disease.

4 14 TIGHE ET AL TABLE 1 General characteristics of patients with ischemic heart disease (IHD) and healthy volunteers at screening 1 IHD patients (n = 101) Healthy group (n = 71) P value 2 Age (y) Male sex (%) BMI (kg/m 2 ) Homocysteine and B vitamin status Plasma homocysteine (lmol/l) 13.4 (10.3, 16.4) (8.8, 13.0), Serum folate (lg/l) 7.9 (5.4, 11.5) 9.8 (6.6, 13.7) Serum vitamin B-12 (ng/l) 249 (180, 350) 305 (233, 382) Plasma vitamin B-6, PLP (nmol/l) 54.9 (40.2, 71.1) 74.1 (51.1, 107.5), Riboflavin (EGRac) 1.30 (1.21, 1.35) 1.26 (1.20, 1.35) MTHFR 677C/T genotype (%) CC CT TT 8 12 Dietary intakes Energy (MJ/d) (5.829, 8.320) (6.558, 9.775) Dietary folate (lg/d) 219 (186, 292) 229 (182, 275) Dietary vitamin B-12 (lg/d) 3.5 (2.8, 4.6) 3.7 (2.6, 4.5) Dietary vitamin B-6 (mg/d) 1.9 (1.6, 2.4) 2.2 (1.8, 2.5) Dietary riboflavin (mg/d) 1.5 (1.2, 1.9) 1.7 (1.3, 2.1) SI conversion factors: to convert folate to nmol/l, multiply by 2.266; to convert vitamin B-12 to pmol/l, multiply by PLP, pyridoxal phosphate; EGRac, erythrocyte glutathione reductase activation coefficient (a higher value indicates lower riboflavin status); MTHFR, methylenetetrahydrofolate reductase [CC (wild type), CT (heterozygous), and TT (homozygous mutant) genotypes for the MTHFR 677C/T polymorphism]. 2 Differences between groups were assessed by using an independent-samples t test. Data were log transformed before statistical analysis when applicable. Categorical variables were assessed by using chi-square analysis. P, 0.05 was considered significant. 3 Mean 6 SD (all such values, normally distributed) 4 Geometric mean; interquartile range in parentheses (all such values, log transformed before statistical analysis). which provided further evidence that compliance was generally high and consistent between treatments. Although the magnitude of the homocysteine response appeared to differ between the patient and healthy groups, no differences in responses were observed after baseline homocysteine was controlled for (data not shown). Therefore, the data were combined and the responses to intervention were examined by using one-factor ANCOVA (with adjustment for baseline homocysteine, age, and sex) and a Bonferroni post hoc test (P, 0.05) (Table 2). Folic acid at all doses resulted in significant homocysteine lowering, and no significant differences in the homocysteine responses to the different doses of folic acid were observed. Although no significant differences in homocysteine responses to folic acid doses of 0.2 and 0.8 mg/d were observed, analysis of the 95% CI for the difference in response between 0.2 and 0.8 mg/d showed that this could be as great as 2.2 lmol/l. The homocysteine response to the intervention was then examined according to initial homocysteine concentrations. In both the high- and low-homocysteine groups, all doses of folic acid were effective at lowering homocysteine, with no significant differences observed between the folic acid doses within either category (Figure 3). Participants with a high homocysteine concentration at baseline (ie, an initial homocysteine concentration greater than the median value) showed the greatest magnitude of homocysteine response to folic acid doses of 0.2 mg (220.6%), 0.4 mg (220.7%), and 0.8 mg (227.8%) (Figure 3). In participants with a low baseline homocysteine concentration (ie, a value below the median), homocysteine responses were 28.2%, 28.9%, and 28.3% for folic acid doses of 0.2, 0.4, and 0.8 mg/d, respectively. When the characteristics of the participants with low or high initial homocysteine concentrations were further examined, those with high homocysteine were typically older and had lower serum concentrations of both folate and vitamin B-12. No differences in sex or MTHFR genotype were observed. After the high- and low-homocysteine groups were matched for age, both serum folate and vitamin B-12 concentrations were found to be significantly lower in those with a higher initial homocysteine concentration (Table 3). A univariate model of analysis showed that vitamin B-12 concentration was a significant determinant of FIGURE 2. Serum folate response to folic acid intervention in patients with ischemic heart disease and healthy volunteers combined (n = 172). Values are expressed as geometric means (error bars indicate 95% CIs). Differences between groups were analyzed by one-factor ANOVA. Different lowercase letters indicate significant differences between postintervention values, P, 0.05 (Bonferroni post hoc test).

5 EFFECTS OF LONG-TERM FOLIC ACID INTERVENTION 15 TABLE 2 Plasma homocysteine response to folic acid intervention in patients with ischemic heart disease and healthy volunteers combined 1 Placebo (n = 40) 0.2 mg Folic acid (n = 44) 0.4 mg Folic acid (n = 43) 0.8 mg Folic acid (n = 45) lmol/l lmol/l lmol/l lmol/l Week (10.7, 13.4) 12.6 (11.8, 15.3) 12.6 (11.7, 15.7) 13.1 (12.3, 15.6) Week (11.0, 13.4) 10.6 (9.9, 12.4) 10.6 (9.9, 12.1) 10.3 (9.8, 11.6) Response 3, (21.1, 0.3) a 22.1 (22.8, 21.5) b 22.5 (23.2, 21.9) b 23.1 (23.7, 22.4) b Percentage response 3,5 1.4 (23.1, 5.9) a (218.6, 210.0) b (218.5, 29.9) b (222.5, 214.0) b 1 Statistical significance of the comparison of responses and percentage responses between treatment groups was conducted by ANCOVA with adjustments for age, sex, and baseline homocysteine. Values within a row with different superscript letters are significantly different, P, 0.05 (Bonferroni post hoc test). Mean differences (in lmol/l) in homocysteine responses between treatment groups (after adjustment for age, sex, and baseline homocysteine): placebo compared with 0.2 mg/d, 1.7 (95% CI: 0.4, 3.0); placebo compared with 0.4 mg/d, 2.1 (95% CI: 0.8, 3.4); placebo compared with 0.8 mg/d, 2.6 (95% CI: 1.3, 3.9); 0.2 compared with 0.4 mg/d, 0.4 (95% CI: 20.8, 1.7); 0.4 compared with 0.8 mg/d, 0.5 (95% CI: 20.7, 1.8); and 0.2 compared with 0.8 mg/d, 0.9 (95% CI: 20.3, 2.2). 2 Values are geometric means; 95% CIs in parentheses. 3 Values are estimated means adjusted for age, sex, and baseline homocysteine; 95% CIs in parentheses. 4 The response to intervention was calculated as the posttreatment minus pretreatment value. 5 The percentage response to intervention was calculated as the posttreatment minus pretreatment value expressed as a percentage of the pretreatment value. postintervention homocysteine concentrations only in those with a high homocysteine concentration at baseline (P = 0.007; data not shown). In addition to the 26-wk postintervention sample examined in all participants, a subset of IHD patients were also sampled at 6 or 12 wk (Figure 4). At 6 wk, the homocysteine response differed considerably between the 3 folic acid treatments. At 12 wk, the response to both 0.8 and 0.4 mg folic acid/d appeared to be optimal in that no further homocysteine-lowering in response to either dose was observed between 12 and 26 wk. The extended intervention period up to a total of 26 wk, however, was required to achieve effective homocysteine lowering with 0.2 mg folic acid/d; in this treatment group the response appeared to be suboptimal at both 6 and 12 wk. FIGURE 3. Mean (6SE) homocysteine response to a 26-wk folic acid intervention in patients with ischemic heart disease and healthy volunteers combined (n = 172) according to initial plasma homocysteine concentrations. Subjects were categorized into categories of low or high initial homocysteine concentrations on the basis of the median baseline homocysteine concentration for each treatment group as a cutoff. The homocysteine response to the intervention was calculated as the posttreatment minus the pretreatment value and expressed as a percentage of the pretreatment value. Differences between groups were analyzed by ANCOVA. Different lowercase letters indicate significant differences within each category, P, 0.05 (ANCOVA with Bonferroni post hoc test). DISCUSSION Our results showed that a dose of folic acid as low as 0.2 mg/d results in an effective decrease in homocysteine concentrations. Although serum folate increased significantly in a dosedependent manner with each increase in folic acid from 0.2 to mg/d, the 2 higher doses did not result in a significantly greater homocysteine response than did the dose of 0.2 mg folic acid/d. These findings, shown after a total intervention period of 26 wk, differ from previous conclusions that much higher folic acid doses were necessary on the basis of intervention trials that were typically 8 wk in duration (16). The prolonged period of intervention in this study was found to be necessary to allow a more optimal homocysteine-lowering effect to be observed in response to all folic acid treatments. The results are timely because population-wide fortification with folic acid on a mandatory basis (introduced over 10 y ago in North America) is under consideration by various governments worldwide. Determining the long-term effects of folic acid at lower doses is important given recent concerns regarding potential adverse effects of overexposure to folic acid at high doses (18, 21 23). Homocysteine at baseline was an important predictor of the homocysteine response in this study as in other studies (12, 13, 16), although it is unclear how much of this effect was due to regression to the mean. However, regardless of the initial homocysteine concentration, 0.2 mg folic acid/d given chronically was found to result in effective homocysteine lowering, and the response to 0.8 mg folic acid/d was not significantly greater than the response to 0.2 mg folic acid/d. Although our data show that the difference in response to the 0.2 and 0.8 mg folic acid/d doses may have been as great as 2.2 lmol/l, we considered the risk of overexposure from the long-term delivery of a dose 4 times more than the effective dose to be too great to justify any additional lowering. Folic acid is generally desirable because it provides a stable and bioavailable form of the vitamin for ingestion; however, at higher intakes, unmetabolized folic acid is known to persist in plasma because of the limited ability of humans to reduce it to the active folate form (31). The persistence of unmetabolized folic acid could in turn contribute to the reported adverse effects of excessive folic acid exposure (18,

6 16 TIGHE ET AL TABLE 3 Characteristics of participants with low or high initial homocysteine concentrations, matched for age 1 Low baseline homocysteine 2 (n = 70) High baseline homocysteine 3 (n = 67) P value 4 Baseline B vitamin status Serum folate (lg/l) 10.6 (10.5, 14.0) 6.8 (6.8, 9.0), Serum vitamin B-12 (ng/l) 303 (295, 343) 235 (229, 278), Plasma vitamin B-6, PLP (nmol/l) 66.6 (66.9, 86.9) 62.4 (56.4, 93.2) 0.48 Riboflavin (EGRac) 1.27 (1.25, 1.32) 1.31 (1.28, 1.35) 0.22 Postintervention Hcy (lmol/l) (8.6, 9.3) 12.3 (11.7, 13.7), All values are geometric means; 95% CIs in parentheses. Low and high homocysteine categories were established arbitrarily by using the median baseline homocysteine concentration as a cutoff and then matched for age between the 2 categories. SI conversion factors: to convert folate to nmol/l, multiply by 2.266; to convert vitamin B-12 to pmol/l, multiply by PLP, pyridoxal phosphate; EGRac, erythrocyte glutathione reductase activation coefficient (a higher value indicates lower riboflavin status); Hcy, homocysteine. 2 Defined as a mean (6SD) of lmol/l. 3 Defined as a mean (6SD) of lmol/l. 4 Differences between groups were assessed by using an independent-samples t test. Data were log transformed before statistical analysis. 5 For all treatment groups combined, excluding placebo ). Although, the variability in the presence of unmetabolized folic acid does not appear to be entirely explained by folic acid intakes in healthy populations (32), clearly a lower target dose to deliver a beneficial effect will lower the risk of adverse effects (whatever the estimated size of such risk) in any emerging fortification policy. Previous studies examined homocysteine-lowering in response to different folic acid doses but the results are inconsistent (15, 33 35). One of these studies investigated the effect of folic acid in the range mg/d in IHD patients (15) and showed that homocysteine decreased in a dose-dependent manner with increasing doses of folic acid up to a maximum effect (ie, 23% reduction) in response to 0.8 mg folic acid/d. However, the study did not take into account initial homocysteine concentrations before randomization, which may have influenced the response. Another dose-finding study (35), which stratified subjects on the basis of pretreatment homocysteine, estimated that 0.4 mg folic acid/d was required to achieve 90% of the maximal homocysteine response. Both of the aforementioned studies, together with 23 others, were then subjected to a meta-analysis that standardized for pretreatment homocysteine (16) and concluded that 0.8 mg folic acid/d was required to maximally lower homocysteine concentrations, whereas a dose of 0.2 mg/d produced only 60% of this maximal effect (16). Few if any of the original studies included in the meta-analysis reported on participant compliance; however, poor compliance could underestimate the response to any given dose and, in turn, become problematic when the results are interpreted for fortification purposes. Because food fortification (unlike supplementation) automatically delivers maximal compliance, ensuring participant compliance was an important focus in our protocol, and the results indicate that this was generally very good. However, we consider the extended duration of treatment in this study to be the most critical factor explaining the difference between our results and earlier findings, concluding that the effective folic acid dose was 4 times higher than in this trial (15, 16). In the aforementioned meta-analysis (16), the treatment duration in 18 of 25 trials examined was 8 wk; 6 had a duration of 12 wk, and only 1 (a 24-wk trial) had a duration similar to that of the current investigation. When the effect of folic acid in our IHD patients FIGURE 4. Median homocysteine response to folic acid over 26 wk of intervention in patients with ischemic heart disease. A subsample of the participants also underwentsamplingat6or12wk(week6,n = 34; week 12, n = 72; week 26, n = 101; not the same participants). The homocysteine response was calculated as the posttreatment minus pretreatment value. See Table 2 for a statistical comparison of homocysteine responses between all participants (n = 172).

7 EFFECTS OF LONG-TERM FOLIC ACID INTERVENTION 17 sampled at intervals during the intervention was examined, the maximal homocysteine response appeared to be achieved as early as 6 wk in the 0.8-mg folic acid/d group and by 12 wk in the 0.4-mg/d group; however, in the 0.2-mg/d group, the response was suboptimal at both 6 and 12 wk compared with that achieved by 26 wk. Thus, the longer treatment duration was necessary for a more optimal homocysteine-lowering effect to be observed in response to a low folic acid dose. The finding in this study that 0.2 mg folic acid/d given for 26 wk was effective at lowering homocysteine is important for emerging folic acid fortification policy in different countries. This is because food fortification is untargeted, and ensuring that the effective dose is reached in a population inevitably means that some people will be exposed to much higher levels. A dose of 0.2 mg/d is similar to the estimated increment in folic acid intake in the United States (ie, lg/d), arising from the introduction of mandatory fortification in 1998 (36). Mandatory folic acid fortification was formally proposed in the United Kingdom in 2000 and again in 2006 after an extensive review of the available evidence (37); however, in 2007, implementation of a new policy was further postponed to allow consideration of new evidence in relation to potential adverse effects of folic acid and cancer risk a process just completed. The current findings suggest that the additional intake of 0.2 mg folic acid/d now being proposed for implementation (38), although primarily aimed at reducing neural tube defects (NTDs), is likely to also have other benefits. The extent of homocysteine-lowering shown here in response to 0.2 mg folic acid/d in participants with higher initial homocysteine was recently shown (by a metaanalysis of randomized trials of folic acid and stroke risk) to be associated with a 23% reduction in the relative risk of stroke (0.77; 95% CI: 0.63, 0.95; P = 0.012) (10). This is also the maternal folic acid intake associated with the lowest risk of NTDs (39, 40) and may, in addition, offer protection against congenital heart defects in infants (41). Apart from folic acid, the findings provide some indirect evidence of the potential benefits of enhancing vitamin B-12, the status of which was found to be significantly higher at baseline in those with lower initial homocysteine concentrations. Furthermore, postintervention homocysteine concentrations in participants with higher baseline homocysteine concentrations appeared to reach a plateau (irrespective of the folic acid dose) at 12.0 lmol/l, compared with a postintervention value of 9.0 lmol/l achieved in those with lower baseline homocysteine. Given that participants in both homocysteine categories were matched for age, it is unlikely that age-related renal impairment (known to be a major determinant of plasma homocysteine concentration) could have explained the difference in homocysteine responses between the 2 categories, although a limitation of the current study was that no measure of renal function was included. It is more likely that the extent of the homocysteine response in those with higher baseline concentrations was limited by vitamin B-12, which is also required for the metabolism of homocysteine through the methionine synthase pathway. Thus, although low-dose folic acid was found to be effective in this study, folic acid alone may not lower homocysteine to desirable concentrations (,10 lmol/l) (42) in those with higher initial homocysteine concentrations. Any small additional decrease in homocysteine from vitamin B-12 intervention could be predicted to confer a further benefit in terms of disease risk. Therefore, although not investigated here, some consideration could be given to the inclusion of vitamin B-12 together with folic acid in food fortification policies, particularly in light of new evidence that low maternal vitamin B-12 is a significant predictor (independent of folate) of NTD risk in women (43). In conclusion, these results show that a dose of folic acid as low as 0.2 mg/d can, if taken for 6 mo, effectively lower homocysteine concentrations. We showed that higher doses of folic acid may not be necessary and, in support of the recent opinion expressed elsewhere (44, 45), may be inappropriate given potential adverse effects of long-term exposure to high folic acid intakes. Previous trials may have overestimated the folic acid dose required because of too short an intervention period to observe a more complete response to lower doses. Thus, the potency of folic acid at low doses given chronically, as demonstrated in this study, should not be underestimated, and it is possible that doses lower than those investigated here would also be effective over a longer time period. Finally, folic acid alone may not lower homocysteine to desirable concentrations, and further research is required to determine whether the inclusion of vitamin B-12 will have benefits over and above the effect of folic acid in existing and emerging fortification policies. We thank all of the study participants and acknowledge the important contribution of nurses Joanne Glass and Donna Harris and the laboratory staff at Causeway Hospital, Coleraine. The authors responsibilities were as follows MW: guarantor; MW, HM, and JMS: planned and designed the study; PT and OF: collected the study data; PT, MW, HM, and JMS: analyzed and interpreted the data; PT, MW, and HM: drafted the manuscript; PT, MW, HM, JMS, AMM, JJS, OF, and AD: critically revised the manuscript for important intellectual content; PT, AD, AMM, and MD: conducted the statistical analysis; MW, HM, and JJS: obtained funding; MW, HM, AMM, JMS, JJS, OF, AD, and KP: provided administrative, technical, or material support; and MW, HM, and OF: supervised the study. All authors read and approved the final manuscript. None of the authors had a conflict of interest to report. REFERENCES 1. The Homocysteine Studies Collaboration. Homocysteine and risk of ischemic heart disease and stroke. JAMA 2002;288: Wald DS, Law M, Morris JK. Homocysteine and cardiovascular disease: evidence on causality from a meta-analysis. BMJ 2002;325: 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: Lonn E, Yusuf S, Arnold MJ, et al. The Heart Outcomes Prevention Evaluation 2 (HOPE-2) Investigators. Homocysteine lowering with folic acid and B vitamins in vascular disease. N Engl J Med 2006;354: Bønaa KH, Njølstad I, Ueland PM, et al. Homocysteine lowering and cardiovascular events after acute myocardial infarction. N Engl J Med 2006;354: Albert CM, Cook NR, Gaziano JM, et al. The NORVIT Trial Investigators. Effect of folic acid and B vitamins on risk of cardiovascular events and total mortality among women at high risk for cardiovascular disease. A randomized trial. JAMA 2008;299: Ebbing M, Bleie Ø, Ueland PM, et al. Mortality and cardiovascular events in patients treated with homocysteine-lowering B vitamins after coronary angiography: a randomized controlled trial. JAMA 2008;300: Study of the Effectiveness of Additional Reductions in Cholesterol and Homocysteine (SEARCH) Collaborative Group. Effects of homocysteine-lowering with folic acid plus vitamin B 12 vs placebo on mortality and major morbidity in myocardial infarction survivors. JAMA 2010;303: Saposnik G, Ray JG, Sheridan P, Sheridan P, McQueen M, Lonn E. The HOPE 2 Investigators. Homcysteine-lowering therapy and stroke risk,

8 18 TIGHE ET AL severity, and disability: additional findings from the HOPE 2 trial. Stroke 2009;40: Wang X, Qin X, Demirtas H, et al. Efficacy of folic acid supplementation in stroke prevention: a meta-analysis. Lancet 2007;369: Yang Q, Botto LD, Erickson JD, et al. Improvement in stroke mortality in Canada and the United States, 1990 to Circulation 2006;113: Homocysteine Lowering Trialists Collaboration. Lowering blood homocysteine with folic acid based supplements: meta-analysis of randomised trials. BMJ 1998;316: 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. QJM 1997;90: Daly S, Mills JL, Molloy AM, et al. Low-dose folic acid lowers plasma homocysteine levels in women of child-bearing age. QJM 2002;95: Wald DS, Bishop L, Wald NJ, et al. Randomized trial of folic acid supplementation and serum homocysteine levels. Arch Intern Med 2001; 161: Homocysteine Lowering Trialists Collaboration. Dose-dependent effects of folic acid on blood concentrations of homocysteine: a metaanalysis of the randomized trials. Am J Clin Nutr 2005;82: Reynolds E. Vitamin B12, folic acid, and the nervous system. Lancet Neurol 2006;5: Morris MS, Jacques PF, Rosenberg IH, Selhub J. Folate and vitamin B12 status in relation to anemia, macrocytosis, and cognitive impairment in older Americans in the age of folic acid fortification. Am J Clin Nutr 2007;85: Kim Y-I. Folate and carcinogenesis: evidence, mechanisms, and implications. J Nutr Biochem 1999;10: Ericson U, Sonestedt E, Gullberg B, Olsson H, Wirfalt E. High folate intake is associated with lower breast cancer incidence in postmenopausal women in the Malmo Diet and Cancer cohort. Am J Clin Nutr 2007;86: Cole BF, Baron JA, Sandler RS, et al, for the Polyp Prevention Study Group. Folic acid for the prevention of colorectal adenomas: a randomized clinical trial. JAMA 2007;297: Figueiredo JC, Grau MV, Haile RW, et al. Folic acid and risk of prostate cancer: results from a randomized clinical trial. J Natl Cancer Inst 2009; 101: Ebbing M, Bønaa KH, Nygård O, et al. Cancer incidence and mortality after treatment with folic acid and vitamin B12. JAMA 2009;302: Leino A. Fully automated measurement of total homocysteine in plasma and serum on the Abbott IMx analyzer. Clin Chem 1999;45: Molloy AM, Scott JM. Microbiological assay for serum, plasma, and red cell folate using cryopreserved, microtiter plate method. Methods Enzymol 1997;281: Kelleher BP, Broin SD. Microbiological assay for vitamin B12 performed in 96-well microtitre plates. J Clin Pathol 1991;44: Bates CJ, Pentieva KD, Prentice A, Mansoor MA, Finch S. Plasma pyridoxal phosphate and pyridoxic acid and their relationship to plasma homocysteine in a representative sample of British men and women aged 65 years and over. Br J Nutr 1999;81: Powers HJ, Bates CJ, Prentice AM, Lamb WH, Jepson M, Bowman H. The relative effectiveness of iron and iron with riboflavin in correcting a microcytic anaemia in men and children in rural Gambia. Hum Nutr Clin Nutr 1983;37: Frosst P, Blom HJ, Milos R, et al. A candidate genetic risk factor for vascular disease: a common mutation in methylenetetrahydrofolate reductase. Nat Genet 1995;10: Crawley H. Food portion sizes. 3rd ed. London, United Kingdom: Her Majesty s Stationery Office, Bailey SW, Ayling JE. The extremely slow and variable activity of dihydrofolate reductase in human liver and its implications for high folic acid intake. Proc Natl Acad Sci USA 2009;106: Bailey RL, Mills JL, Yetley EA, et al. Unmetabolized serum folic acid and its relation to folic acid intake from diet and supplements in a nationally representative sample of adults aged. or =60 y in the United States. Am J Clin Nutr 2010;92: de Bree A, Verschuren WM, Blom HJ, Kromhout D. Association between B vitamin intake and plasma homocysteine concentration in the general Dutch population aged y. Am J Clin Nutr 2001;73: Rydlewicz A, Simpson JA, Taylor RJ, Bond CM, Golden MH. The effect of folic acid supplementation on plasma homocysteine in an elderly population. QJM 2002;95: van Oort FV, Melse-Boonstra A, Brouwer IA, et al. Folic acid and reduction of plasma homocysteine concentrations in older adults: a dose-response study. Am J Clin Nutr 2003;77: Quinlivan EP, Gregory JF. Effect of food fortification on folic acid intake in the United States. Am J Clin Nutr 2003;77: Scientific Advisory Committee on Nutrition. Report on folate and disease prevention Available from: reports_position_statements (cited 10 December 2009). 38. Scientific Advisory Committee on Nutrition. Folic acid and colorectal cancer risk: review of recommendation for mandatory folic acid fortification. Available from: (cited 10 December 2009). 39. Daly LE, Kirke PN, Molloy A, Weir DG, Scott JM. Folate levels and neural tube defects. Implications for prevention. JAMA 1995;274: Daly S, Mills JL, Molloy AM, et al. Minimum effective dose of folic acid for food fortification to prevent neural-tube defects. Lancet 1997; 350: van Beynum IM, Kapusta L, Bakker MK, den Heijer M, Blom HJ, de Walle HE. Protective effect of periconceptional folic acid supplements on the risk of congenital heart defects: a registry-based case-control study in the northern Netherlands. Eur Heart J 2010;31: Malinow MR, Bostom AG, Krauss RM. Homocyst(e)ine, diet, and cardiovascular diseases: a statement for healthcare professionals from the Nutrition Committee, American Heart Association. Circulation 1999;99: Molloy AM, Kirke PN, Troendle JF, et al. Maternal vitamin B12 status and risk of neural tube defects in a population with high neural tube defects and no folic acid supplementation. Pediatrics 2009;123: Heseker HB, Mason JB, Selhub J, Rosenberg IH, Jacques PF. Not all cases of neural-tube defect can be prevented by increasing the intake of folic acid. Br J Nutr 2009;102: Miller JW, Garrod MG, Allen LH, Haan MN, Green R. Metabolic evidence of vitamin B-12 deficiency, including high homocysteine and methylmalonic acid and low holotranscobalamin, is more pronounced in older adults with elevated plasma folate. Am J Clin Nutr 2009;90: