Vitamin Intake and Risk of Coronary Disease: Observation versus Intervention Caroline Moats, MS, and Eric B. Rimm, ScD

Corresponding author Eric B. Rimm, ScD Department of Nutrition, Harvard School of Public Health, 665 Huntington Avenue, Boston, MA 02115, USA. E-mail: [email protected] Current Atherosclerosis Reports 2007, 9:508–514 Current Medicine Group LLC ISSN 1523-3804 Copyright © 2007 by Current Medicine Group LLC

The evidence that specific vitamins may be beneficial in the prevention of cardiovascular disease (CVD) is supported by mechanistic models of atherogenesis. We and others have published observational epidemiologic studies in support of vitamins in the primary prevention of CVD, but the results from intervention studies are mixed. This article summarizes the recent results for vitamin E, vitamin D, and the B vitamins, comparing study populations, study designs, and potential methodologic reasons for differences in findings. For vitamin E, observational data suggest benefit at doses of 100 to 400 IU/d. Results from recent large-scale trials are mixed, with some showing modest benefit but others suggesting no benefit, especially for secondary prevention. Results for B vitamins are also mixed and further complicated by the recent folate fortification of the flour supply. If greater B vitamin intake does reduce CVD, the benefits are likely to be greatest for primary prevention and in populations with intake below dietary reference standards. Research on vitamin D and CVD is just beginning to emerge, but current data suggest that if there is benefit it likely needs to be at intake levels much higher than the current reference intakes of 200 to 600 IU/d for American adults.

Introduction Dietary supplements are estimated to be more than a $20 billion industry in the United States [1]. Consumer spending on supplements nearly doubled from 1994 to 2000, and it continues to grow more than 10% a year. The reasons for taking supplements are varied, but for many it is the perception that they can treat or prevent chronic disease [2]. In the 1999 to 2000 National Health and Examination Survey [3], 52% of adults reported taking a dietary supplement in the past month, up from the 40% reported in the same survey conducted 10 years earlier [4].

In the late 1980s and early 1990s, several articles were published suggesting that higher intake of specific micronutrients, usually achieved through long-term supplementation, leads to lower risk of several chronic diseases, including prostate cancer, colon cancer, and coronary heart disease (CHD) [5–11]. We cannot discuss each of these chronic disease endpoints within the space of this article, but many of the methodologic issues are the same. To help focus this review, we limit the discussion to vitamin E, folate (B vitamins), and vitamin D and their effect on risk of cardiovascular disease (CVD) because results from large-scale observational and intervention studies of these supplements are available. Vitamins C and K also have a growing body of evidence to suggest direct benefits, but they are not discussed in detail here. We precede the summary of the literature on vitamins and CVD with a brief discussion of the different scenarios that may explain the apparent discrepancies between results from observational and intervention studies.

Methodologic Issues Early observational studies reported an inverse association between vitamins and CVD, usually at doses well above the recommended daily allowance [5–7,10]. Subsequently, many large intervention studies have found little or no effect [12,13•,14,15,16•,17,18], and more recently a meta-analysis has linked high-dose vitamin use to a modest increase in mortality [19]. Before we review the results for specific vitamins, we discuss potential methodologic differences that may explain discrepant findings. The most frequently cited explanation for differences between observational and intervention studies is that purported benefits of vitamin supplementation are confounded by healthier lifestyles among participants with higher vitamin intake. For example, in the Nurses’ Health Study, we [7] reported that women in the highest quintile of both folate and vitamin B6 intake had a 45% reduction in risk of CHD compared with women with low intake of both vitamins. Some have hypothesized that this benefit is not due to B vitamin intake but rather to differences in other factors, such as a better diet, exercise, social status, and less smoking among the women with higher intake. If this inverse association was solely due to residual confounding, then all micronutrients associ-

Vitamin Intake and Risk of Coronary Disease  Moats and Rimm  509

ated with a healthier lifestyle should be associated with lower risk, but this was not the case in our study or in several others. Furthermore, accounting for many other important lifestyle characteristics and known CHD risk factors only modestly attenuated the results. Thus, for further confounding to explain the inverse association, some other set of unknown factors would need to be very strongly associated with vitamin intake and also strongly inversely associated with CHD [20]. This is possible, but it is very unlikely. Also, for the vitamins we discuss in the following text, there is general biologic plausibility, supportive animal models, a dose response for benefit, and reproducibility across populations and across dietary or supplement sources. Thus, it makes it highly unlikely that this first scenario (ie, vitamins have no benefit in the primary prevention of CVD) explains the discrepancy in the published literature. A second scenario is that vitamins provide a benefit in CVD prevention, but the pertinent etiologic period is not always captured within a study design or study population, and thus results are inconsistent across studies. The development of atherosclerosis and CVD is a multifaceted progression of biologic processes that likely include hyperlipidemia, endothelial dysfunction, oxidation, inflammation, coagulation, and other pathophysiologic phenomena. It seems biologically implausible that a single vitamin should be instrumental across all etiologic pathways from early fatty streak production to the ultimate clinical event, sometimes occurring 60 to 70 years later. Recent large, randomized, secondary prevention studies have lasted 2 to 7 years [21–23]. This short time-frame may be too late in the disease progression to play a meaningful role. Some have argued against this because the design of secondary prevention trials has been an effective tool to test for CVD prevention from lipid-lowering drugs. However, hypothesized mechanisms for lipid-lowering medications are very different than the mechanisms for vitamins. A third (and related) scenario that could explain the differences between observational and intervention results is the large differences in patient populations and the highly selective inclusion criteria. Typically, participants in secondary prevention trials are already managed well medically, to a point where vitamin supplementation may not provide additional benefit. For example, in the Heart Outcomes Prevention Evaluation (HOPE) 2 trial [13•], a secondary prevention study of 5522 patients 55 years of age and older randomized to 2.5 mg of folic acid, 50 mg of vitamin B6, and 1 mg of vitamin B12 , baseline use of aspirin or an antiplatelet agent was reported by 78% of the treatment group, use of β-blockers was reported by 46%, and use of angiotensin-converting enzyme inhibitors was reported by 65.9%. In the Norwegian Vitamin Trial (NORVIT) [15], a parallel 4-year secondary prevention study of B vitamins and coronary disease, baseline use of statins, antihypertensives, and aspirin ranged from

80% to 92% of participants. Even if vitamins are beneficial in these selective patient populations, at higher doses of vitamins, the benefits of supplements may be offset by a modest adverse interaction with existing medications, as has been previously suggested for statins and vitamin E [24]. Thus, the results from trials of medically managed patients are still valid, but they may not be generalizable to the primary prevention setting for CVD. Observational studies can also be modestly selective. Initial response rates frequently range from 25% to 75%, and the general distribution of characteristics of participants is somewhat healthier than nonparticipants. This too can impact the generalizability of results. In intervention studies, the inclusion criteria are usually so highly selective with respect to the baseline health status and other study requirements that frequently only 1% to 20% of the patient pool is recruited and eventually randomized to treatment. The generalizability and validity of the results may further be compromised by subsequent compliance because lower compliance can bias results toward not finding a difference in effect between intervention and placebo. The highly selective nature of clinical studies may help to explain why clinical trials themselves find discrepant results [25–32]. For example, in trials of atherosclerosis progression and CVD events, the Vitamin E Atherosclerosis Prevention Study (VEAPS) [17], which randomized only 17.3% of those initially screened, found that vitamin E (400 IU/d) did not slow the progression of subclinical atherosclerosis in hypercholesterolemic men and women. However, the Antioxidant Supplement and Atherosclerosis Prevention (ASAP) study [33] did find that a supplement containing vitamin E (approximately 250 IU/d) slowed the progression of atherosclerosis in hypercholesterolemic men. Because both are trials, differences are not likely due to confounding, as both studies had well-balanced baseline characteristics. Thus, it is likely that subtle differences in assessment of atherosclerosis progression, length of follow-up, vitamin formulations, or differences in the selective nature of the baseline population explain the discrepant results. Ultimately, the true differences in results between studies (regardless of design) are likely due to a combination of these various scenarios. Thus, in our discussions of the substantive findings in the following text, we refer back to the methodologic issues just described.

Antioxidants As with many chronic diseases, oxidative damage likely plays a crucial role in the development and progression of CVD. Few argue against this theory, first described broadly after discoveries by Brown and Goldstein [34], that macrophage selection uptake oxidatively modified low-density lipoprotein through the scavenger receptor. Steinberg et al. [35] later described in more detail the implications of oxidation on atherogenesis. The epidemio-

510  Nutrition

logic literature has been extensive in this area, starting decades ago with studies of beta-carotene intake, then vitamin C, and more recently vitamin E. The large-scale primary and secondary prevention trials of beta-carotene did not show benefit and are reviewed elsewhere. The data on vitamin C are also not reviewed here, but in general results from observational studies are equivocal and only one long-term trial of CHD has been conducted with vitamin C as a single supplement. It is likely that if vitamin C reduces CVD, it is most beneficial for populations with very low dietary intake. Hundreds of other dietary constituents have antioxidant properties but are beyond the scope of this review of vitamins. Thus, we focus specifically on vitamin E because the data are voluminous and the methodologic issues most interesting. In 1993, we [5,6] found an inverse association between supplemental vitamin E and a risk of CHD in men and women in two large, ongoing observational studies. In both studies, we found that men and women who reported taking 100 IU/d or more of vitamin E for 2 or more years had a 40% lower risk of developing CHD, even after controlling for classic coronary risk factors. As discussed earlier, it is possible that other unknown or unmeasured confounders explained this strong inverse association. Interestingly, in the Health Professionals Follow-up Study, we also looked simultaneously at the association for vitamin C because it too was associated with a healthier lifestyle [5,6]. After adjustment for vitamin E, however, vitamin C was not associated with a reduction in coronary disease. This result supports a specific effect of vitamin E, at least within this population that is free of pre-existing CHD [5,6]. Since these observational studies, numerous clinical and observational trials have been conducted to analyze the relationship between antioxidants and CVD. In the HOPE study [12], the first large-scale, multicountry, double-blind, randomized clinical trial, investigators evaluated high-dose (400 IU/d) vitamin E in a 5-year randomized 2 x 2 factorial study of 9541 high-risk patients. Vitamin E did not reduce the incidence of cardiovascular events compared with those receiving placebo [12]. In HOPE-TOO [22], a subset of the initial HOPE trial, 3994 patients agreed to continue the study intervention. After 7 years of total follow-up, there were no significant differences for major cardiovascular events between the vitamin E and placebo groups, although patients on vitamin E had modestly higher risk of heart failure (RR of 1.13; 95% CI, 1.01–1.26) and hospitalization for heart failure (RR of 1.21; 95% CI, 1.00–147) [22]. These two important studies provide strong evidence that high-dose vitamin E among at-risk patients (with an average age of 66 years) does not lower incidence of CVD. However, there is modest concern, as discussed previously, that these studies do not address the pertinent etiologic time period for benefit, and that secondary prevention with vitamin E among patients with presumably extensive

disease may not test the hypothesis that antioxidants prevent the incidence of primary CVD. Interestingly, in the most recently published secondary prevention trial for vitamin E, investigators in the Women’s Antioxidant Cardiovascular Study (WACS) [36•] randomized 8171 female health professionals to a 2 x 2 factorial design that included 600 IU of vitamin E every other day for 9.4 years. The women were modestly younger (60 years on average) than those in the HOPE and HOPE-TOO trials. Although the overall risk of major CVD was not lower (RR of 0.94; 95% CI, 0.85–1.04) among the 74% who were compliant to the active vitamin E intervention over the period of follow-up, the risk of major CVD (RR of 0.87; 95% CI, 0.76–0.99) and many of the secondary endpoints was significantly reduced. Only about 25% of the subjects in the HOPE studies were woman, and the studies did not report sex-specific relative risks. Therefore, it is difficult to determine if the differences in the results of these trials are due to chance, a true sex difference, a difference in age at recruitment, or the other differences in the inclusion characteristics of each population. The best test of vitamin E in the primary prevention of CVD is from the Women’s Health Study (WHS) [16•], a 2 x 2 factorial design study of aspirin and vitamin E among 39,876 women with an average age of 55 years who were followed for 10 years. Women were randomized to 600 IU of natural-source (typically derived from vegetable oils) vitamin E every other day. The vitamin E was taken on opposite days from the aspirin/placebo intervention. Because vitamin E is fat soluble, every other day is sufficient to sustain blood levels two- to threefold higher than when on placebo. For the primary CVD endpoint under study, the authors reported a nonsignificant 7% reduction in major cardiovascular events (RR of 0.93; 95% CI, 0.82–1.05). The study also found no significant effect of vitamin E on the incidence of myocardial infarctions (RR of 1.01; 95% CI, 0.82–1.23), but there was a significant 24% reduction in cardiovascular mortality (RR of 0.76; 95% CI, 0.59–0.98). Further, when the main outcome of major cardiovascular events was limited to women greater than 65 years of age at baseline, vitamin E supplementation was significantly inversely associated (RR of 0.74; 95% CI, 0.59–0.93) with risk. In this cohort, the incidence of primary CHD events was four times higher among women 65 years and older than among younger women. It cannot be determined from the existing epidemiologic and biologic literature whether this group of women at much higher risk of a primary CVD event would benefit more from intervention. If these subanalyses showing benefit for certain groups and endpoints in the WHS and the WACS are real, how can we explain the differences between these studies and the secondary prevention trials discussed here (ie, HOPE, HOPE-TOO)? Because these studies are all clinical trials, it is very unlikely that confounding by some unbalanced factor explains the difference. Chance is always an

Vitamin Intake and Risk of Coronary Disease  Moats and Rimm  511

explanation for findings from subgroup analyses. Several different prespecified endpoints were examined in the WHS, which increases the likelihood that a relative risk was significant by chance. However, the WACS study found many statistically significant findings for benefit from vitamin E among the subset of compliant subjects. If the protective effect of intervention was solely due to chance, we would expect vitamin E also to increase risk in other subanalyses, but none were reported. It is more plausible that differences between study populations explain differences in trial results. As discussed previously, in both primary and secondary prevention trials of vitamin E, stringent selection criteria affect participation. In true secondary prevention trials of CVD, participants must survive their first CVD event. This in itself may lead to a selection of individuals who are less susceptible to the benefits of vitamins. However, in primary prevention studies where follow-up needs to be sufficiently long to accrue endpoints, recruitment can sometimes be very difficult because participants must agree to stay on intervention for the length of the study. In the WHS, only 2.3% of the women initially contacted were eventually randomized to intervention [16•]. Thus, primary or secondary intervention studies may be too highly selective, so much so that results may be less generalizable, and in the case of primary versus secondary prevention with vitamin E, even results from comparable trials do not agree with each other. Recently, a systematic meta-analysis of 68 randomized trials of antioxidant supplements did not find substantial effects of supplements on mortality [19]. When the authors limited the analyses to the 47 studies that they deemed properly conducted and with the least methodologic biases, the antioxidants beta-carotene, vitamin A, and vitamin E were modestly positively associated with risk. The authors only assessed all-cause mortality and did not present CVD-specific analyses. The results are very difficult to interpret because most studies were conducted among highly selective, at-risk populations. In many cases, participants were selected because of a history of cancer, eye disease, or reduced liver function, or because they were elderly. Furthermore, the meta-analysis did not adequately take into account dose or duration. Thus, it is difficult to use a single summary statistic from this metaanalysis to evaluate the effects of vitamin E or any other antioxidants on cardiovascular disease. The effects of vitamin E from observational and intervention studies are mixed. Although most large secondary prevention studies show no benefit for vitamin E supplementation, the data from observational studies among men and women and data from subgroup analyses from large primary prevention and secondary prevention intervention trials among women suggest that benefit may be achieved at moderate doses over a long period of time.

B Vitamins The study of B vitamins in the prevention of CVD derives from the initial observations of premature atherosclerosis in adolescents with homocysteinuria caused by a rare genetic disorder in the cystathione-β-synthase gene. Over the past several decades, numerous studies have found a modest elevation in CVD risk associated with blood homocysteine levels in the normal range, well below those initially found in the adolescent patients [37,38]. Because folate and vitamins B6 and B12 are important direct or indirect co-factors in the metabolism of homocysteine and because supplements of these vitamins have been shown to lower plasma homocysteine, a combination of these vitamins could be instrumental in CVD prevention. It is important to note that this family of vitamins is involved in many metabolic pathways and could also affect CVD risk through other mechanisms not related to homocysteine. In 1998, we [7] examined the relation of intakes of folate and vitamin B6 and 14-year risk of CHD among 80,082 women enrolled in the Nurses’ Health Study without previous CVD. Compared with women in the lowest quintile of both folate and vitamin B6 intake, the relative risk of CHD for women in the highest quintile of intake was 0.55 (95% CI, 0.41–0.74). These first results from a large, prospective observational study of diet provided evidence that folate and B6 intake may be important in the prevention of CHD. Subsequently, several large primary and secondary intervention studies were initiated to test the hypothesis that folate alone or in combination with other B vitamins could prevent CVD. However, concurrently, in the United States and in many other countries worldwide, folate fortification was mandated to reduce neural tube defect in the offspring of mothers with substandard folate intake. The law went into effect in 1998 in North America and significantly reduced the prevalence of low folate and high plasma homocysteine levels in middleaged individuals [18,39]. Post-fortification, folate intake increased by at least 100 to 200 mg/d on a population level; thus, it became more challenging to identify high-risk patients with elevated homocysteine for secondary prevention trials. For example, in the Vitamin Intervention for Stroke Prevention (VISP) study [18], a secondary prevention trial of 3680 stroke patients from North America, the threshold for the inclusion criteria of “elevated homocysteine” needed to be lowered twice because the recruitment time-frame paralleled the mandated fortification. The difference in homocysteine levels between the high-dose and low-dose intervention arms was substantially smaller at the end of the trial compared with the beginning, further suggesting that folate fortification across the population diluted the effective difference between treatment arms. In the vernacular of trialists, this would be a modest folate “drop-in” in 100% of the participants in the placebo arm. In general, “drop-ins” reduce the difference in exposure between the intervention and placebo arm and bias the results toward the null. After 2 years of follow-up in the

512  Nutrition

VISP study, the relative risk of ischemic stroke was 1.0 (95% CI, 0.8–1.3), with virtually identical Kaplan-Meier survival curves between patients randomized to high-dose versus low-dose B vitamins. Three other large secondary prevention trials also found no benefit of supplementation among patients with existing CVD or at high risk for CVD [13•,15,36•]. In HOPE 2, patients with prior vascular disease or diabetes were assigned to treatment with 2.5 mg/d of folic acid, 50 mg/d of vitamin B6, and 1 mg/d of vitamin B12 or a placebo [13•]. The mean duration of treatment was 5 years. The authors reported no benefit in the primary outcomes of death from CVD. They did report significantly lower risk of stroke (mostly due to ischemic stroke) but adverse risk of angina. Approximately 72% of the patients were in countries with folate fortification; thus, as with VISP, it is possible that there were too few patients with substantially increased homocysteine, which would be the population most likely to benefit from folate supplementation [40]. We have insufficient space to cover all of the recent studies on B vitamin intake or homocysteine and CVD. Several recent meta-analyses have summarized past results and in general have found that the effect of homocysteine on CVD risk, if it exists at all, is modest at best. For example, in a recent meta-analysis of 12 homocysteinelowering trials, the authors concluded that in secondary prevention the strength of the benefit of homocysteine lowering with risk of CVD may not be as strong as previously believed [41]. In the review, seven of the trials included patients with prior CHD, three included patients with prior stroke, and two included patients with prior renal disease. As discussed earlier, these studies may not address the pertinent etiologic time period. Further, the median duration of treatment for the selected trials ranged from 2 to 7.4 years. In all cases the intervention lowered homocysteine, but approximately half the trials were conducted in populations with folate fortification. Thus, the difference in exposure between intervention and placebo would be lessened and true hyperhomocysteinemia would be rare. In a meta-analysis on the efficacy of folate supplementation specifically for stroke prevention [42•], the authors highlighted populations where greatest benefit could be achieved. In the eight randomized trials summarized, folate intervention significantly reduced risk of stroke by 18% (95% CI, 0%–32%). The risk reduction (percent reduction) was only in strata defined by supplementation longer than 3 years (29%), effective homocysteine lowering by 20% or more (23%), in countries without grain fortification (25%), and in patients without previous stroke (25%). One final concern that has arisen that could also explain discrepancies in studies (regardless of design) is the wide range of dosages of B vitamins. These have ranged from 0.5 to 40.0 mg/d for folic acid, 0.02 to 1.0 mg/d for vitamin B12 , and 3 to 100 mg/d for vitamin B6. For folate, this is up to 250 times the recommended daily

allowance of 400 mg/d, and in many cases above the upper limit of 1 mg/d for folic acid supplementation recommended by the US Dietary Reference Intakes [43]. Excess folate may lead to circulation of unmetabolized folic acid in plasma, which has been shown to reduce immune function [44]. The effects on coronary disease progression are unknown. It is possible that the modest benefit of homocysteine lowering is offset by systemic complications from excess circulating unmetabolized folate. In the future, this should be an active area of research because folate fortification is now worldwide and the number of people with folate intake in excess of the upper limit is growing. Further complicating the effects of excess folic acid supplementation is the metabolic interaction folate may have with alcohol consumption [7,45]. Future primary and secondary prevention studies of folate should consider the effects of folate stratified by alcohol consumption. In summary, if there is benefit from greater folate intake (or other B vitamins) on CVD, the overall benefit is likely to be modest and may only be seen in patients without prior CVD or with intake well below the US Estimated Average Requirement of 320 mg/d.

Vitamin D Vitamin D has recently gained attention because insufficient dietary levels have been linked with cancer, multiple sclerosis, Parkinson’s disease, and CVD [46–48]. In most populations, sensible sun exposure is the main source of vitamin D for both children and adults. Interestingly, the rate of CVD death is higher at higher latitudes and also increases during the winter months. As has been noted previously by others, this pattern is consistent with an adverse effect of hypovitaminosis D, which is more prevalent in higher latitudes and during the winter [49,50]. In the liver, vitamin D is metabolized by a hepatic hydroxylase into 25-hydroxyvitamin D and then in the kidney by a renal 1α-hydroxylase into the active vitamin D hormone calcitriol. Calcitriol regulates systemic calcium and calcium resorption in bone and reabsorption in the kidney. Parathyroid hormone controls this conversion into calcitriol, and vitamin D deficiency can lead to severe secondary hyperparathyroidsism [51]. Research in patients with end-stage renal disease has shown that severe secondary hyperparathyroidism from vitamin D deficiency increases blood pressure and cardiac contractility and can lead to cardiomyocyte hypertrophy [51]. To date, there have been no large-scale, prospective observational studies of vitamin D intake or blood levels and risk of CVD. However, the Women’s Health Initiative trial [40] evaluated the risk of coronary and cerebrovascular events associated with modest vitamin D supplementation. The 36,282 postmenopausal women in the study were randomized to calcium (1000 mg/d) and vitamin D3 (400 IU/d) or to placebo. The calcium/vitamin D supplementation had no substantial effect on CHD or

Vitamin Intake and Risk of Coronary Disease  Moats and Rimm  513

stroke over 7 years of treatment in healthy postmenopausal women [40]. The strengths of this study are the size and the double-blind randomization of treatment. However, the dose of vitamin D was relatively low (200 IU twice daily) and the compliance was poor (60% of study participants took at least 80% of their study medication through year 6). The influence of the treatment on the 25-hydroxyvitamin D levels was not reported, but based on the dose and compliance, it is estimated to cause only a 2-ng/mL increase in blood levels, a difference likely not to cause appreciable change in CVD risk [52]. Finally, the trial was designed to evaluate the effect of vitamin D intervention on fractures and not CVD [40]. As such, additional data in observational and intervention studies are needed to explore the importance of vitamin D on incidence and progression of CVD.

Conclusions Results from observational and intervention studies for these vitamins that show the greatest promise in CVD prevention have been mixed and inconclusive. Some of the differences in results between studies, regardless of design, may be due to methodologic limitations (ie, confounding in observational studies or generalizability in trials). The potential problem of confounding in observational studies has been well discussed elsewhere. However, it has not been previously adequately noted that in most clinical trials for vitamins, participation is very highly selective and most participants are already sufficiently medically managed such that vitamins may not further lower risk. In the future, research on vitamins may be most fruitful if conducted in studies of the primary prevention of CVD, specifically in populations with insufficient intake. Finally, it is important to note that vitamin supplements cannot replace a healthy lifestyle. A healthy diet, regular exercise, weight maintenance, and not smoking are significant factors in reducing incidence of CVD [53]. Public health programs, policy, and clinical advice should strongly reinforce these primary prevention areas before using vitamin supplements as a sole “therapy” to reduce coronary disease.

References and Recommended Reading Papers of particular interest, published recently, have been highlighted as: • Of importance •• Of major importance 1. 2. 3.

Fraunfelder FW: The science and marketing of dietary supplements. Am J Ophthalmol 2005, 140:302–304. US Food and Drug Administration: Dietary supplement enforcement report. Available at http://www.fda.gov/oc/ nutritioninitiative/report.html. Accessed May 5, 2007. Radimer K, Bindewald B, Hughes J, et al.: Dietary supplement use by US adults: data from the National Examination Survey. Am J Epidemiol 2004, 160:339–349.

4.

Erwin RB, Wright JD, Reed-Gillette D: Prevalence of Leading Types of Dietary Supplements Used in the Third National Health and Nutrition Examination Survey, 1988-94. Advance Data from Vital and Health Statistics; Number 349. Hyattsville, Maryland: National Cener for Health Statistics; 2004. 5. Rimm EB, Stampfer MJ, Ascherio A, et al.: Vitamin E consumption and the risk of coronary heart disease in men. N Engl J Med 1993, 328:1450–1456. 6. Stampfer MJ, Hennekens CH, Manson JE, et al.: Vitamin E consumption and the risk of coronary disease in women. N Engl J Med 1993, 328:1444–1449. 7. Rimm EB, Willett WC, Hu FB, et al.: Folate and vitamin B6 from diet and supplements in relation to risk of coronary heart disease among women. JAMA 1998, 279:359–364. 8. Giovannucci E: Tomatoes, tomato-based products, lycopene, and cancer: review of the epidemiologic literature. J Natl Cancer Inst 1999, 91:317–331. 9. Giovannucci E, Stampfer MJ, Colditz GA, et al.: Multivitamin use, folate, and colon cancer in women in the Nurses’ Health Study. Ann Intern Med 1998, 129:517–524. 10. He K, Merchant A, Rimm EB, et al.: Folate, vitamin B6, and B12 intakes in relation to risk of stroke among men. Stroke 2004, 35:169–174. 11. Kushi LH, Fee RM, Sellers TA, et al.: Intake of vitamin A, C, and E and postmenopausal breast cancer. Am J Epidemiol 1996, 144:165–174. 12. Yusuf S, Dagenais G, Pogue J, et al.: Vitamin E supplementation and cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. N Engl J Med 2000, 342:154–160. 13.• The Heart Outcomes Prevention Evaluation (HOPE) 2 Investigators: Homocysteine lowering with folic acid and B vitamins in vascular disease. N Engl J Med 2006, 354:1567–1577. This is a multicountry, long-term, secondary prevention trial that found no benefit for B vitamin supplementation on CVD. 14. The HOPE and HOPE-TOO Trial Investigators: Effects of long-term vitamin E supplementation on cardiovascular events and cancer—a randomized controlled trial. JAMA 2005, 293:1338–1347. 15. Bonaa K, Njolstad I, Magne-Ueland P, et al.: Homocysteine lowering and cardiovascular events after acute myocardial infarction. N Engl J Med 2006, 354:1578–1588. 16.• Lee IM, Cook NR, Gaziano JM, et al.: Vitamin E in the primary prevention of cardiovascular disease and cancer: the Women’s Health Study: a randomized controlled trial. JAMA 2005, 294:56–65. This is the first large-scale, long-term, primary prevention study of vitamin E. The investigators found no overall benefit but did report reduction in women over 65 years of age and for overall CVD mortality. 17. Hodis HN, Mack WJ, LaBree L, et al.: Alpha-tocopherol supplementation in healthy individuals reduces low-density lipoprotein oxidation but not atherosclerosis. The Vitamin E Atherosclerosis Prevention Study (VEAPS). Circulation 2002, 106:1453–1459. 18. 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, 291:565–575. 19. Bjelakovic G, Nikolova D, Gluud LL: Mortality in randomized trials of antioxidant supplements for primary and secondary prevention. JAMA 2007, 297:842–857. 20. Rothman KJ, Greenland S: Modern Epidemiology. Philadelphia: Lippincott-Raven Publishers; 1998. 21. Goodman GE, Thornquist MD, Balmes J, et al.: The Beta-Carotene and Retinol Efficacy Trial: incidence of lung cancer and cardiovascular disease mortality during 6-year follow-up after stopping beta-carotene and retinol supplements. J Natl Cancer Inst 2004, 96:1742–1750.

514  Nutrition Lonn E, Bosch J, Yusuf S, et al.: Effects of long-term vitamin E supplementation on cardiovascular events and cancer: a randomized controlled trial. JAMA 2005, 293:1338–1347. 23. Waters DD, Alderman EL, Hsia J, et al.: Effects of hormone replacement therapy and antioxidant vitamin supplements on coronary atherosclerosis in post-menopausal women: a randomized controlled trial. JAMA 2002, 288:2432–2440. 24. Brown BG, Zhao XQ, Chait A, et al.: Simvastatin and niacin, antioxidant vitamins, or the combination for the prevention of coronary disease. N Engl J Med 2001, 345:1583–1592. 25. Rapola JM, Virtamo J, Ripatti S, et al.: Randomised trial of alpha-tocopherol and beta-carotene supplements on incidence of major coronary events in men with previous myocardial infarction. Lancet 1997, 349:1715–1720. 26. Virtamo J, Rapola JM, Ripatti S, et al.: Effect of vitamin E and beta carotene on the incidence of primary nonfatal myocardial infarction and fatal coronary heart disease. Arch Intern Med 1998, 158:668–675. 27. Stephens NG, Parsons A, Schofield PM, et al.: Randomised controlled trial of vitamin E in patients with coronary disease: Cambridge Heart Antioxidant Study (CHAOS). Lancet 1996, 347:781–786. 28. GISSI-Prevenzione Investigators: Dietary supplementation with n-3 polyunsaturated fatty acids and vitamin E after myocardial infarction: results of the GISSI-Prevenzione trial. Lancet 1999, 354:447–455. 29. The Heart Outcomes Prevention Evaluation Study Investigators: Vitamin E supplementation and cardiovascular events in high-risk patients. N Engl J Med 2000, 342:154–160. 30. Collaborative Group of the Primary Prevention Project: Low-dose aspirin and vitamin E in people at cardiovascular risk: a randomised trial in general practice. Collaborative Group of the Primary Prevention Project. Lancet 2001, 357:89–95. 31. Boaz M, Smetana S, Weinstein T, et al.: Secondary prevention with antioxidants of cardiovascular disease in endstage renal disease (SPACE): randomised placebo-controlled trial. Lancet 2000, 356:1213–1218. 32. Blot WJ, Li JY, Taylor PR, et al.: Nutrition intervention trials in Linxian, China: supplementation with specific vitamin/mineral combinations, cancer incidence, and disease-specific mortality in the general population. J Natl Cancer Inst 1993, 85:1483–1492. 33. Salonen JT, Nyyssonen K, Salonen R, et al.: Antioxidant Supplementation in Atherosclerosis Prevention (ASAP) study: a randomized trial of the effect of vitamins E and C on 3-year progression of carotid atherosclerosis. J Intern Med 2000, 248:377–386. 34. Brown MS, Goldstein JL: Lipoprotein metabolism in the macrophage: implications for cholesterol deposition in atherosclerosis. Annu Rev Biochem 1983, 52:223–261. 35. Steinberg D, Pathasarathy S, Carew TE, et al.: Beyond cholesterol. Modifications of low-density lipoprotein that increase its atherogenicity. N Engl J Med 1989, 320:915–924. 36.• Cook NR, Albert CM, Gaziano JM, et al.: A randomized factorial trial of vitamins C and E and beta carotene in the secondary prevention of cardiovascular events in women: Results from the Women’s Antioxidant Cardiovascular Study. Arch Intern Med 2007, 167:1610–1618. A large secondary prevention study of antioxidants that found no overall CVD benefit from any antioxidant supplement but did report 25% reduction in CVD among the 74% of compliant subjects. 37. The Homocysteine Studies Collaboration: Homocysteine and risk of ischemic heart disease and stroke: a meta-analysis. JAMA 2002, 288:2015–2022. 22.

Wald DS, Law M, Morris JK: Homocysteine and cardiovascular disease: evidence on causality from a meta-analysis. BMJ 2002, 325:1202. 39. Jacques PF, Selhub J, Bostom AG, et al.: The effect of folic acid fortification on plasma folate and total homocysteine concentrations. N Engl J Med 1999, 340:1449–1454. 40. Hsia J, Heiss G, Hong R: Calcium/vitamin D supplementation and cardiovascular events. Circulation 2007, 115:846–854. 41. B-Vitamin Treatment Trialists’ Collaboration: Homocysteine-lowering trials for prevention of cardiovascular events: a review of the design and power of the large randomized trials. Am Heart J 2006, 151:282–287. 42.• Wang X, Qin X, Demirtas H, et al.: Efficacy of folic acid supplmentation in stroke prevention: a meta-analysis. Lancet 2007, 369:1876–1882. This is a meta-analysis of folate supplementation and stroke that found approximately 20% reduction in CVD with intervention. Benefit was greatest in countries that did not have existing folate fortification. 43. Food and Nutrition Board, Institute of Medicine: Dietary Reference Intakes for Thiamin, Riboflavin, Niacin, Vitamin B6, Folate, Vitamin B12, Pantothenic Acid, Biotin, and Choline—a Report of the Standing Committee on the Scientific Evaluation of Dietary Reference Intakes and Its Panel on Folate, Other B Vitamins, and Choline and Subcommittee on Upper Reference Levels of Nutrients. Washington, DC: The National Academies Press; 2000. 44. Troen AM, Mitchell B, Sorensen B, et al.: Unmetabolized folic acid in plasma is associated with reduced natural killer cell cytotoxicity among postmenopausal women. J Nutr 2006, 136:189–194. 45. Chiuve SE, Giovannucci EL, Hankinson SE, et al.: Alcohol intake and methylenetetrahydrofolate reductase polymorphism modify the relation of folate intake to plasma homocysteine. Am J Clin Nutr 2005, 82:155–162. 46. Giovannucci E, Liu Y, Willett WC: Cancer incidence and mortality and vitamin D in black and white male health professionals. Cancer Epidemiol Biomarkers Prev 2006, 15:2467–2472. 47. Giovannucci E, Liu Y, Rimm EB, et al.: Prospective study of predictors of vitamin D status and cancer incidence and mortality in men. J Natl Cancer Inst 2006, 98:451–459. 48. Munger KL, Levin LI, Hollis BW, et al.: Serum 25hydroxyvitamin D levels and risk of multiple sclerosis. JAMA 2006, 296:2832–2838. 49. Zittermann A, Schleithoff SS, Koerfer R: Putting cardiovascular disease and vitamin D insufficiency into perspective. Br J Nutr 2005, 94:483–492. 50. Holick MF: The influence of vitamin D on bone health across the life cycle. J Nutr 2005, 135:2739S–2748S. 51. Zittermann A: Vitamin D and disease prevention with special reference to cardiovascular disease. Progress Biophysics Molecular Biol 2006, 92:39–45. 52. Lappe JM, Travers-Gustafson D, Davies KM, et al.: Vitamin D and calcium supplementation reduces cancer risk: results of a randomized trial. Am J Clin Nutr 2007, 85:1586–1591. 53. Chiuve SE, McCullough ML, Sacks FM, Rimm EB: Healthy lifestyle factors in the primary prevention of coronary heart disease among men: benefits among users and nonusers of lipid-lowering and antihypertensive medications. Circulation 2006, 114:160–167. 38.