הסרטן של פיליס

http://www.ynet.co.il/articles/0,7340,L-3353692,00.html

Current Opinion in Gastroenterology Volume 23(2), March 2007, p 159–163 Vitamins AND prevention of cardiovascular disease AND cancer: should we give supplements? Nutrition: Edited by David H. Alpers AND William F. Stenson] Alpers, David H Washington University School of Medicine, St Louis, Missouri, USA Abbreviations 25(OH)D: 25-hydroxyvitamin D; CI: confidence interval; CVD: cardiovascular disease; DRI: dietary recommended intake; LDL: low-density lipoprotein; OR: odds ratio; RR: relative risk. Introduction One of the factors that most altered the recent dietary recommended intakes (DRIs) was the increased use of dietary vitamin AND mineral supplements. Much of this increased use has been due to perceived health benefits, in particular as they apply to prevention of chronic disease [1•]. Cardiovascular disease AND cancer are the diseases that cause the most concern AND for which supplements are most commonly taken. This article will review the recent evidence regarding the use of vitamins (folate especially) to prevent hyperhomocysteinemia, AND whether that test is a significant risk factor for cardiovascular disease in the general population. The issue of the level of body stores that define vitamin D deficiency will be discussed, AND whether supplemental vitamin D may be useful in prevention of cardiovascular disease AND cancer. Finally, the antioxidant theory of prevention of cancer (and also cardiovascular disease) will be reviewed. In all cases the data are incomplete (more so for vitamin D than for the homocysteine AND antioxidants), so this review will serve as an interim report. Enough data have accumulated in some of these areas that preliminary conclusions can be attempted [2•]. Hyperhomocysteinemia AND atherosclerosis Factors leading to atherosclerosis have included low-density lipoprotein (LDL), oxidized-LDL, elevated blood pressure, increased angiotensin II AND increased serum total homocysteine. The homocysteine story started with the observation of McCully [3] that accelerated atherosclerosis was found in a genetic disorder of homocysteine metabolism, cystathionine synthase deficiency. Subsequently the same phenotype was seen in genetic deficiencies of methionine synthase AND methylene-tetrahydrofolate reductase (MTHFR), enzymes that convert homocysteine to methionine AND that provide the 5-methyl donor group for that conversion. Rats lack choline oxidase, AND on a choline-free diet cannot convert homocysteine to methionine, AND develop atheromatous changes. Monkeys on a soy protein (methionine limited) diet develop atheromatous changes. The Physicians' Health Study [4] showed that serum homocysteine was a risk factor for myocardial infarction [4], AND the Nurses' Health Study [5] showed that deficiencies of the enzymes needed to convert homocysteine to methionine (folate, vitamin B6 AND vitamin B12) were associated with an increased mortality due to vascular disease. These AND other studies have led to a huge program of studies designed to prove the value of supplemental vitamins in lowering serum homocysteine AND in preventing cardiovascular disease (CVD). It is incontrovertible that folate can reduce serum homocysteine: 0.8 mg of folate/day reduces the levels by as much as one-third, if serum homocysteine is elevated to at least 20 µmol/l AND serum folate is low (5 nmol/l) [6]. When folate AND homocysteine levels are normal, the effect is more modest (~15%). A metaanalysis in 2000 [7] of 33 observational studies confirmed a modest association [odds ratio (OR) 1.58] between increased serum homocysteine AND CVD [7]. A review of 81 studies (31 355 individuals) [8] examined for the MTHFR polymorphism C677T (that increased serum homocysteine) showed a weighted mean difference for serum homocysteine in stroke risk of 1.93 µmol/l, AND the only major source of heterogeneity was serum folate. A metaanalysis [9] of 20 prospective studies of disease risk demonstrated that a plasma homocysteine level higher by 5 µmol/l corresponded to an OR for stroke of 1.59. The difference of 1.93 µmol/l, which was observed by Casas, would predict an OR for stroke of 1.2, AND a review of 30 studies showed an observed OR of 1.26 [confidence interval (CI) 1.14–1.4]. It was concluded that a causal relationship existed between plasma homocysteine AND stroke. Another metaanalysis of studies examining the MTHFR C677T polymorphism [10] showed a similarly weak association (OR 1.2) with coronary heart disease. The message of these observational studies on large databases was that very large trials would be needed to detect plausible differences in major outcomes, such as death, myocardial infarction, stroke, OR tumor, but that only moderate effects should be expected [11]. The B-vitamin Treatment Trialists' Collaboration [12•] reviewed all the homocysteine-lowering trials in progress, AND concluded that most of them were underpowered for showing a difference in stroke, only four out of 12 were adequately powered to show a difference in major coronary events, AND only the occurrence of the sum of all major vascular events was adequately powered in eight out of 12 studies. It may take a metaanalysis of all the trial results to answer the question of whether lowering serum homocysteine prevents CVD. Meanwhile two of the studies have reported initial results, but neither of them was adequately powered for major coronary events OR stroke alone, although both were powered to detect a change in all vascular events. The Norwegian Vitamin Trial (NORVIT) [13•] provided 0.8 mg folate, 0.4 mg vitamin B12, AND 40 mg vitamin B6, OR placebo to 3749 patients within a week of a myocardial infarction (MI), AND followed them for up to 40 months. The primary endpoint was a composite of recurrent MI, stroke, AND sudden death attributed to CVD; the relative risk (RR) was 1.22 (CI 1.0–1.5). For all MIs the RR was 1.23 (CI 0.99–1.52). These differences were significant, but the CI just barely crossed 1.0. The Health Outcomes Prevention Evaluation trial (HOPE-2) [14•] treated 5522 older adults with chronic stable vascular disease for over 5 years with folate 2.5 mg, vitamin B6 50 mg, AND vitamin B12 1 mg OR placebo. Serum homocysteine fell from a mean of 12.2 to 9.7 µmol/l but no advantage was seen in total cardiovascular events OR mortality. Fewer patients in the treated group had a stroke (RR 0.75, CI 0.59–0.97), but more patients were hospitalized for unstable angina. Some excitement has been reported from a reanalysis of the data from the Vitamin Intervention for Stroke Prevention (VISP) Trial in which 3680 adults at risk were treated with folate 2.5 mg, vitamin B6 25 mg, AND vitamin B12 0.4 mg/day versus low-dose therapy of the three vitamins. The study was stopped because of futility as the interim analysis showed very small differences. Arguing that folate fortification could have equalized the folate effect, a posthoc analysis selected patients most likely to respond by removing those who may be supplementing OR malabsorbing, who had low OR high serum values, OR with decreased glomerular filtration rate. Of the 2155 patients left, AND combining all clinical outcomes into a composite score, the high vitamin group had a 21% reduction compared with the low-dose group [15]. This result must be accepted very cautiously, as is the case for nearly all posthoc analyses that do not confirm a real trend in the overall sample results. In summary, epidemiological data show that elevated serum homocysteine is a well established risk factor for CVDs, AND strongly predicts both first AND recurring cardiovascular events. Serum homocysteine is also directly linked to the status of nutrients related to its metabolism, especially folate, vitamin B6, AND vitamin B12. The first three large trials to report efficacy, however, show no OR very little effect. There are small signals from each of the studies within subgroups, but unless these are confirmed by subsequent studies, the recommendation of the American Heart Association not to test for serum homocysteine nor to treat elevated levels with vitamin supplements should be respected for managing most patients with a risk of CVD [16]. Vitamin D deficiency AND chronic disease The body stores of vitamin D are reflected by the serum level of 25-hydroxyvitamin D [25(OH)D]. The issue that has been raised is whether adequacy of stores is reflected by the 95% CI of values from a ‘normal’ population, OR whether adequacy should reflect the level that depresses parathormone secretion AND minimizes bone turnover. ‘Normal’ values of 25(OH)D range from 10 to 50 ng/ml (25–125 nmol/l), a range that is quite large. This issue is related to the relative value of dietary intake OR supplements as compared with the much more efficient stimulus of exposure to sunlight. One minimal erythemic dose (MED) of ultraviolet radiation releases from 10 000 to 20 000 IU of vitamin D3 from the skin in 24 h [17•]. In practical terms, 1 MED can be delivered to a nontanned Caucasian in a bathing suite within 10–12 min of peak July summer sun in Boston. It would take 120 min to deliver a similar dose to an African-American. One MED elevates serum 25(OH)D to a concentration of around 20 ng/ml [18]. At this level pararthormone secretion is decreased maximally to around 40 pg/ml. Secondary hyperparathyroidism AND calcium absorption decrease in elderly patients at 25(OH)D concentrations over 30 ng/ml. When vitamin D is provided in oral supplements, however, the recommended dietary intake (400 IU) raises 25(OH)D levels by only 2.8 ng/ml after 5 months. To achieve larger increases of 7 OR 28 ng/ml the dose of ingested vitamin D would need to be 1000 OR 5000 IU, respectively [17]. Using linear regression with plasma 25(OH)D as the dependent variable, vitamin D intake predicted a relatively small proportion of the variance in plasma concentrations in a follow-up to the Health Professionals' Study [19•]. This result is consistent with the small increment in plasma 25(OH)D effected by oral vitamin D. In this study each increment of 100 IU/day (the contents of a glass of milk) in vitamin D3 intake increased plasma 25(OH)D by only 0.7 mg/ml. Leisure time activity, perhaps a proxy for sun exposure, was one of the strongest determinants of plasma 25(OH)D. An increment of 10 ng/ml (25 nmol/l) in plasma 25(OH)D provided significantly decreased relative risks for gastrointestinal cancers, ranging from 0.4 (oral/pharyngeal) to 0.8 (colorectal cancer). Prospective studies will be needed to confirm the results from this cohort study in men. Another review of 63 observational studies [20] included studies of cancer of the colon (30 studies), breast (13 studies), prostate (26 studies), AND ovary (seven studies). The majority of studies suggested that when vitamin D status was sufficient, the risk of cancer was lower. Once again, prospective intervention studies are needed. Neither epidemiological studies nor prospective supplement trials have yet tested the effect of vitamin D on protection of CVD, despite extensive preclinical data to suggest such an effect [18]. In animal models (mostly mice), vitamin D has shown antiinflammatory AND antiatherogenetic activity. It has decreased cardiac hypertrophy AND myocyte proliferation, AND is inversely associated with elevated renin AND blood pressure in mice with a disrupted vitamin D receptor (VDR) gene. The antioxidant theory AND antioxidant vitamins The antioxidant theory is based on the ability of cellular free radicals derived from lipid peroxides to produce inflammation AND contribute to malignant transformation, AND the reversal of these findings with antioxidant agents. The natural antioxidant system is complex, AND involves factors that decrease tissue levels of lipid peroxides, at least vitamin E, methionine (as a precursor for reduced glutathione) AND selenium (as a cofactor for glutathione peroxidase). In addition, vitamin C is thought to act as a general reducing agent. There is a large amount of epidemiological evidence linking diet, nutrients, AND cancer. None of the evidence, however, is yet considered convincing for antioxidants [21]. The link has been judged as probable for selenium AND folate for colorectal cancer, but fruit AND vegetables have also been associated with prevention of gastrointestinal cancers involving esophagus AND stomach [22]. Most studies involving vitamins, minerals, AND nonnutrient plant constituents (flavonoids, isoflavones) have been considered insufficient to make a significant link to cancer. Carotenoid intake has not been associated with lowering the risk of lung cancer in seven cohort studies in North America AND Europe, involving nearly 400 000 participants followed for 7–16 years [23]. The only specific carotenoid which was found to be inversely associated (RR = 0.76) was [beta]-crytpoxanthin. Similar results have been found on prevention of gastrointestinal cancers when [beta]-carotene was used as a supplement, either by itself OR in conjunction with vitamins A, C, OR E [24]. No prevention was noted with vitamin E, selenium, OR combinations of vitamins AND minerals. The striking observation made in this systematic review was that the range of RR was so large between many studies. For example, [beta]-carotene supplement produced a RR of 0.15 for esophageal cancer, but the confidence interval was 0.01–3.7. This wide range suggests caution in interpreting individual studies. Whether the reason for such a wide range is the inability to control for variations within a population, OR insufficient powering of each study, OR a combination of many reasons is not known. Antioxidant intake (vitamin E, C, [beta]-carotene) did not alter the risk of prostatic cancer in 29 361 men followed for 8 years in the Prostate, Lung, Colorectal, AND Ovarian Cancer Screening Trial [25], despite some earlier suggestions for an effect of selenium. In one subgroup (recent smokers with >400 IU/day intake of vitamin E), the RR was 0.29 (CI 0.12–0.68). In another subgroup with [beta]-carotene less than the median AND supplemented with 2.0 mg/day, the RR was 0.52 (CI 0.33–0.81). Overall, however, the data were not sufficiently strong to recommend supplementation to a general population. The intake of vitamins A, C, AND E AND folate was inversely related to lung cancer when eight prospective studies were analyzed [26]. The difference, however, was greatly reduced after adjusting for smoking AND other risk factors. Six prospective primary prevention studies using vitamin E alone OR in combination with other antioxidants (vitamin C, [beta]-carotene) have been reviewed recently [27]. In most studies no clinically significant effects were seen. The striking exception is the study from Linxian, China, that showed a decrease in mortality from stomach cancer AND from all causes in those taking 33 IU/day of an unspecified source of vitamin E. The data from two secondary prevention trials suggested a twofold decrease in nonfatal MI (Cambridge Heart Antioxidant Study – CHAOS), AND a composite score of cardiovascular events (Secondary Prevention with Antioxidants of Cardiovascular Disease in End-stage Renal disease – SPACE). Another review [28] of nine prospective intervention studies found that vitamin C, but not vitamin E, was associated with a lower RR for coronary heart disease events. Mortality, however, was not altered. A review of eight population studies [29] of carotenoid intake concluded that a higher intake of fruit AND vegetables appears to be useful in prevention of heart disease; the data on intake of single nutrients (e.g. carotenoids) was not convincingly helpful. What are the reasons for the discrepancy between observational studies of association AND prospective studies of supplementation? The reasons are not known, but some differences seem clear. First, most evidence from diet AND disease prevention has come from studies in groups consuming dietary constituents in foods AND not as separate supplements. Second, recent population AND intervention studies are confounded by the widespread use of supplements in Western countries [30]. Third, older epidemiological studies were cross-sectional cohort studies, but more recent studies have been prospective. Fourth, most intervention studies do not deliver as supplements the full biological profile of antioxidants (vitamins A, C, AND E, selenium, methionine, coenzyme Q, etc.). Fifth, studies cannot control for cofactors that affect risk of cancer OR CVD, such as exercise, food preparation, among others. Sixth, pharmacological doses of specific nutrients given to middle-aged adults may not have predictable effects on diseases with latency periods longer than the period of observation. One study found that regular use of vitamin supplements over 10 years before lowered the risk of colorectal cancer, but supplement use in the decade prior to the study was without effect [31]. Finally, it is possible that the reason for the discrepancy is that the antioxidant theory of chronic disease may not be correct. Some doubts have started to surface regarding the LDL oxidation story [32]. In-vitro experiments showed that modification of LDL required redox-active metal ions AND were reversed by lipid-soluble antioxidants. This led to the theory that lipid oxidation was essential for LDL atherogenesis. Probucol, an antioxidant, inhibited atherosclerosis in animal models of hypercholesterolemia. As discussed above, however, vitamin E supplements do not prevent CVD in humans. It is now clear that the pathways that promote oxidative stress in animals AND humans differ. For example, myeloperoxidase is present in human arterial walls, but not in the mouse arteries. Furthermore, vitamin E does not scavange hypochlorous acid, a major product of myeloperoxidase [32]. A recent study [33•] showed that the benefit of probucol is unrelated to its antoxidant properties in three animal models, in that the degree of lipid oxidation in vascular tissue was unrelated to the extent of the lesions. Thus, it may be that at least part of the antioxidant theory may not be correct. Since the signal for decreasing risk does not appear to be large, if even part of the theory is incorrect, it may prevent a positive reproducible response in the intervention trials. Moreover, it should be remembered that there are suggestions (equally inconclusive) that supplemental vitamin E OR [beta]-carotene may increase cancer risk [34,35]. Conclusion Folate lowers total serum homocysteine levels, but any effect on CVD OR stroke prevention is uncertain. Therefore, routine screening for serum homocysteine AND folate supplementation are not indicated at this time for prevention of CVD. The DRI may be set too low for vitamin D at 400 IU/day, but a role of vitamin D supplementation to produce 25(OH)D concentrations in the upper half of the normal range has not yet been demonstrated, either to prevent cancer OR CVD. Epidemiological studies have suggested that intake of antioxidants (vitamins A, C, AND E in particular) is inversely correlated with the incidence of cancer AND CVD, but intervention studies have not yet been supportive of this association. The current recommendation on the use of vitamins for the general population should be directed at ensuring an adequate intake as outline by the DRI recommendations. Of course, recommendations can be individualized for persons at special risk OR who are deficient in body stores of one OR more vitamin [36].