Vitamin supplement consumption and breast cancer risk: a review

Background: Breast cancer is the most frequently diagnosed cancer globally, and studies provide contradictory results about the possible effects of vitamin supplementation to reduce cancer risk. Our aim was to conduct a review to better investigate whether vitamin supplements given orally modify breast cancer risk. Methods: We conducted a comprehensive, systematic bibliographic search of the medical literature to identify relevant studies. Case-control, cohort studies, and randomised controlled trials (RCTs) published up to August 2013 that reported cancer risk estimates for vitamin supplementation were included. For each study, we retrieved study characteristics, study population, exposure evaluation, and risk estimates. Results: We identified 26 studies (14 cohort, 11 case-control, and one RCT) and overall, we found 104 estimates. We grouped all the estimates into six supplementation categories: vitamin A and beta-carotene, B-group vitamins and folic acid, vitamin C, vitamin D, vitamin E, and multivitamins. Only a few studies showed a statistically significant association between the consumption of supplemental vitamins and the occurrence of breast cancer, and most of the significant estimates were found in case-control studies. The results found in prospective studies seem to be in the opposite direction. Conclusion: The role of vitamin supplements in preventing breast cancer still remains unclear, considering our review. Although biologic mechanisms exist to support the anticancer effects of vitamins, there is no clear evidence for an effect in cancer prevention for vitamin supplements. Further investigations are warranted to elucidate the mechanisms by which vitamin supplementation can modify breast cancer development.


Introduction
Breast cancer is the most frequently diagnosed cancer globally and also the main cause of death among women [1]. The highest incidence rates are reported in Western and Northern Europe, Australia, and North America, which is mainly owing to changes in reproductive and hormonal factors, and the availability of screening programs [1]. Breast cancer is a complex disorder and can occur as a result of a process that includes oxidative stress and lipid peroxidation mechanisms [2] beyond gene--environment interactions and hormonal expression [3]. Biologic mechanisms exist to support the anticancer effects of dietary antioxidant vitamins and the overall impression is that intake of some vitamins may play a substantial role in the prevention of cancer [4][5][6], but the findings are not strong in this regard.
Observational studies and randomised controlled trials provide contradictory results about the possible preventive effects of vitamin supplementation to reduce cancer risk [7][8][9][10]. Bjelakovic et al [11], in a recent meta-analysis, found no evidence that antioxidant supplements can prevent gastrointestinal cancers; in contrast, they found a possible increasing effect on mortality for excessive consumption. The second World Cancer Research Fund/American Institute for Cancer Research expert report, published in 2007 [6], states that high-dose nutrient supplements can be protective or can cause cancer. The strongest evidence refers to beta-carotene supplementation and lung cancer; actually, high doses of beta-carotene can cause lung cancer in tobacco smokers. The evidence is limited for other vitamins, and there are no recommendations for breast cancer [6].
In view of evaluating the number of published reports on the relation between supplement use and breast cancer risk, we decided to conduct a review to investigate whether vitamin supplements given orally modify breast cancer incidence in humans.

Methods
We conducted a comprehensive, systematic bibliographic search of the medical literature to identify relevant studies on vitamin supplementation use and breast cancer risk. A literature search was performed up to August 2013 using the following databases: PubMed, ISIWeb of Science (Science Citation Index Expanded), and Embase. To identify all relevant published papers, our search consisted of a combination of the following MeSH terms: breast neoplasms, dietary supplements, vitamin E, vitamin D, vitamin A, ascorbic acid, vitamin B complex, vitamin B12, carotenoids, folic acid. We also searched for other non-indexed citations, using keywords such as 'breast' and 'vitamin$ or ascorbic' and 'supplement$'. The related-records feature was used to amplify the search. In addition, reference lists of relevant articles and reviews were reviewed. The search was limited to human studies but no language or time restrictions were applied. We searched for publications with frequencies or estimates of the relative risk (RR), odd ratio (OR), incidence rate ratio (IRR), and hazard ratio (HR) for breast cancer. We decided to consider, for the review, only cohort studies and case-control studies that analysed the correlation between vitamin supplementation and breast cancer risk. Randomised controlled trials (RCTs) were included whenever they reported analysis on baseline vitamin consumptions, but they were excluded if they provided estimates on experimental vitamin supplementation. Reviews and meta-analyses were not considered eligible but were used to retrieve articles.
Studies reporting the biological mechanism of vitamins in cancer, analysis on dietary vitamins, and supplementation were excluded. We excluded two reports [23,24] because we found more recent publications based on the same study population [25,26].
We considered only data about vitamin intake coming exclusively from supplementation. Studies that analysed food plus supplementation intake of vitamins were not considered. We also rejected studies regarding other tumour sites than breast cancer. www.ecancer.org ecancer 2013, 7:365 For each study, we retrieved the following information: • Study characteristics: first author, publication year, country and name of the cohort (if available), study design, starting year of the study, type of supplement • Study population: number of cases and controls or cohort size, age range, menopausal status (if available) • Exposure evaluation: the upper category of daily intake of supplement (Q max versus the lowest category Q1) or duration in years or users versus non-users • Statistic: risk estimates, RR, OR, IRR, or HR and 95% confidence intervals (CIs).

Vitamin A and beta-carotene
We identified seven studies to investigate the associations of vitamin A supplementation with breast cancer risk. Only one case-control study reported a statistically significant increased risk. Moorman et al [43] found that any use of supplemental vitamin A can increase the risk of breast cancer (OR = 1.86; 95% CI: 1.00-3.48). Examining the other studies, three cohort studies reported a decrease of cancer risk but this effect is not statistically significant [4,25,27]. Similar results were found in two case-control studies reported by Longnecker et al [40] and Rohan et al [38]. On the contrary, a case-control study by Dorjgochoo et al [44] found a not significant increased risk (OR = 1.20; 95% CI: 0.80-1.60) for the longest supplemental use.
For beta-carotene supplementation, we found four studies, two cohort [26,31] and two case-control [2,43]. The case-control by Pan et al [2] reported a significantly inverse association between high intake of beta-carotene and breast cancer risk (OR = 0.11; 95% CI: 0.01-0.88) in premenopausal women, but this effect was not found in postmenopausal women.

B-group vitamins
We identified six studies to investigate the associations of vitamin B supplementations with breast cancer risk. The authors reported estimates for single vitamin B, for vitamin B complex, and for folic acid ( Table 2).

Vitamin D
The association between breast cancer risk and vitamin D supplementation has been analysed by three cohort studies [29,30,37]. The results suggest a decreased risk of breast cancer, although not significant. We found a statistically significant decreased risk in two case-control studies: Shamsi et al [46] found a reduced risk for vitamin D supplementation lasting more than three years (OR=0.27; 95% CI: 0.13-0.56) and Rollison et al [45] found a protective effect of more than 400 IU/day vitamin D supplementation (OR=0.79; 95% CI: 0.65-0.96). www.ecancer.org ecancer 2013, 7:365

Discussion
In the present review, we examined the intake of vitamin supplements and its association with breast cancer risk. Out of the 26 studies, we identified 104 estimates evaluating the effect of vitamin A and beta-carotene, B-group vitamins and folic acid, vitamin C, vitamin D, vitamin E, and multivitamin. We found 14 significant estimates. This review addresses the supposition that the use of multivitamin or single vitamins does not have an effect on breast cancer risk.
A point to note is that most of the significant estimates were found in a large-population case-control study within the National Enhanced Cancer Surveillance System in Canada [2] ( Table 2). Pan et al found that supplementations of ten years or longer of multiple vitamins, betacarotene, vitamin C, and vitamin E were associated with statistically significant reductions for breast cancer risk for postmenopausal women. Moreover, they found a reduced cancer risk also for high intake of beta-carotene in premenopausal women and vitamin E in postmenopausal women. It is known that the oxidative stress and lipid peroxidation are linked to the aetiology of breast cancer, and antioxidants may exert their effect, blocking free radicals, inducing apoptosis in cancer cells, and inhibiting cell proliferation [48][49]. However, it seems that other factors might interfere with cancer development, such as menopausal status. This status is physiological but it produces significant modifications in women (body fat distribution, bone and breast density, ovarian and endometrial changes, cardiovascular modifications). It may influence breast cancer development and/or the interaction between antioxidants and the breast tissue, but it is difficult to fully understand the consequences and the potential clinical implications. Another factor that might interfere with cancer development is the receptor status [estrogen receptor (ER) or progesterone receptor (PR)]. These are the most widely studied markers in breast tissue, and their presence has an important clinical relevance. Cui et al found, in the Women's Health Initiative Observational Study on postmenopausal women, an inverse association between breast cancer risk and dietary carotenoids in subject with ER+ and PR+, but not with other breast cancer groups [31]. However, whether hormone receptor-defined breast cancers are etiologically different is not well understood, and it is difficult to identify an effect linked to the vitamin supplementation. We can underline that the results presented by Pan et al can be affected by recall bias inherent in case-control design. Cases might bias their responses to questions on diet after cancer diagnosis. In contrast, cohort and prospective studies have an important advantage because diet assessment is made before diagnosis, and it should be unbiased by cancer experience.
The results seem to be in the opposite direction when found in prospective studies (Table 2), and they are also limited, making it difficult to draw any conclusions.
Regarding the use of multivitamins, Larsson et al [33] highlighted an increase in the risk of developing breast cancer both for high frequency of consumption (19%) and for long duration of multivitamin supplementation (22%) in the Swedish Mammography cohort study. It is not possible to know the composition of the multivitamins and to separate the effect of the single component for the observed associations. Multivitamins are usually a heterogeneous group of products with no standard composition [2,33]. The composition may depend on the manufacturer, year of production, and batches [2,12]. Manufacturers frequently change formulations, and detailed information is not reported [50]. www.ecancer.org ecancer 2013, 7:365 For vitamin C supplementations, we found an increased breast cancer risk of 16% and 46%, respectively, in the Women's Health Initiative Observational Study cohort [31] and in a nested case-control Study (National Breast Screening) [38]. Regarding vitamin A, only Moorman et al [43] in their case-control found an increased risk of breast cancer of 86%, considering any use of supplements. These effects may be due to an increment of oxidative damage [51,52].
We found an increased risk of breast cancer associated with folic acid supplementation. Stolzenberg-Solomon et al [47] found that a supplementation of more than 400 μg/day of folic acid is associated with a 19% increase of the risk of developing breast cancer in the PLCO cancer screening trial. A high folate concentration could contribute to epigenetic changes in gene-regulatory mechanisms, which may result in enhanced cancer development and its promotion. However, results suggest that the role of folic acid in breast cancer development may be more complex, and these mechanisms must be considered with caution [47].
Biological as well as epidemiologic data suggest that vitamin D status could affect cancer risk and play a role in cancer prevention. All the studies included about vitamin D supplementation seem to show a decreased risk in breast cancer, though only two estimates were statistically significant. Rollison et al [45] and Shamsi et al [46] found a decreased risk of 21% and 73% in case-control studies. The link between vitamin D and breast cancer is based on the concept that the vitamin D receptor (VDR) and its ligand 1,25D (the biologically active form of vitamin D) promotes or maintains the differentiated phenotype in normal mammary cells [53]. Nevertheless, other differences in the vitamin D synthesis from sunlight exposure, circulating vitamin D levels, and ethnicity can affect the association [29] and this may explain the null association found in the other studies.
In addition, we found other differences between studies that made it difficult to compare results. Most of the studies collected information about vitamin supplement consumption throughout food frequency questionnaires, estimating average intakes over the preceding year at baseline, or open-ended questions [28,33]. Some researchers reconstructed the consumption of supplements over long periods of time [2,32,43], estimating average intakes over the 5, 10, 20 years preceding (at baseline) or re-administering the questionnaire (cohort study). Their aim is to study the effect of latency time (time from exposure to cancer diagnosis). The specific time intervals after exposure or the relevant exposure time is unknown in diseases with long incubation periods such as cancer. Therefore, it is difficult to identify the relevant exposure period and antioxidants may exert their effect over a long period of time. The lack of evidence underlines the need to develop a more reliable and standardised method to assess supplement consumption.
Another interesting aspect is the dose of vitamin supplementation reported in selected studies ( Table 2). The 'upper level of intake' and also the measurement unit are very heterogeneous. For instance, vitamin A doses vary from 5,000 to 23,000 IU/day, while vitamin C supplementation varies from 64 to 1,300 mg/day. Vitamin E varies from 4 to 424 mg/day. Vitamin consumption is reported in milligrams/micrograms or International Units. However, few studies provided evidences of a dose--response relationship between dose of supplementation and breast cancer risk [31,45,47]. Moreover, in some studies, authors reported only the frequency of consumption, the duration of use, or the user versus non-user data, not enabling us to make a direct comparison between estimates.
We decided to evaluate only vitamin consumption from supplementation, not including dietary intake. It is known that nutritional status and food intake may vary and may influence cancer risk [6], but we would separate these effects, trying to provide a picture of the role of vitamin supplementation in breast cancer risk. The food is constituted of a wide variety of intended and unintended chemicals which may act singly on human metabolism, but more likely act in groups in a synergistic way. Vitamins and bioactive components may interact between them, not acting in isolation [26,54,55].
We retrieved all relevant studies through our systematic search. The data extraction from studies have been performed and checked by individual reviewers, in order not to miss any data. Unfortunately, it should also be noted that some limitations exist. First, the number of studies was relatively small, which may limit our evaluation. We extracted fully adjusted estimates, for known confounding factors, however, in some studies the estimate can be modified by other aspects. It is possible that the observed relationships could be partly due to unmeasured or residual confounding such as other medical conditions or other factors (dietary intake, alcohol intake, sun exposure, lifestyle factors, physical activities, hormone receptor status) that could modify the estimates. www.ecancer.org ecancer 2013, 7:365

Conclusion
In conclusion, while dietary factors may be crucial in modifying cancer risk, the role of vitamin supplements in preventing breast cancer still remains unclear, considering our review. Although biologic mechanisms exist to support the anticancer effects of vitamins, this review did not address whether supplement use may have an effect on breast cancer risk. Considering the current recommendations and expert report published [6,55], there is no clear evidence for an effect in cancer prevention for vitamin supplements due to the fact that few studies have generated relevant data and some results have been contradictory. Further investigations are warranted to elucidate the mechanisms by which vitamin supplementation may modify breast cancer development.

Conflicts of interest
The authors declare that they have no conflicts of interest. www.ecancer.org ecancer 2013, 7:365