Advertisement
Research Article

Serum α-Tocopherol and γ-Tocopherol Concentrations and Prostate Cancer Risk in the PLCO Screening Trial: A Nested Case-Control Study

  • Stephanie J. Weinstein mail,

    weinstes@mail.nih.gov (SJW); daa@nih.gov (DA)

    Affiliation: Nutritional Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Department of Health and Human Servies, Bethesda, Maryland, United States of America

    X
  • Ulrike Peters,

    Affiliations: Cancer Prevention Program, Fred Hutchinson Cancer Research Center, Seattle, Washington, United States of America, Department of Epidemiology, School of Public Health, University of Washington, Seattle, Washington, United States of America

    X
  • Jiyoung Ahn,

    Affiliation: Division of Epidemiology, Department of Population Health, New York University School of Medicine, New York, New York, United States of America

    X
  • Marlin D. Friesen,

    Affiliation: Department of Environmental Health Sciences, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, United States of America

    X
  • Elio Riboli,

    Affiliation: Department of Epidemiology and Public Health, Imperial College, London, United Kingdom

    X
  • Richard B. Hayes,

    Affiliation: Division of Epidemiology, Department of Population Health, New York University School of Medicine, New York, New York, United States of America

    X
  • Demetrius Albanes mail

    weinstes@mail.nih.gov (SJW); daa@nih.gov (DA)

    Affiliation: Nutritional Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Department of Health and Human Servies, Bethesda, Maryland, United States of America

    X
  • Published: July 05, 2012
  • DOI: 10.1371/journal.pone.0040204

Abstract

Background

Vitamin E compounds exhibit prostate cancer preventive properties experimentally, but serologic investigations of tocopherols, and randomized controlled trials of supplementation in particular, have been inconsistent. Many studies suggest protective effects among smokers and for aggressive prostate cancer, however.

Methods

We conducted a nested case-control study of serum α-tocopherol and γ-tocopherol and prostate cancer risk in the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial, with 680 prostate cancer cases and 824 frequency-matched controls. Multivariate-adjusted, conditional logistic regression models were used to estimate odds ratios (OR) and 95% confidence intervals (CIs) for tocopherol quintiles.

Results

Serum α-tocopherol and γ-tocopherol were inversely correlated (r = −0.24, p<0.0001). Higher serum α-tocopherol was associated with significantly lower prostate cancer risk (OR for the highest vs. lowest quintile = 0.63, 95% CI 0.44–0.92, p-trend 0.05). By contrast, risk was non-significantly elevated among men with higher γ-tocopherol concentrations (OR for the highest vs. lowest quintile = 1.35, 95% CI 0.92–1.97, p-trend 0.41). The inverse association between prostate cancer and α-tocopherol was restricted to current and recently former smokers, but was only slightly stronger for aggressive disease. By contrast, the increased risk for higher γ-tocopherol was more pronounced for less aggressive cancers.

Conclusions

Our findings indicate higher α-tocopherol status is associated with decreased risk of developing prostate cancer, particularly among smokers. Although two recent controlled trials did not substantiate an earlier finding of lower prostate cancer incidence and mortality in response to supplementation with a relatively low dose of α-tocopherol, higher α-tocopherol status may be beneficial with respect to prostate cancer risk among smokers. Determining what stage of prostate cancer development is impacted by vitamin E, the underlying mechanisms, and how smoking modifies the association, is needed for a more complete understanding of the vitamin E-prostate cancer relation.

Introduction

Vitamin E compounds are thought to have potential prostate cancer preventive effects, but randomized controlled trials have been inconsistent. Earlier findings of a one-third reduction in prostate cancer incidence in response to daily supplementation with 50 mg (50 IU) of α-tocopherol from the Alpha-Tocopherol, Beta-Carotene Cancer Prevention (ATBC) Study of smokers [1] were not substantiated by two recent trials, the Selenium and Vitamin E Cancer Prevention Trial (SELECT) and the Physicians’ Health Study II Randomized Controlled Trial (PHS-II), which included primarily nonsmokers and tested either a 400 IU daily dose [2], [3] or a 400 IU alternate day dose [4] of vitamin E (α-tocopherol). In addition to the differences in smoking status and supplementation dose across the three trials, the SELECT protocol included pre-randomization (for exclusion) prostate cancer screening which resulted in a preponderance of stage Ia prostate cancers being diagnosed during the trial [2]. Observational data also suggest a vitamin E-prostate cancer-smoking interaction, with a beneficial association for supplemental vitamin E or higher tocopherol status in smokers and for aggressive, and not non-aggressive, disease [5][13].

Given the high incidence of prostate cancer in the U.S. and elsewhere, the biological plausibility that vitamin E could impact cancer risk through several mechanisms [14][18], and the conflicting observational and controlled trial data, further examination of the vitamin E - prostate cancer relationship is needed. To this end, we conducted a prospective nested case-control study of serum concentrations of the two major tocopherols, α- and γ-, in relation to prostate cancer risk in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO). Key a priori aims of our analysis included examination of the serologic dose-risk relation, the relative impact of the two vitamin E congeners, and effect modification by smoking and disease aggressiveness.

Methods

Ethics Statement

The study was approved by the institutional review boards of the U.S. National Cancer Institute and the ten PLCO screening centers, and participants provided written informed consent.

Study Population

We conducted a case-control study nested within the PLCO Screening Trial, an on-going community-based study evaluating the effectiveness of cancer screening tests on site-specific cancer mortality. Participants (ages 55–74) were recruited from ten centers in the United States (Birmingham, AL; Denver, CO; Detroit, MI; Honolulu, HI; Marshfield, WI; Minneapolis, MN; Pittsburgh, PA; Salt Lake City, UT; St Louis, MO; and Washington, DC) between September 1993 and June 2001.

Men randomized to the screening arm of the trial were offered prostate cancer screening by serum prostate-specific antigen (PSA) and digital rectal examination (DRE) at entry and annually for 5 and 3 years, respectively. Men with a PSA test result >4 ng/ml or a DRE exam suspicious for prostate cancer were referred to their medical-care providers for follow-up. Trial participants were asked to provide information regarding recent cancer diagnoses through annual mailed follow-up questionnaires, and medical and pathologic records related to diagnostic follow-up of prostate cancer were obtained by study personnel from medical providers. Periodic searches of the National Death Index were also conducted and death certificates and medical and pathology records related to death were obtained. Data were abstracted by trained medical record specialists.

Data Collection

At enrollment, all participants were asked to complete a risk factor questionnaire including age, ethnicity, education, occupation, smoking history, personal and family medical history, use of selected drugs, recent history of screening exams, and prostate related health factors. In addition, usual dietary intake over the 12 months prior to enrollment was assessed with a 137-item food frequency questionnaire (http://www.cancer.gov/prevention/plco/DQ​X.pdf). Sex- and age-specific portion size and nutrient values were quantified [19]. Total vitamin and mineral intake was calculated by adding dietary and supplemental intake. Non-fasting blood samples were obtained at baseline and in subsequent screening exams from participants in the screening arm [20]. All samples were shipped overnight to a central biorepository and stored at –70oC.

Case Identification and Control Selection

This prostate cancer nested case-control set has been previously described [21]. Briefly, the study included men randomized to the screening arm of the trial, whose first valid prostate cancer screen (PSA test or DRE) was before October 1, 2001. All men were followed from their initial screen to the earliest of: prostate cancer diagnosis, loss-to-follow-up, death, or censor date (October 1, 2001). Cases were defined as non-Hispanic white participants diagnosed with adenocarcinoma of the prostate at least 1 year after blood draw (n = 803). Aggressive cases were defined as those with stage III or IV of the tumor-node-metastasis staging system, as defined by the American Joint Committee on Cancer [22], or Gleason score ≥7. Controls (n = 949) were selected by incidence-density sampling [23] with a case-control ratio of 1:1.2, frequency-matched by age (5-year intervals), time since initial screening (1-year time windows), and year of blood collection. Baseline serum was available for 692 of these cases and 844 controls. We excluded subjects with missing serum cholesterol data, resulting in an analytic set of 680 cases (including 267 aggressive cases) and 824 controls. In a sub-sample of 46 controls, we measured serum tocopherols in samples drawn at baseline and one year follow-up.

Laboratory Analyses

Serum concentrations of α- and γ-tocopherol were determined using reversed-phase high-performance liquid chromatography, with ultraviolet detection [24]. Cholesterol was determined enzymatically using a Hitachi 912 autoanalyzer with a standard procedure at 37°C. Batches of serum samples were organized to include cases, their matched controls, and randomly inserted blinded quality controls. The overall coefficients of variation estimated from the 171 of the latter samples were 5.8% for α-tocopherol and 8.9% for γ-tocopherol. Serum retinol, β-carotene, and lycopene were previously measured [21], [24].

Statistical Analyses

Case and non-case characteristics were compared using t-tests and chi-square tests, for continuous and categorical variables, respectively. Spearman correlations were calculated for tocopherol measurements among the controls taken at two time points, as well as for each tocopherol with age, body mass index (BMI), and several dietary and serum factors. Partial Spearman correlations were used to adjust for factors such as month of blood draw, serum cholesterol concentration, smoking, BMI, age, and energy intake. Conditional logistic regression models were used to estimate odds ratios (OR) and 95% confidence intervals (CIs) for the association between prostate cancer and serum tocopherols. Quintile categories of the nutrients were created based on the distribution among the controls, and entered into the models as indicator variables with the lowest quintile as the referent category. Quintile categories were also calculated separately for analyses stratified on vitamin E supplement dosage, categorized as ≤50 IU/day (defined for purposes in this manuscript as “non-users of vitamin E supplements”) vs. >50 IU/day (defined as “vitamin E supplement users”) from either individual or multivitamin supplements. This categorization was selected because the 50 IU/day supplement in the ATBC Study resulted in a significant increase in serum α-tocopherol [25], and because PLCO participants reported vitamin E intake from multivitamin supplements only of greater than 50 IU/day. Tests for linear trend were obtained by assigning to each nutrient quintile the median value and treating this as a continuous variable. The multivariate models were conditioned on the matching factors (age, time since initial screening, and year of blood draw), and adjusted for serum cholesterol, serum β-carotene, and study center. The following were not confounders in our sample (i.e., when adding each to the model, a <10% change in any of the nutrient coefficients resulted): height, weight, BMI, smoking status, physical activity, educational attainment, marital status, aspirin and ibuprofen use, history of diabetes, history of benign prostatic hyperplasia, family history of prostate cancer, average number of prostate screens (PSA or DRE) per year, serum selenium, month of blood draw, vitamin supplement use, and intakes of total energy, total fat, fruits, vegetables, alcohol, red meat, heterocyclic amine from meat (2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine), lycopene, vitamin C, vitamin E, and calcium. Results are also presented stratified by stage (non-aggressive/aggressive as described above) and smoking status (never-smokers, current smokers, current smokers combined with former smokers who quit <10 years ago, and former smokers who quit ≥10 years ago). Additional subgroup analyses were based on high/low (medians) of age, BMI; serum cholesterol, selenium, and β-carotene; dietary selenium, vitamin C, vitamin E (dietary only and total); vitamin E supplement dose; and follow-up time (1–2 yrs vs. 3 or more years). All stratified models were run using unconditional logistic regression. Multiplicative interactions were tested statistically by comparing models with and without a cross-product interaction term (tocopherol quintiles crossed with the effect modifier split at the median) using the log-likelihood ratio test. Statistical analyses were performed using SAS software version 9.2 (SAS Institute, Inc., Cary, North Carolina) and all p-values were two-sided.

Results

Selected baseline characteristics of cases and controls are shown in Table 1. Cases were more likely to have a history of benign prostatic hyperplasia or familial prostate cancer, and less likely to take aspirin more than once per week. Cases tended to smoke less than controls, although this difference was not statistically significant. Average daily intake of dietary and total (diet plus supplements) vitamin E did not differ by case status. These patterns were similar when aggressive cases were compared with controls, with the exception that aggressive cases had a significantly lower history of diabetes compared with controls (p = 0.04). As was the situation for all cases, aggressive cases tended to smoke less than controls, but this difference was not statistically significant (p = 0.16). Reported doses of supplemental vitamin E ranged from 4 to 1060 IU/day, but 30 IU/day (from a multivitamin supplement), 400 IU/day (from an individual vitamin E supplement), and 430 IU/day (from a combination of a multivitamin and an individual supplement) were most prevalent −27%, 16% and 17%, respectively, of controls who took vitamin E. Approximately 30% of controls reported consuming 50 IU/day or greater of vitamin E from any type of supplement.

thumbnail

Table 1. Selected baseline characteristics by case or control statusa, PLCO Study.

doi:10.1371/journal.pone.0040204.t001

Among controls, serum α-tocopherol and γ-tocopherol were weakly inversely correlated (r = −0.24, p<0.0001; r = −0.39, p<0.0001 with adjustment for serum total cholesterol). Both tocopherols were strongly associated with the α-tocopherol:γ-tocopherol molar ratio and weakly associated with serum β-carotene and BMI, but in opposite directions, and weakly correlated with serum cholesterol, retinol, and lycopene (Table 2). Total vitamin E intake, but not dietary vitamin E intake alone, was positively correlated with serum α-tocopherol and inversely correlated with serum γ-tocopherol; adjustment for month of blood draw, serum cholesterol concentration, smoking, BMI, age, and energy intake had no material impact on these relations (i.e., with adjustment, r = 0.58 for α-tocopherol and −0.56 for γ-tocopherol). Vitamin E supplement use was associated with higher median serum α-tocopherol (14.8 mg/L for non-users vs. 23.7 mg/L for users, p<0.0001) and lower serum γ-tocopherol (3.5 mg/L for non-users vs. 1.4 mg/L for users, p<0.0001). In a sub-sample of 46 controls, two measurements of serum α-tocopherol and γ-tocopherol from baseline and one year follow-up were well-correlated (r = 0.58 and r = 0.80, respectively; both p<0.0001, data not shown). Tocopherol concentrations did not differ by smoking status. For example, median serum α-tocopherol was 16.1, 17.0, 17.2 and 16.8 mg/L for never smokers, former smokers (quit ≥10 years ago), former smokers (quit <10 years ago, and current smokers, respectively (p = 0.28).

thumbnail

Table 2. Correlations between baseline characteristics and α-tocopherol and γ-tocopherol among controls, PLCO Study.

doi:10.1371/journal.pone.0040204.t002

Higher serum α-tocopherol was associated with statistically significantly lower risk of prostate cancer (OR = 0.63, 95% CI 0.44–0.92, for the highest vs. lowest quintile, p-trend 0.05, Table 3). By contrast, there was no clear association with serum γ-tocopherol, although elevated risk was suggested for the four higher quintiles, but with no evidence of dose-response. The patterns were similar when mutually-adjusting for both tocopherols (data not shown). The molar ratio of α-tocopherol to γ-tocopherol was not related to risk of prostate cancer, although risk was significantly elevated for the second quintile (Table 3). The exclusion of vitamin E supplement users (50 IU or greater/day) resulted in an attenuated risk for α-tocopherol [OR = 0.87 (95% CI 0.55–1.38) for the highest quintile (>19.9 mg/L, median = 23.3 mg/L) vs. lowest quintile (≤11.4 mg/L, median = 10.0 mg/L)] and no association for γ-tocopherol [OR = 1.15 (95% CI 0.74–1.79) for the highest quintile (>5.32 mg/L, median = 6.52) vs. lowest quintile (≤2.28 mg/L, median = 1.67 mg/L)]. However, among supplemental vitamin E users, risk appeared lower for higher α-tocopherol [OR = 0.54, 95% CI 0.26–1.01 for the highest quintile (>33.2 mg/L, median = 40.7 mg/L) vs. lowest quintile (≤17.0 mg/L, median = 14.2 mg/L)] with no association for γ-tocopherol [OR = 0.97 (95% CI 0.47–2.01) for the highest quintile (>2.82 mg/L, median = 3.64) vs. lowest quintile (≤0.83 mg/L, median = 0.65 mg/L)]. Post-hoc joint classification using as the reference category men at elevated risk based on both α-tocopherol and γ-tocopherol (i.e., in quintile 1 of α-tocopherol and quintiles 2–5 of γ-tocopherol) revealed an OR of 0.53 (95% CI 0.36–0.77) for men with the lowest risk profile (i.e. in quintiles 2–5 of α-tocopherol and quintile 1 of γ-tocopherol), and an OR = 0.74 (95% CI 0.55–1.00) for the two intermediate risk categories combined (p-trend = 0.001).

thumbnail

Table 3. Association between baseline serum α-tocopherol, γ-tocopherol, and the α-tocopherol:γ-tocopherol molar ratio and risk of prostate cancer, PLCO Study.

doi:10.1371/journal.pone.0040204.t003

Serum α-tocopherol was inversely associated with both non-aggressive and aggressive prostate cancer, although the odds ratios for each quintile were stronger for aggressive disease (Table 4). By contrast, elevated risk for higher serum γ-tocopherol appeared stronger for non-aggressive disease, with a threshold above the lowest quintile and confidence intervals for three OR’s excluding 1.0, but with no significant trend. There was no clear relationship between the molar ratio of the two tocopherols and risk of either non-aggressive or aggressive prostate cancer (data not shown). When comparing the top four quintiles to the lowest quintile of α-tocopherol, the odds ratios for non-aggressive and aggressive disease were 0.76 (95% CI 0.55–1.06, p = 0.11) and 0.67 (95% CI 0.46–0.98, p = 0.04), respectively. Similar risks for γ-tocopherol were OR = 1.64 (95% CI 1.16–2.32, p = 0.01) and OR = 1.17 (95% CI 0.80–1.72, p = 0.41), and for the α-tocopherol:γ-tocopherol ratio were OR = 1.19 (95% CI 0.86–1.66, p = 0.30) and OR = 1.41 (95% CI 0.92–2.16, p = 0.12).

thumbnail

Table 4. Association between baseline serum α-tocopherol and γ-tocopherol and risk of prostate cancer, stratified by disease stage and grade, PLCO Study.

doi:10.1371/journal.pone.0040204.t004

Analyses stratified by smoking status showed lower risk with increasing serum α-tocopherol primarily among current smokers and the combined group of current smokers and those who recently quit smoking (i.e., within the past 10 years), with a significant test for interaction (Table 5). The latter combined subgroup showed a marginally significant dose-risk trend for serum α-tocopherol. Adding current cigar and pipe smokers yielded similar associations in each of these subgroups; for example, the odds ratio for the highest vs. lowest quintile of serum α-tocopherol was 0.41 (95% CI 0.19–0.90, p-trend = 0.02) in the current smoker-recent quitter category. Prostate cancer risk in the current smoker-recent quitter category was also similar when vitamin E supplement users were excluded: OR = 0.33 (95% CI 0.09–1.24) for the highest quintile (>19.9 mg/L) vs. lowest quintile (≤11.4 mg/L); p-trend = 0.02. When smoking strata were further subdivided by disease severity, the inverse association for serum α-tocopherol among current smokers and recent quitters appeared stronger for aggressive prostate cancer (OR for the highest vs. lowest quintile = 0.24, 95% CI 0.05–1.17, p-trend = 0.06) compared with non-aggressive disease (OR = 0.55, 95% CI 0.17–1.79, p-trend = 0.35). Serum α-tocopherol was not associated with prostate cancer among never-smokers (p-trend = 0.49). For γ-tocopherol, the positive risk association appeared strongest among current smokers and recent quitters, although the tests for trends and the interaction test were not significant.

thumbnail

Table 5. Association between baseline serum α-tocopherol and γ-tocopherol and risk of prostate cancer, stratified by smoking status, PLCO Study.

doi:10.1371/journal.pone.0040204.t005

Analyses of other selected subgroups relevant to the vitamin E – prostate cancer association showed that the inverse association with serum α-tocopherol was limited to subjects with total vitamin E intake above the median (OR for highest vs. lowest quintile = 0.36, 95% CI 0.20–0.64, p-trend = 0.01) compared with vitamin E intake below the median (OR = 1.63, 95% CI 0.78–3.39, p-trend = 0.42; p-interaction = 0.03). Risk was also significantly lower for men with high serum α-tocopherol in subgroups defined by older age or lower BMI (data not shown), and significantly higher for men with high γ-tocopherol in subgroups defined by lower serum total cholesterol or shorter follow-up time (data not shown); however, these interactions were not statistically significant. No other subgroups we examined indicated risk interactions for either serum α-tocopherol or γ-tocopherol.

Discussion

Consistent with some prior studies, we found serum α-tocopherol to be inversely associated with prostate cancer risk. This relationship did not differ materially by disease stage, but appeared restricted to current smokers and recently former smokers (p-interaction = 0.049). By contrast, prostate cancer risk appeared elevated among men in all quintiles of γ-tocopherol above the first. Risk was reduced for men who had both high α-tocopherol and low γ-tocopherol concentrations, but was unrelated to the serum tocopherol molar ratio.

Our findings are supported by other studies where inverse associations between serum α-tocopherol or supplemental vitamin E and prostate cancer were limited to current or recent smokers [6], [8][13] or smokers with aggressive disease [5], [7], [26], including a previous analysis of dietary and supplemental vitamin E in PLCO which found lower risk among current and recent smokers for aggressive prostate cancer only [26]. (See Table S1 for a summary of these studies.) By contrast, a protective association for vitamin E supplement use was only evident among never and former smokers in another cohort analysis [27]. Other studies showed no interaction among smoking, serum α-tocopherol or supplemental vitamin E use, and prostate cancer risk [28][32], or non-significant inverse or null associations for serum α-tocopherol overall [33][36]. Also, several [5], [7], [10], [13], [26], [27], [32], but not all [12], [28], [29], [31], prior studies found stronger inverse relations for advanced disease, with some indicating this only among current smokers or recent quitters [5], [26]. In the present analysis, the protective association for higher α-tocopherol status was slightly stronger for aggressive prostate cancer. While we did observe inverse associations for α-tocopherol in both vitamin E supplement users and non-users, the association was somewhat stronger in the supplement users, suggesting that the higher attained serum α-tocopherol concentrations in the supplement users were related to the findings (median α-tocopherol was 14.8 mg/L for non-users vs. 24.2 mg/L for users). However, although the exclusion of vitamin E supplement users attenuated the risk reduction observed for higher serum α-tocopherol, and vitamin E supplement use was higher across increasing α-tocopherol quintiles, the risk reduction observed in current smokers and recently former smokers persisted even with exclusion of the vitamin E supplement users. This is consistent with other studies where the prevalence of vitamin E supplement use was low and/or the median serum α-tocopherol concentrations were lower than those in PLCO [7][10], [13]. For example, in the ATBC and the Physicians’ Health Studies, median α-tocopherol concentrations were, respectively, 11.6 mg/L and 11.1 mg/L, vitamin E supplements were used by 10% and 8% of the men, and ORs were 0.80 (95% CI 0.66–0.96) and 0.51 (95% CI 0.26–0.98, for smokers with aggressive disease) [7], [10]. This indicates that lower prostate cancer risk for higher α-tocopherol concentrations observed among smokers is not limited to vitamin E supplement users.

We found no clear association for serum γ-tocopherol, although prostate cancer risk appeared elevated for men in all quintiles above the first quintile, and adjustment for α-tocopherol had no impact. Circulating γ-tocopherol has been inversely associated with prostate cancer risk in three cohorts [the ATBC Study, the Washington County, MD Study (CLUE), and among smokers for aggressive disease in the Carotene and Retinol Efficacy Trial (CARET)] [9], [13], [31], [33] of eight cohorts in which it was examined [7][9], [13], [28], [29], [31], [33], [34]. The median γ-tocopherol concentration in the present investigation (i.e., 3.0 mg/L) is higher than in previous studies (i.e., 1.0–2.9 mg/L), and while the distribution differs greatly (i.e., higher) from that in the ATBC Study [9], it is fairly similar to that in the CLUE and CARET studies [13], [31], [33]. Similar to our current findings, an inverse correlation between the tocopherols was also reported in the National Health and Nutrition Examination Survey (r = −0.37) [37], which contrasts with a positive correlation in the ATBC Study conducted in Finland (r = 0.51) [9]. This difference could be due to the different food sources of tocopherols in Finland and the United States, or greater vegetable oil consumption and α-tocopherol supplement use in the United States [38]. Given the relatively small number of studies that have measured circulating γ-tocopherol, the inverse relationship between serum α-tocopherol and γ-tocopherol, the suppressive effect of vitamin E supplement use (most of which is α-tocopherol) on circulating γ-tocopherol [39], [40], and the identification of both similar and unique biological activities for the two compounds [38], [41], [42], further study of γ-tocopherol is warranted.

The ATBC Study of male Finnish smokers (n = 29,133, 246 prostate cancer cases) was the first controlled trial to report a significant reduction in the incidence and mortality of prostate cancer in response to daily supplementation with 50 mg (50 IU) of α-tocopherol for a median of 6.1 years [1], [43]. Incidence and mortality were reduced 32% and 41%, respectively, with a 40% reduction in incidence of advanced prostate cancer and no reduction for early stage disease [1] (see Table S2 for a review of the trials described here). A subsequent trial in France, SU.VI.MAX (n = 5,141, 103 prostate cancer cases), reported that daily supplementation with 30 mg α-tocopherol for 8 years (along with other antioxidants in the combination supplement) significantly reduced the incidence of prostate cancer among men with normal PSA at baseline (HR = 0.52, 95% CI 0.29–0.92) [44]. Risks did not differ by smoking status, but only 15% of participants were current smokers. Two cardiovascular disease/diabetes trials of daily α-tocopherol supplementation, the Heart Outcomes Prevention Evaluation Trial (400 IU, n = 6,996, 235 prostate cancer cases) and the Heart Protection Study (600 mg in combination with other antioxidants, n = 15,454, 290 prostate cancer cases), showed no effect on prostate cancer incidence [45], [46]. These trials included approximately only 14% and 25% current smokers, respectively, however. Most recently, two trials of healthy men, SELECT (400 IU α-tocopherol daily for a median of 5.5 years, n = 35,533, 2,279 cases) [2], [3] and PHS-II (400 IU alternate days for a median of 7.6 years, n = 14,641, 1,008 cases) [4], also reported no beneficial effect of α-tocopherol supplementation, while additional follow-up of SELECT showed significantly elevated prostate cancer incidence [3]. These two recent trials also included very few current smokers (only 8% and 4% of participants, respectively). In addition to smoking status, another factor potentially related to the inconsistent findings across ATBC, SELECT and PHS-II is the substantially lower vitamin E dose used in ATBC (50 IU/day with beneficial effects), compared with PHS-II (on average 200 IU/day with no effect), and SELECT (400 IU/day with harmful effects). Interestingly, the different dosages resulted in very similar increases in average on-study blood concentrations (from 11.5 to 17.3 mg/L in ATBC versus 12.8 to 18.4 mg/L in SELECT, for example). Another factor that differed among these trials was the baseline eligibility requirement for normal PSA and digital rectal examinations in SELECT, which resulted in few diagnoses of advanced prostate cancer (only approximately 1.1% of all prostate cancers in SELECT) [2], [47]; i.e., precisely the diagnostic category exhibiting lower incidence in the vitamin E arm of the ATBC Study [1]. Although the PHS-II protocol did not require prostate cancer screening at study entry [4], that trial also observed fewer advanced cases than expected, possibly as a result of greater prostate screening consciousness in that population of U.S. physicians [4], [47]. Given that the protective association for vitamin E supplementation or status also appeared stronger for advanced prostate cancer in several observational studies [5], [7], [10], [27], [32], in addition to the ATBC trial [1], which would be consistent with a tumor growth inhibitory effect, the original null results from SELECT and PHS-II may not be surprising. The follow-up findings of significantly greater prostate cancer incidence in the vitamin E groups in SELECT [3] are singular and difficult to explain given substantial previous research.

Cigarette smokers have increased oxidative stress [48], and although circulating tocopherol concentrations tend not to differ between smokers and nonsmokers [48], [49], smokers have increased rates of α-tocopherol disappearance [49]. Therefore, the stronger risk reduction with higher serum α-tocopherol concentrations among smokers is biologically plausible. Although higher vitamin E status could theoretically lower prostate cancer risk in smokers through its chain-breaking antioxidant or anti-inflammatory functions, experimental data indicating tocopherol and tocotrienol inhibition of cell proliferation, cell adhesion, and protein kinase C activity are more consistent with a reduction in tumor progression [14], [15]. For example, prostate cancer cell line growth is inhibited by α-tocopheryl succinate by suppressing androgen receptor expression, prostate-specific antigen, and cell cycle regulatory elements [50], [51]. In a recent analysis of adult men in the National Health and Nutrition Examination Survey III, strong inverse relations between serum α-tocopherol and testosterone, estradiol, and sex hormone binding globulin (SHBG) were observed in cigarette smokers and those with elevated serum cotinine concentrations [52]. These findings corroborated an earlier report from the ATBC Study showing decreased circulating androgens in male smokers supplemented with α-tocopherol [53] and provide a biologically plausible mechanism for the inhibitory influence in the development of prostate cancer. α-Tocopherol supplementation also decreased vascular epithelial growth factor (VEGF) concentrations, which could reduce prostate tumor angiogenesis and growth [54], [55], but this may not be limited to smokers. Cytochrome P450 (CYP) enzymes and other enzymes responsible for metabolism or activation of carcinogens in tobacco [56] are also involved in the metabolism of vitamin E compounds (e.g., CYP3A4 and CYP4F2) [57][60]. Induction or competitive antagonism of these enzymes could be consistent with the present findings of a protective association between α-tocopherol and prostate cancer in current smokers, although this may be more relevant for tumor initiation than for growth inhibition.

Properties of γ-tocopherol that differ from α-tocopherol include reduced hepatic secretion into very low density lipoproteins, resulting from the preferential uptake of α-tocopherol by the α-tocopherol transfer protein; selective inhibition of prostaglandin E2 synthesis and cyclooxygenase activity; protection against reactive nitrogen species; and inhibition of prostate cancer cells in vitro [18], [38], [41], [42], [61], [62]. How these might account for the suggested positive γ-tocopherol - prostate cancer risk association is unclear and should be examined in further studies. Alternatively, this could simply reflect the suppressive effect of α-tocopherol supplement use on circulating γ-tocopherol [39], [40], as 80% of vitamin E supplement users were in the two lowest quintiles of γ-tocopherol, including 29% that were in both the highest α-tocopherol and the lowest γ-tocopherol quintile. Adjustment for serum α-tocopherol did not alter the risk estimates for γ-tocopherol, however.

The prospective design of our study and the availability of pre-diagnostic blood samples reduced the potential for an effect of prostate cancer on the serum tocopherol measurements (i.e., reverse causality). The availability of data for numerous prostate cancer risk factors facilitated testing and adjustment for potential confounding, and reduced the likelihood of residual confounding. The large number of prostate cancer cases permitted stratification by several factors, notably disease aggressiveness and smoking status, although the number of current smokers was still relatively small. All subjects were selected from the PLCO Trial screening arm (within which biospecimens were collected) and received annual prostate cancer screening under standardized procedures, thereby reducing the likelihood of differential screening practices and diagnostic bias. Fasting status prior to blood collection was not ascertained, and participants were not specifically instructed to fast. Our analyses are based only on baseline tocopherol measures, although in a subset of controls, tocopherol measures one year later were well-correlated with baseline values (r = 0.58 and 0.80 for α- and γ-tocopherol, respectively), and 15-yr repeatability data indicate correlations of 0.46–0.61 for α-tocopherol and 0.48–0.53 for γ-tocopherol [63]. The prevalent use of supplemental vitamin E in this population (31% among controls) and the relatively high serum tocopherol concentrations allowed a robust examination of the hypothesis, although the follow-up time of 8 years was relatively short compared with some other longitudinal studies.

In conclusion, our data indicate a significant inverse association between serum α-tocopherol and prostate cancer risk, observed primarily among smokers. Serum γ-tocopherol appeared to be directly related to risk, but this association was not statistically significant. Our finding of a stronger association among smokers is supported by previous studies, and suggests a biological mechanism for vitamin E that is particularly beneficial among smokers, possibly related to greater tumor growth inhibition in this population. This, along with other factors described above, may partially explain the inconsistent outcomes from clinical trials of vitamin E and prostate cancer, and suggests that further examination of vitamin E and prostate cancer, with careful attention to smoking exposure, disease screening and stage, and mechanisms, is needed.

Supporting Information

Table S1.

Observational studies of vitamin E and prostate cancer which stratified on smoking status: selected risk estimates.

doi:10.1371/journal.pone.0040204.s001

(XLS)

Table S2.

Randomized controlled intervention trials which examined vitamin E and prostate cancer.

doi:10.1371/journal.pone.0040204.s002

(XLS)

Acknowledgments

The authors thank Drs. Christine Berg and Philip Prorok, Division of Cancer Prevention, National Cancer Institute, the Screening Center investigators and staff of the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial, Mr. Tom Riley and staff, Information Management Services, Inc., and Ms. Barbara O’Brien and staff, Westat, Inc. Most importantly, we acknowledge the study participants for their contributions to making this study possible.

Author Contributions

Conceived and designed the experiments: UP JA RBH DA. Performed the experiments: MDF. Analyzed the data: SJW DA. Contributed reagents/materials/analysis tools: MDF ER. Wrote the paper: SJW DA.

References

  1. 1. Heinonen OP, Albanes D, Virtamo J, Taylor PR, Huttunen JK, et al. (1998) Prostate cancer and supplementation with alpha-tocopherol and beta-carotene: incidence and mortality in a controlled trial. J Natl Cancer Inst 90: 440–446.
  2. 2. Lippman SM, Klein EA, Goodman PJ, Lucia MS, Thompson IM, et al. (2009) Effect of selenium and vitamin E on risk of prostate cancer and other cancers: the Selenium and Vitamin E Cancer Prevention Trial (SELECT). JAMA 301: 39–51.
  3. 3. Klein EA, Thompson IM, Tangen CM, Crowley JJ, Lucia M, et al. (2011) Vitamin E and the Risk of Prostate Cancer. The Selenium and Vitamin E Cancer Prevention Trial (SELECT). JAMA 306: 1549–1556.
  4. 4. Gaziano JM, Glynn RJ, Christen WG, Kurth T, Belanger C, et al. (2009) Vitamins E and C in the prevention of prostate and total cancer in men: the Physicians’ Health Study II randomized controlled trial. JAMA 301: 52–62.
  5. 5. Chan JM, Stampfer MJ, Ma J, Rimm EB, Willett WC, et al. (1999) Supplemental vitamin E intake and prostate cancer risk in a large cohort of men in the United States. Cancer Epidemiol Biomarkers Prev 8: 893–899.
  6. 6. Eichholzer M, Stahelin HB, Gey KF, Ludin E, Bernasconi F (1996) Prediction of male cancer mortality by plasma levels of interacting vitamins: 17-year follow-up of the prospective Basel study. Int J Cancer 66: 145–150.
  7. 7. Gann PH, Ma J, Giovannucci E, Willett W, Sacks FM, et al. (1999) Lower prostate cancer risk in men with elevated plasma lycopene levels: results of a prospective analysis. Cancer Res 59: 1225–1230.
  8. 8. Goodman GE, Schaffer S, Omenn GS, Chen C, King I (2003) The association between lung and prostate cancer risk, and serum micronutrients: results and lessons learned from beta-carotene and retinol efficacy trial. Cancer Epidemiol Biomarkers Prev 12: 518–526.
  9. 9. Weinstein SJ, Wright ME, Pietinen P, King I, Tan C, et al. (2005) Serum alpha-tocopherol and gamma-tocopherol in relation to prostate cancer risk in a prospective study. J Natl Cancer Inst 97: 396–399.
  10. 10. Weinstein SJ, Wright ME, Lawson KA, Snyder K, Mannisto S, et al. (2007) Serum and dietary vitamin E in relation to prostate cancer risk. Cancer Epidemiol Biomarkers Prev 16: 1253–1259.
  11. 11. Watters JL, Gail MH, Weinstein SJ, Virtamo J, Albanes D (2009) Associations between alpha-tocopherol, beta-carotene, and retinol and prostate cancer survival. Cancer Res 69: 3833–3841.
  12. 12. Rodriguez C, Jacobs EJ, Mondul AM, Calle EE, McCullough ML, et al. (2004) Vitamin E supplements and risk of prostate cancer in U.S. men. Cancer Epidemiol Biomarkers Prev 13: 378–382.
  13. 13. Cheng T, Barnett MJ, Kristal AE, Ambrosone CB, King IB, et al. (2011) Genetic variation in myeloperoxidase modifies the association of serum alpha-tocopherol with aggressive prostate cancer among current smokers. J Nutr 141: 1731–1737.
  14. 14. Fleshner NE (2002) Vitamin E and prostate cancer. Urol Clin North Am 29: 107–13, ix.
  15. 15. Ricciarelli R, Zingg JM, Azzi A (2001) Vitamin E: protective role of a Janus molecule. FASEB J 15: 2314–2325.
  16. 16. Brigelius-Flohe R, Kelly FJ, Salonen JT, Neuzil J, Zingg JM, et al. (2002) The European perspective on vitamin E: current knowledge and future research. Am J Clin Nutr 76: 703–716.
  17. 17. Venkateswaran V, Klotz LH (2010) Diet and prostate cancer: mechanisms of action and implications for chemoprevention. Nat Rev Urol 7: 442–453.
  18. 18. Traber MG (1999) Vitamin E. In: Shils ME, Olson JA, Shike M, Ross AC, editors. Modern Nutrition in Health and Disease. Baltimore, MD: Lippincott Williams & Wilkins. pp. 347–362.
  19. 19. Subar AF, Midthune D, Kulldorff M, Brown CC, Thompson FE, et al. (2000) Evaluation of alternative approaches to assign nutrient values to food groups in food frequency questionnaires. Am J Epidemiol 152: 279–286.
  20. 20. Hayes RB, Reding D, Kopp W, Subar AF, Bhat N, et al. (2000) Etiologic and early marker studies in the prostate, lung, colorectal and ovarian (PLCO) cancer screening trial. Control Clin Trials 21: 349S–355S.
  21. 21. Peters U, Leitzmann MF, Chatterjee N, Wang Y, Albanes D, et al. (2007) Serum lycopene, other carotenoids, and prostate cancer risk: a nested case-control study in the prostate, lung, colorectal, and ovarian cancer screening trial. Cancer Epidemiol Biomarkers Prev 16: 962–968.
  22. 22. Fleming ID, Cooper JS, Henson ED, Hutter RVP, Kennedy BJ, Murphy GP, O’Sullivan B, Sobin LH, Yarbro JW (1997) AJCC Cancer Staging Manual, 5th edition. Philadelphia: Lippincot-Raven.
  23. 23. Rothman K, Greenland S (1998) Modern Epidemiology. Philadelphia: Lippincott Raven.
  24. 24. Steghens JP, van Kappel AL, Riboli E, Collombel C (1997) Simultaneous measurement of seven carotenoids, retinol and alpha-tocopherol in serum by high-performance liquid chromatography. J Chromatogr B Biomed Sci Appl 694: 71–81.
  25. 25. The ATBC Cancer Prevention Study Group (1994) The alpha-tocopherol, beta-carotene lung cancer prevention study: design, methods, participant characteristics, and compliance. Ann Epidemiol 4: 1–10.
  26. 26. Kirsh VA, Hayes RB, Mayne ST, Chatterjee N, Subar AF, et al. (2006) Supplemental and dietary vitamin E, beta-carotene, and vitamin C intakes and prostate cancer risk. J Natl Cancer Inst 98: 245–254.
  27. 27. Peters U, Littman AJ, Kristal AR, Patterson RE, Potter JD, et al. (2008) Vitamin E and selenium supplementation and risk of prostate cancer in the Vitamins and lifestyle (VITAL) study cohort. Cancer Causes Control 19: 75–87.
  28. 28. Key TJ, Appleby PN, Allen NE, Travis RC, Roddam AW, et al. (2007) Plasma carotenoids, retinol, and tocopherols and the risk of prostate cancer in the European Prospective Investigation into Cancer and Nutrition study. Am J Clin Nutr 86: 672–681.
  29. 29. Gill JK, Franke AA, Steven MJ, Cooney RV, Wilkens LR, et al. (2009) Association of selenium, tocopherols, carotenoids, retinol, and 15-isoprostane F(2t) in serum or urine with prostate cancer risk: the multiethnic cohort. Cancer Causes Control 20: 1161–1171.
  30. 30. Stram DO, Hankin JH, Wilkens LR, Park S, Henderson BE, et al. (2006) Prostate cancer incidence and intake of fruits, vegetables and related micronutrients: the multiethnic cohort study (United States). Cancer Causes Control 17: 1193–1207.
  31. 31. Huang HY, Alberg AJ, Norkus EP, Hoffman SC, Comstock GW, et al. (2003) Prospective study of antioxidant micronutrients in the blood and the risk of developing prostate cancer. Am J Epidemiol 157: 335–344.
  32. 32. Wright ME, Weinstein SJ, Lawson KA, Albanes D, Subar AF, et al. (2007) Supplemental and dietary vitamin E intakes and risk of prostate cancer in a large prospective study. Cancer Epidemiol Biomarkers Prev 16: 1128–1135.
  33. 33. Helzlsouer KJ, Huang HY, Alberg AJ, Hoffman S, Burke A, et al. (2000) Association between alpha-tocopherol, gamma-tocopherol, selenium, and subsequent prostate cancer. J Natl Cancer Inst 92: 2018–2023.
  34. 34. Nomura AM, Stemmermann GN, Lee J, Craft NE (1997) Serum micronutrients and prostate cancer in Japanese Americans in Hawaii. Cancer Epidemiol Biomarkers Prev 6: 487–491.
  35. 35. Knekt P, Aromaa A, Maatela J, Aaran RK, Nikkari T, et al. (1988) Serum vitamin E and risk of cancer among Finnish men during a 10-year follow-up. Am J Epidemiol 127: 28–41.
  36. 36. Hayes RB, Bogdanovicz JF, Schroeder FH, De Bruijn A, Raatgever JW, et al. (1988) Serum retinol and prostate cancer. Cancer 62: 2021–2026.
  37. 37. Ford ES, Schleicher RL, Mokdad AH, Ajani UA, Liu S (2006) Distribution of serum concentrations of alpha-tocopherol and gamma-tocopherol in the US population. Am J Clin Nutr 84: 375–383.
  38. 38. Wagner KH, Kamal-Eldin A, Elmadfa I (2004) Gamma-Tocopherol - An Underestimated Vitamin? Ann Nutr Metab 48: 169–188.
  39. 39. Huang HY, Appel LJ (2003) Supplementation of diets with alpha-tocopherol reduces serum concentrations of gamma- and delta-tocopherol in humans. J Nutr 133: 3137–3140.
  40. 40. Handelman GJ, Machlin LJ, Fitch K, Weiter JJ, Dratz EA (1985) Oral alpha-tocopherol supplements decrease plasma gamma-tocopherol levels in humans. J Nutr 115: 807–813.
  41. 41. Jiang Q, Christen S, Shigenaga MK, Ames BN (2001) Gamma-tocopherol, the major form of vitamin E in the US diet, deserves more attention. Am J Clin Nutr 74: 714–722.
  42. 42. Saldeen K, Saldeen T (2005) Importance of tocopherols beyond alpha-tocopherol: evidence from animal and human studies. Nutrition Research 25: 877–889.
  43. 43. The Alpha-Tocopherol Beta Carotene Cancer Prevention Study Group (1994) The effect of vitamin E and beta carotene on the incidence of lung cancer and other cancers in male smokers. N Engl J Med 330: 1029–1035.
  44. 44. Meyer F, Galan P, Douville P, Bairati I, Kegle P, et al. (2005) Antioxidant vitamin and mineral supplementation and prostate cancer prevention in the SU.VI.MAX trial. Int J Cancer 116: 182–186.
  45. 45. Lonn E, Bosch J, Yusuf S, Sheridan P, Pogue J, et al. (2005) Effects of long-term vitamin E supplementation on cardiovascular events and cancer: a randomized controlled trial. JAMA 293: 1338–1347.
  46. 46. Heart Protection Study Collaborative Group (2002) MRC/BHF Heart Protection Study of antioxidant vitamin supplementation in 20,536 high-risk individuals: a randomised placebo-controlled trial. Lancet 360: 23–33.
  47. 47. Gann PH (2009) Randomized trials of antioxidant supplementation for cancer prevention: first bias, now chance–next, cause. JAMA 301: 102–103.
  48. 48. Bruno RS, Traber MG (2005) Cigarette smoke alters human vitamin E requirements. J Nutr 135: 671–674.
  49. 49. Bruno RS, Ramakrishnan R, Montine TJ, Bray TM, Traber MG (2005) Alpha-tocopherol disappearance is faster in cigarette smokers and is inversely related to their ascorbic acid status. Am J Clin Nutr 81: 95–103.
  50. 50. Zhang Y, Ni J, Messing EM, Chang E, Yang CR, et al. (2002) Vitamin E succinate inhibits the function of androgen receptor and the expression of prostate-specific antigen in prostate cancer cells. Proc Natl Acad Sci U S A 99: 7408–7413.
  51. 51. Ni J, Chen M, Zhang Y, Li R, Huang J, et al. (2003) Vitamin E succinate inhibits human prostate cancer cell growth via modulating cell cycle regulatory machinery. Biochem Biophys Res Commun 300: 357–363.
  52. 52. Mondul AM, Rohrmann S, Menke A, Feinleib M, Nelson WG, et al. (2011) Association of serum alpha-tocopherol with sex steroid hormones and interactions with smoking: implications for prostate cancer risk. Cancer Causes Control 22: 827–836.
  53. 53. Hartman TJ, Dorgan JF, Woodson K, Virtamo J, Tangrea JA, et al. (2001) Effects of long-term alpha-tocopherol supplementation on serum hormones in older men. Prostate 46: 33–38.
  54. 54. Woodson K, Triantos S, Hartman T, Taylor PR, Virtamo J, et al. (2002) Long-term alpha-tocopherol supplementation is associated with lower serum vascular endothelial growth factor levels. Anticancer Res 22: 375–378.
  55. 55. Mondul AM, Rager HC, Kopp W, Virtamo J, Albanes D (2011) Supplementation with alpha-tocopherol or beta-carotene reduces serum concentrations of vascular endothelial growth factor-D, but Not -A or -C, in male smokers. J Nutr 141: 2030–2034.
  56. 56. Rossini A, de Almeida ST, Albano RM, Pinto LF (2008) CYP2A6 polymorphisms and risk for tobacco-related cancers. Pharmacogenomics 9: 1737–1752.
  57. 57. Traber MG (2010) Regulation of xenobiotic metabolism, the only signaling function of alpha-tocopherol? Mol Nutr Food Res 54: 661–668.
  58. 58. Parker RS, Sontag TJ, Swanson JE, McCormick CC (2004) Discovery, characterization, and significance of the cytochrome P450 omega-hydroxylase pathway of vitamin E catabolism. Ann N Y Acad Sci 1031: 13–21.
  59. 59. Sontag TJ, Parker RS (2002) Cytochrome P450 omega-hydroxylase pathway of tocopherol catabolism. Novel mechanism of regulation of vitamin E status. J Biol Chem 277: 25290–25296.
  60. 60. Major JM, Yu K, Wheeler W, Zhang H, Cornelis MC, et al. (2011) Genome-wide association study identifies common variants associated with circulating vitamin E levels. Hum Mol Genet.
  61. 61. Jiang Q, Elson-Schwab I, Courtemanche C, Ames BN (2000) gamma-tocopherol and its major metabolite, in contrast to alpha-tocopherol, inhibit cyclooxygenase activity in macrophages and epithelial cells. Proc Natl Acad Sci U S A 97: 11494–11499.
  62. 62. Moyad MA, Brumfield SK, Pienta KJ (1999) Vitamin E, alpha- and gamma-tocopherol, and prostate cancer. Semin Urol Oncol 17: 85–90.
  63. 63. Comstock GW, Burke AE, Hoffman SC, Norkus EP, Gross M, et al. (2001) The repeatability of serum carotenoid, retinoid, and tocopherol concentrations in specimens of blood collected 15 years apart. Cancer Epidemiol Biomarkers Prev 10: 65–68.