Large-scale trials over the past several years have shown a general lack of effect of single vitamins or small numbers of vitamins given at high doses in preventing cancer. However, as recently reported in JAMA by Gaziano and colleagues, the Physicians’ Health Study II has found a modest but significant reduction in risk for “total cancer”—ie, all new cancers—in men taking a daily multivitamin compared with those taking placebo.1 A small reduction in cancer mortality was not statistically significant.
Physicians’ Health Study
The Physicians’ Health Study II was a randomized, double-blind, placebo-controlled 2×2×2×2 factorial trial assessing multivitamins, vitamin E, vitamin C, and beta-carotene use. The components of the trial other than the multivitamin component were terminated several years ago.
In the multivitamin component of the study, 14,641 male physicians aged ≥ 50 years were randomly assigned to receive a daily multivitamin (n = 7,317) or placebo (n = 7,324). Patients had a mean age of 64.3 years. At entry, 1,312 men had a history of cancer (8.9% of multivitamin group, 9.1% of placebo group), excluding melanoma skin cancer. The two groups were well balanced with respect to age, body mass index, smoking status, frequency of exercise, alcohol consumption, aspirin use, parental history of cancer, self-reported history of cancer, and general dietary habits. Enrollment began in 1997, and follow-up continued through June 1, 2011.
Reduced Risk of Total Cancer
Over a median follow-up of 11.2 years, 2,669 men (18.2%) had confirmed cancer, including 1,373 (9.4%) with prostate cancer and 210 (1.4%) with colorectal cancer. The incidence of total cancer was 17.0 events/1,000 person-years in the multivitamin group vs 18.3 events/1,000 person-years in the placebo group, representing a significant 8% reduction in risk with multivitamin use (hazard ratio [HR] = 0.92, 95% confidence interval [CI] = 0.86–0.988, P = .04).
No significant effect of multivitamin use was found for prostate cancer (9.1 vs 9.2 events/1,000 person-years (HR = 0.98, P = .76), colorectal cancer (1.2 vs 1.4 events/1,000 person-years (HR = 0.88, P = .39), or other site-specific cancers, including lung cancer (HR = 0.84, P = .26) and bladder cancer (HR = 0.72, P = .10). As noted by the authors, the power to detect differences was limited in site-specific cancers with low event rates.
Multivitamin use was associated with a significant 12% reduction in risk for all cancers when prostate cancer was excluded from analysis (HR = 0.88, P = .02) and a significant 8% reduction in risk for all epithelial cancers (carcinomas including all cancers except lymphomas and leukemias; HR = 0.92, P = .04).
Among men with a history of cancer at baseline, multivitamin use was associated with a significant 23% reduction in risk for total cancer (HR = 0.73, 95% CI = 0.56–0.96, P = .02). However, this reduction did not differ significantly (P = .07 for interaction) from the nonsignificant reduction in risk among subjects without baseline history of cancer (HR = 0.94, 95% CI = 0.87–1.02, P = .15). The effect of multivitamin use was stronger for total epithelial cancer in men with history of cancer (HR = 0.66, P = .004) than among those without history of cancer (HR = 0.95, P = .21), and the P value for interaction was significant (P = .02).
Few Subgroup Differences
Subgroup analyses showed that multivitamin use significantly reduced risk of cancer in men aged ≥ 70 years (HR = 0.82, 95% CI = 0.72–0.93). However, the test for heterogeneity across age groups (HR = 0.96 for ages 50–59 and 1.01 for ages 60–69) did not reach significance (P = .06 for interaction). Risk was reduced by multivitamin use among subjects with no parental history of cancer (HR = 0.86, 95% CI = 0.76–0.98, P = .02) but not among those with parental history (HR = 1.05, 95% CI = 0.94–1.17, P = .37), and the P value for the interaction was significant (P = .02). No significant heterogeneity was found according to other clinical, lifestyle, or select dietary factors or according to the previously terminated beta-carotene, vitamin C, or vitamin E interventions in the study.
In men with history of cancer at baseline, subgroup analyses showed no significant differential effect of multivitamin use according to diagnosis < 5 years before baseline (HR = 0.80, P = .33) vs diagnosis ≥ 5 years before baseline (HR = 0.70, P = .04; P = .70 for the interaction). Likewise, there was no differential effect according to whether the most recent type of cancer was prostate cancer (HR = 0.66, P = .21) vs other/unknown cancer (HR = 0.78, P = .10; P = .62 for interaction).
No Significant Effect on Mortality
A total of 2,757 men (18.8%) died during follow-up, including 859 (5.9%) due to cancer. A reduction in risk of cancer mortality with multivitamin use fell short of significance (4.9 vs 5.6 events/1,000 person-years, HR = 0.88, 95% CI = 0.77–1.01, P = .07), and no significant reductions in risk for death from site-specific cancers were observed. Multivitamin use was not associated with reduction in risk for overall mortality (HR = 0.94, 95% CI 0.88–1.02, P = .13).
There were no significant differences in rates of reported adherence between the multivitamin and placebo groups at 4 years (76.8% vs 77.1%), 8 years (72.3% vs 70.7%), or end of follow-up (67.5% vs 67.1%). At the end of follow-up, there was no difference between groups with regard to proportions of subjects reporting avoidance of nonstudy multivitamin use (< 30 days per year; 81.0% vs 80.3%).
With regard to safety, no differences between groups were found in for gastrointestinal symptoms, fatigue, drowsiness, skin discoloration, or migraine. Multivitamin use was associated with a higher frequency of rash (HR = 1.07, P = .03). Inconsistent effects on minor bleeding were observed, with multivitamin use being associated with reduced hematuria (HR = 0.91, P = .02), increased epistaxis (HR = 1.10, P = .01), and no effect on easy bruising or other bleeding (HR = 0.99, P = .77).
The authors note that total cancer rates in the trial were likely influenced by increased prostate-specific antigen surveillance and subsequent diagnoses of prostate cancer starting in the late 1990s. Approximately half of all confirmed cancers in the study were prostate cancers, and the majority of these were earlier-stage, lower-grade disease with high survival rates. Thus, the significant reduction in risk for total cancer other than prostate cancer may indicate that multivitamin use has a greater benefit in more clinically relevant cancer diagnoses.
The authors also acknowledge that adherence is of concern, as it is in any long-term trial. Drop-in rates of nonstudy vitamin use increased over time, mirroring increased use of vitamin supplements in the general U.S. population. However, analyses accounting for adherence did not greatly affect the hazard ratios for total or site-specific cancers.
The authors conclude, “[A] daily multivitamin supplement significantly but modestly reduced risk of total cancer during a mean of 11 years of treatment and follow-up. Although the main reason to take multivitamins is to prevent nutritional deficiency, these data provide support for the potential use of multivitamin supplements in the prevention of cancer in middle-aged and older men.” ■
Reference
1. Gaziano JM, Sesso HD, Christen WG, et al: Multivitamins in the prevention of cancer in men: The Physicians’ Health Study II randomized controlled trial. JAMA. October 17, 2012 (early release online).
