The Physicians’ Health Study II - PHS II


Although many people routinely consume multivitamins to prevent chronic diseases, data on benefit from randomized clinical trials (RCTs) have been mostly negative. The current trial sought to study the impact of daily consumption of multivitamins on the incidence of cancer and cardiovascular disease (CVD).


Daily consumption of multivitamins would be associated with a significant reduction in all cancers and CVD in healthy subjects.

Study Design

  • Blinded
  • Factorial
  • Parallel
  • Placebo Controlled
  • Randomized
  • Stratified

Patient Populations:

  • Male physicians who agreed to participate
  • Took ≥2/3 of their pills

    Number of enrollees: 14,641
    Duration of follow-up: 11.2 years
    Mean patient age: 64.2 years
    Percentage female: 0%


  • Cirrhosis or active liver disease
  • On anticoagulants
  • Reported a serious illness that might preclude participation
  • Willing to forego the use of multivitamins or individual supplements containing more than 100% of the recommended dietary allowance of vitamin E, vitamin C, beta carotene, or vitamin A

Primary Endpoints:

  • All cancers (excluding nonmelanoma skin cancers)
  • Major adverse cardiac events (nonfatal MI, nonfatal stroke, CV death)

Secondary Endpoints:

  • Prostate cancer
  • Colorectal cancer
  • Total MI
  • Total stroke
  • Ischemic and hemorrhagic stroke
  • CV mortality
  • All-cause mortality 

Drug/Procedures Used:

Patients were randomized in a 2 x 2 x 2 x 2 factorial design to a multivitamin (Centrum Silver or its placebo daily); vitamin E (400 IU synthetic α-tocopherol or its placebo on alternate days); vitamin C (500 mg synthetic ascorbic acid or its placebo daily); and beta carotene (50 mg Lurotin or placebo on alternate days). The beta carotene component was terminated in March 2003. Treatment and follow-up of the vitamin E and vitamin C components continued through August 2007, with findings of no overall association reported for cancer and cardiovascular disease. Participants were sent monthly calendar packs containing a multivitamin or placebo (taken daily) every 6 months for the first year, then annually thereafter.

Principal Findings:

A total of 14,641 patients were randomized, 7,317 to a multivitamin and 7,324 to placebo. Baseline characteristics were fairly similar between the two arms. The mean body mass index was approximately 26.0 kg/m2. Approximately 56% were never smokers (only 3.6% current smokers), 6% had diabetes, 37% had hyperlipidemia, and 42% had hypertension. Approximately 61% exercised ≥1 time/week, and 19% never consumed alcohol. The median intake per day was 4.2 servings of fruits and vegetables, 1.1 servings of whole grains, and 0.6 servings of red meat. Approximately 9% had a self-reported history of cancer (4.5% with prostate cancer) and approximately 53% had a family history of cancer (10% with prostate cancer, 12% with colorectal cancer). Similarly, about 5% had a personal history of CVD and 10% a family history of premature CVD. Adherence was 77% at 4 years, 71% at 8 years, and 67% at the end of follow-up.

Cancer: Over a median duration of follow-up of 11.2 years, there was a significant, but modest reduction in total cancer incidence (17 vs. 18.3/1,000 person-years; hazard ratio [HR], 0.92; 95% confidence interval [CI], 0.86-0.998; p = 0.04). Total epithelial cancers were similarly reduced (HR, 0.92; 95% CI, 0.85-0.997; p = 0.04). About one-half of all incident cancers were prostate cancer (most of them early stage), with no difference between the two arms (HR, 0.98; 95% CI, 0.88-1.09; p = 0.76). On exclusion of these cancers, all cancers were still significantly reduced in the multivitamin arm (HR, 0.88; 95% CI, 0.79-0.98; p = 0.02). No reductions were noted in site-specific cancers. Total (HR, 0.94; 95% CI, 0.88-1.02; p = 0.13) and cancer-related (HR, 0.88; 95% CI, 0.77-1.01; p = 0.07) mortality were similar, but numerically lower in the multivitamin arm.

There was significant effect modification by parental history of cancer (p for interaction = 0.02). Men with no parental history of cancer had a beneficial effect of a daily multivitamin on total cancer (HR, 0.86; 95% CI, 0.76-0.98; p = 0.02), although men with a parental history of cancer did not (HR, 1.05; 95% CI, 0.94-1.17; p = 0.37). No significant heterogeneity by age, other clinical, lifestyle, dietary factors, or by the previously terminated vitamin C, vitamin E, and beta carotene interventions of PHS II was found. Benefit appeared to be greater for secondary prevention of cancer (HR, 0.73; 95% CI, 0.56-0.96; p = 0.02), especially for total epithelial cancers (HR, 0.66; 95% CI, 0.50-0.88; p = 0.004). No differences were noted with duration of prior cancer.

CV events: Over a median follow-up of 11.2 years, the primary MACE endpoint was similar between the multivitamin and placebo arms (11.0 vs. 10.8/1,000 person-years; HR, 1.01; 95% CI, 0.91-1.10; p = 0.91). Individual endpoints including MI (3.9 vs. 4.2/1,000 person-years, p = 0.39) and stroke (4.1 vs. 3.9/1,000 person-years, p=0.48) were similar between the two arms. Similarly, other endpoints such as congestive heart failure, angina, and coronary revascularization were all similar between the two arms. Fewer MI deaths were noted in the multivitamin arm (p = 0.048). There were no differences for the primary endpoint based on gender, age, and baseline CVD status.

Men taking the multivitamin were more likely to report rashes (HR,1.07; 95% CI, 1.01-1.14; p = 0.03) and epistaxis (HR, 1.10; 95% CI, 1.02-1.18; p = 0.01).


The results of the PHS II trial in middle-aged to elderly male physicians demonstrate that daily intake of a multivitamin results in a small, but statistically significant decrease in all cancers over 11 years of follow-up, especially nonprostate cancer, with numerically lower cancer-related mortality. There was no difference in CV endpoints. For the cancer outcome, benefit appeared maximal in patients with no parental history of cancer and also in those with a personal history of cancer. This is one of the largest RCTs on the topic, and also one of the only trials to demonstrate a beneficial effect on cancer with daily multivitamin use. Other trials have either shown no benefit or even possibly an increased risk of certain cancers in women. For CV endpoints, this is the only large RCT to assess the impact of long-term multivitamin use.

A few points need to be considered. All subjects in this trial were male physicians. They had healthier lifestyles than the general population, with a low rate of baseline risk factors. It is thus unclear if these results can be applied to women, other members of the general population, and non-US populations. These subjects were also nutritionally replete, and thus, these results may not be applicable to nutritionally deficient populations. Given the nature of the pill used (combination multivitamin), it is impossible to tell if one specific component or several components in synergy were responsible for the beneficial effects on cancer. Similarly, it is possible that a balanced diet, with a high proportion of fruits and vegetables, may provide similar benefits on cancer. A better understanding of the potential role of these nutrients on cancer modification is also necessary. There is no role for multivitamins in the primary or secondary prevention of CV events.


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 2012;Oct 17:[Epub ahead of print].

Sesso HD, Christen WG, Bubes V, et al. Multivitamins in the Prevention of Cardiovascular Disease in Men: The Physicians' Health Study II Randomized Controlled Trial. JAMA 2012;308:1751-60

Presented by Dr. Howard Sesso at American Heart Association Scientific Sessions, Los Angeles, CA, November 5, 2012.

Clinical Topics: Dyslipidemia, Heart Failure and Cardiomyopathies, Prevention, Acute Heart Failure, Diet, Hypertension

Keywords: Stroke, Life Style, Vitamin E, Follow-Up Studies, beta Carotene, Hyperlipidemias, Risk Factors, Epistaxis, Body Mass Index, alpha-Tocopherol, Heart Failure, Colorectal Neoplasms, Diet, Confidence Intervals, Hypertension, Prostatic Neoplasms, Diabetes Mellitus

< Back to Listings