Routine Vitamin Supplementation

to Prevent Cancer: A Summary

of the Evidence From Randomized
Controlled Trials for the
U.S. Preventive Services Task Force
Cheryl Ritenbaugh, PhD, MPH; Kelly Streit, MS, RD; Mark Helfand, MD, MS

Cancer is the second leading cause of death in randomized controlled trials designed to test the
the United States, accounting for 1 of every 4 efficacy of vitamin supplements in the primary
deaths.1 Nutritional status has long been speculated prevention of cancer have also been undertaken.
to play a significant role in certain cancers. One These randomized trials have examined the effect
theory holds that oxidative damage to cells of vitamin supplements on cancer biomarkers, on
contributes to carcinogenesis. In laboratory the incidence and progression of precancerous
experiments, the antioxidant vitamins, vitamin C, lesions, on the incidence of invasive cancer, and
vitamin E (alpha-tocopherol), and beta-carotene, on cancer-specific and all-cause mortality.
counteract damage to biomolecules due to oxidants,2
raising the possibility that increased intake of these Most of the trials that examined cancer incidence
vitamins might prevent cancer. Vitamin A (retinol), and mortality have been published in the last
which acts within the cell to control gene expression, decade. In light of the large body of new evidence,
and folic acid,3 which is involved in DNA the U.S. Preventive Services Task Force decided
methylation and purine and pyrimidine synthesis, that it was timely to review the benefits of vitamin
may also have a role in preventing cancer.4 supplementation. This summary reviews randomized
trials that addressed this question, posed by the
In epidemiological studies, low dietary intake and Task Force: Do antioxidant vitamin supplements
blood levels of certain antioxidant vitamins have reduce all-cause mortality, cancer mortality, or the
been associated with a higher incidence of certain incidence of cancer or certain precancerous lesions in
cancers and higher cancer mortality.5 Several the general adult population of the United States?

From: Oregon Evidence-based Practice Center, Department of Medical Informatics & Clinical Epidemiology, Oregon Health &
Science University (Helfand, Streit) Portland, OR; Kaiser Permanente Center for Health Research, Portland, OR (Ritenbaugh).
Corresponding Author: Cheryl Ritenbaugh, PhD, MPH, Senior Investigator, Kaiser Permanente Center for Health Research,
3800 N. Interstate Avenue, Portland, OR 97227-1110. Phone: (503) 335-2400; fax: (503) 335-2424; e-mail:
Cheryl.Ritenbaugh@kp.org.
The authors of this evidence summary are responsible for its contents, including any clinical or treatment recommendations.
No statement in this article should be construed as an official position of the Agency for Healthcare Research and Quality or the
U.S. Department of Health and Human Services.
Reprints are available from the AHRQ Web site (www.preventiveservices.ahrq.gov) and through the National Guideline Clearinghouse
(www.guideline.gov). The summaries of the evidence on these topics are also available from the AHRQ Publications Clearinghouse in
print or through subscription to the Guide to Clinical Preventive Services, Third Edition: Periodic Updates. To order, contact the
Clearinghouse at 1-800-358-9295 or e-mail ahrqpubs@ahrq.gov. This chapter first appeared on the AHRQ Web site on July 1, 2003.
The USPSTF recommendations based on this evidence review and another review, Routine Vitamin Supplementation to Prevent
Cardiovascular Disease, can be found in Routine Vitamin Supplementation to Prevent Cancer and Cardiovascular Disease:
Recommendations and Rationale, available on the AHRQ Web site and in the Guide to Clinical Preventive Services, Third Edition:
Periodic Updates.

1
 Routine Vitamin Supplementation to Prevent Cancer

Methods The searches identified 932 citations, of which 102

were reports from 36 randomized controlled trials.
Literature Search and Study Ten of these trials were included in this review. The
excluded trials either had no eligible cancer endpoints,
Selection combined included with excluded nutrients, or had
The criteria for inclusion in the review were not yet reported results (Appendix II).
developed in consultation with members of the
USPSTF. English-language randomized controlled
Analysis/Synthesis
trials and prospective cohort studies concerning
adults in developed countries were eligible for Two reviewers independently abstracted
inclusion. Case-control studies were excluded unless descriptive data from the included trials, using one
they were performed in the context of a prospective form for abstraction of information about the study
cohort study (ie, a nested case-control study). design and another form for results. To assess study
Studies of supplementation with vitamin A, quality, we used the system developed by the
vitamin C, vitamin E, beta-carotene, folic acid, USPSTF, which includes a set of 6 criteria to rate
combinations of these vitamins, or a multiple the internal validity of each study as “good,” “fair,”
vitamin were eligible if they reported a) the or “poor.”6 For clinical trials we also assessed study
incidence of or mortality from any invasive cancer quality using the Jadad score.7 We summarized the
other than nonmelanoma skin cancer or b) the results of studies in evidence tables organized by
incidence of colonic polyps. Studies of other type of study, nutrient, and outcome. For
precancerous lesions, carcinoma in situ, and supplement/outcome combinations with sufficient
regression of cancer or of precancerous lesions evidence, we assessed heterogeneity among studies
were excluded. and conducted meta-analyses using a pairwise,
sequential procedure based on maximum likelihood
This summary reports the results of our review methods (M. Aickin, PhD, and M. Helfand, MD,
of randomized trials that address this question. The MPH, unpublished data, 2000).
results of the cohort studies were presented to the
USPSTF, but they are excluded from this report Role of the Funding Source
because they did not contribute to the Task Force’s This research was funded by AHRQ under a
recommendations. Appendix I summarizes the contract to support the work of the USPSTF.
results of the included cohort studies. AHRQ staff and USPSTF members participated in
the initial design of the study and reviewed interim
We searched the Cochrane Controlled Trials analyses and the final manuscript.
Registry (December, 2001) and the MEDLINE
database from 1966 to December 2001 using
terms for the 5 nutrients (A, C, E, beta-carotene, Results
and folate) as well as multivitamin and antioxidant Table 1 summarizes the characteristics of the 9
supplements and terms for cancer and randomized controlled trials that met the criteria for
precancerous lesions. We also searched the inclusion. Beta-carotene supplementation has been
reference lists of review articles and, in several studied in 5 controlled trials. Only 1 trial examined
rounds of review of earlier manuscripts, asked the effect of vitamin E supplementation on cancer
experts for additional references. Finally, we incidence and mortality. For these outcomes, no
searched MEDLINE again (December 2001) using studies of vitamin A or vitamin C alone versus
the acronyms or full titles of the major trials and placebo have been completed. Vitamin C has been
cohort studies to identify additional publications. studied only in combination with other vitamins.
Two reviewers applied the eligibility criteria listed None of the included trials addressed supplemen-
above after reviewing the titles and abstracts of tation with folic acid. Six studies examined vitamin
retrieved citations and again after selecting full-text combinations; 4 did so as part of a factorial design
articles for closer review. and 2 did not.

2
 Routine Vitamin Supplementation to Prevent Cancer

Beta-Carotene a placebo group. The vitamin A group had lower

Evidence from 5 randomized controlled trials8–12 total mortality than the beta-carotene group, largely
indicates that beta-carotene supplementation due to a significantly lower risk for developing
increases lung cancer incidence in smokers and has mesothelioma (RR, 0.24; 95% CI, 0.07–0.86; 3 vs
no effect on cancer incidence in nonsmokers. These 12 cases). Incidence of lung cancer, however, was
trials were designed to test beta-carotene at levels not significantly lower in the subjects randomized to
above those that could be achieved through food, vitamin A (RR, 0.66; 95% CI, 0.19–2.32). Without
approximately 10 times the average U.S. intake.13 untreated controls, it is uncertain whether these
They achieved substantially higher blood levels of findings indicate a harmful effect of beta-carotene
beta-carotene than those associated with benefit or a beneficial effect of vitamin A, but they confirm
in epidemiological studies. the findings of other trials of a lack of benefit of
beta-carotene.

Lung Cancer The Physicians’ Health Study found no impact

of beta-carotene on lung cancer incidence in an
Lung cancer was the primary endpoint in the
average-risk population with a low prevalence of
Alpha-Tocopherol Beta-Carotene (ATBC) trial.9 In
smokers (RR, 0.93; 95% CI, 0.69–1.25).8 There
ATBC, 29,133 male smokers were randomized in
was also no significant increase in lung cancer
geographically defined blocks to vitamin E, beta-
incidence in the beta-carotene arm of the Women’s
carotene, both, or neither.9 Baseline characteristics
Health Study (RR, 1.43; 95% CI, 0.82–2.48).10
were similar in the 4 groups. There were only 113
However, this arm was terminated early after a
exclusions after randomization, 97 of whom were
median duration of treatment of only 2.1 years
excluded because Finnish cancer registry records
due to the lack of benefits of beta-carotene (and
or a chest x-ray showed that cancer was present
suggestion of possible adverse effects) observed in
prior to randomization. Analysis was by intention-
the other trials discussed above.10
to-treat. The study was terminated early for a
significant adverse effect of beta-carotene on lung
cancer incidence among heavy smokers (relative Prostate Cancer
risk [RR], 1.19; 95% confidence interval [CI], For prostate cancer, there was no significant effect
1.03–1.35). Total mortality was 8% higher (RR, of beta-carotene on the incidence of prostate cancer
1.08; 95% CI, 1.01–1.16) among the participants in ATBC (RR, 1.23; 95% CI, 0.95–1.60).16 In the
who received beta-carotene than among those who Physicians’ Health Study, there was no effect on
did not, primarily because there were more deaths lung cancer incidence among all men randomized
from lung cancer and ischemic heart disease. to beta-carotene (RR, 0.99; 95% CI, 0.88–1.11).8
However, among subjects in the lower quartile of
The Beta-Carotene and Retinol Efficacy Trial
serum beta-carotene at baseline, the incidence of
(CARET)14 was designed to test the combination
prostate cancer was significantly reduced among
of beta-carotene (30 mg/d) and vitamin A (retinyl
subjects who received beta-carotene compared with
palmitate 25,000 IU/d) to prevent lung cancer in
those receiving placebo.17, 18
18,000 high-risk participants—asbestos workers and
heavy smokers. After 4 years, there were statistically
significant increases in the incidence of lung cancer Colon Cancer
(RR, 1.28; 95% CI, 1.04–1.57; P=0.02), mortality Beta-carotene supplementation had no effect
from lung cancer (RR, 1.46; 95% CI, 1.07–2.00), on the incidence of colon cancer in ATBC21, (RR,
and all-cause mortality (RR, 1.17; 95% CI, 1.05; 95% CI, 0.75–1.47), in the Physicians’
1.03–1.33). Health Study8 (RR, 0.96; 95% CI, 0.78–1.18), or
in the Women’s Health Study (RR, 0.99; 95% CI,
A randomized controlled trial among 1,024
0.62–1.60).10
asbestos workers in Australia15 compared a group
receiving beta-carotene (30 mg) with a group Two studies tested the impact of beta-carotene on
receiving vitamin A (25,000 IU/d retinol) without the recurrence of adenomatous polyps. A US trial11

3
 Routine Vitamin Supplementation to Prevent Cancer

Table 1. Randomized controlled trials of vitamin supplementation to

prevent cancer (beta-carotene)*

Study, Duration of follow-up;

publication Setting/population Treatment follow-up rate
Hennekens 22,071 US male physicians 50 mg beta-carotene Mean 12 years;
19968 aged 40–84 years; on alternate days; 99.99%
Physicians’ no history of: cointervention with 325 mg
Health Study • cancer aspirin on alternate days
• myocardial infarction in 2 x 2 factorial design
• stroke
• cerebral ischemia
• noncompliance in run-in phase

Alpha-Tocopherol SW Finland; 20 mg/day beta-carotene; 5–8 years;

Beta-Carotene 29,133 male smokers 50 mg/day alpha- tocopherol median 6.1 years
(ATBC) aged 50–69 years in 2 x 2 factorial design
Study Group
19949

Greenberg 6 US gastroenterology clinics; 25 mg/day beta-carotene 3 years between year 1

199411 864 subjects; vs placebo; cointervention with and year 4 colonoscopies
Polyp Prevention 79% male < 80 years; combination of 1 g/day vitamin C (mean 36.6 months; no
Study at least 1 polyp and 400 mg/day vitamin E or difference between groups)
placebo in 2 x 2 factorial design
MacLennan Australian gastrointestinal 20 mg/day beta-carotene; 2–4 years
199512 clinics; fat reduction (25% of total
Australian 424 subjects; energy target); wheat bran fiber
Polyp 67% male; (25 g/day of finely milled
Prevention aged 30–74 years; raw wheat bran) in 2 x 2 x 2
Study at least 1 polyp factorial design

I-M Lee 39,876 US female health 50 mg beta-carotene on 4 years

199910 professionals > 45 years alternate days vs placebo;
Women’s cointervention with 100 mg
Health aspirin or placebo on alternate
Study days and 600 IU vitamin E or
placebo on alternate days in
2 x 2 x 2 factorial design.
Beta-carotene study terminated
early.

DeKlerk 1,024 Australian male (92%) 30 mg/day beta-carotene (n=512) Through 5/95;
199815 and female (8%) asbestos mine or 25,000 IU/day retinol (n=512); mean FU 4.5 years
Mesothelioma workers; mean age 57; No placebo group
Prevention Study 21% current smokers, 52% ex-
smokers, 27% never smokers

*See Appendix III for additional information on these trials.

Note: ATBC indicates Alpha-Tocopherol Beta-Carotene; g, gram; mg, milligram; n, number; FU, followup; NR, not reported;
CARET, Beta-Carotene and Retinol Efficacy Trial.

4
 Routine Vitamin Supplementation to Prevent Cancer

Outcomes (95% confidence interval)

Lung Prostate Breast Colon All-cause
cancer cancer cancer cancer Other cancer mortality
Incidence: Incidence: NR Colon/ Incidence: 1.01
0.93 0.99 Rectum Stomach: 0.90 (0.49–1.67) (0.93–1.10)
(0.69–1.25; (0.88–1.11) Incidence: Pancreas: 1.38 (0.79–2.40)
NS) 0.96 Brain: 0.81 (0.48–1.36)
(0.78–1.18) Melanoma: 0.88 (0.63–1.22)
Leukemia: 0.83 (0.53–1.30)
Lymphoma: 1.07 (0.79–1.45)
Bladder: 1.51 (1.02–2.24)
Thyroid: 8.00 (2.05–31.23)
All cancer mortality: 1.02 (0.88–1.17)

Incidence: Incidence: NR Incidence: 1.05 Pancreatic incidence: 0.75 (0.49–1.14) 1.08

1.19 1.23 (0.75–1.47) Pancreatic mortality: 0.81 (0.53–1.26) (1.01–1.16)
(1.03–1.36) (0.95–1.60) Adenomatous
polyps
recurrence:
0.98
(0.71–1.35)

NR NR NR Adenomatous NR 1.03
polyps (0.82–1.30)
recurrence:
1.01
(0.85–1.20)

NR NR NR Incidence: NR NR
1.5 (0.9–2.5)*

Incidence: NR Incidence: Incidence: 0.99 Incidence: 1.07

1.43 1.01 (0.62–1.60) Uterus: 1.15 (0.69–1.91;NS) (0.74–1.56)
(0.82–2.48) (0.81–1.24) Ovary: 1.33 (0.73–2.43)
Thyroid: 0.75 (0.32–1.74;NS)
Bladder: 0.83 (0.27–2.57)
Brain: 0.67 (0.20–2.20)
Pancreas: 1.50 (0.45–4.94)
Cervix: 0.67 (0.13–3.33)
Stomach: 0.99 (0.10–9.58)
All cancer: 1.03 (0.89–1.18)
All cancer mortality: 1.11 (0.67–1.85)

Incidence: NR NR NR Malignant mesothelioma 0.56

0.66 incidence: 0.24 (0.07–0.86) (0.33–0.95)
(0.19–2.32) Other cancer
mortality: 0.97 (0.24–3.90)

Note: ATBC indicates Alpha-Tocopherol Beta-Carotene; g, gram; mg, milligram; n, number; FU, followup; NR, not reported;
CARET, Beta-Carotene and Retinol Efficacy Trial.
*Unadjusted odds ratio

5
 Routine Vitamin Supplementation to Prevent Cancer

Table 1. Randomized controlled trials of vitamin supplementation to prevent cancer

(vitamin E and multivitamin and antioxidant combinations) (continued)

Study, Duration of follow-up;

publication Setting/population Treatment Follow-up rate

Vitamin E

ATBC Study Group SW Finland; 20 mg/day 5–8 years;

19949 29,133 beta-carotene; median 6.1 years
aged 50–69 years 50 mg/day
male smokers alpha-tocopherol
in 2 x 2 factorial design

Multivitamin and antioxidant combinations

Greenberg 6 US gastroenterology clinics; 1 g/day vitamin C and 3 years between year 1 and
199411 864 subjects; 400 mg/day vitamin E vs year 4; colonoscopies
Polyp Prevention 79% male; placebo; cointervention (mean=36.6 months,
Study < 80 years; 25 mg/day beta-carotene no difference between groups)
at least 1 polyp or placebo
in 2 x 2 factorial design

Omenn 4,060 West Coast male 30 mg/day beta-carotene 5.5 years

199614 asbestos workers; and
CARET 14,254 US male and female 25,000 IU/day retinol or
heavy smokers (66% male placebo
overall)

McKeown-Eyssen Toronto hospital; 400 mg/day vitamin C 2 years

199825 185 subjects; 60% male; and 400 mg/day vitamin
at least 1 polyp E or placebo

Ponz de Leon 199726 Italy; Antioxidant vitamins 36 months

Italian Polyp 255 subjects; (30,000 IU vitamin A,
Prevention Study % male not known; 70 mg vitamin E, 1 g
at least 1 polyp vitamin C per day);
lactulose (crystalline
powder or syrup);
placebo

Note: ATBC indicates Alpha-Tocopherol Beta-Carotene; g, gram; mg, milligram; n, number; FU, followup; NR, not reported;
CARET, Beta-Carotene and Retinol Efficacy Trial.

6
 Routine Vitamin Supplementation to Prevent Cancer

Outcomes (95% confidence interval)

Lung Prostate Breast Colon All-cause
cancer cancer cancer cancer Other cancer mortality

0.98 Incidence: NR Incidence: Cancer mortality: 1.05 1.02

(0.86–1.12) 0.68 0.78 Pancreatic cancer (0.95–1.09)
(0.53–0.88) (0.55–1.09) incidence: 1.34 (0.88–2.05)
Mortality: Adenomatous Pancreatic cancer
0.59 polyps mortality: 1.11 (0.72–1.72)
(0.35–0.99) recurrence:
1.66
(1.19–2.32)

NR NR NR Adenomatous NR NR
polyps
recurrence:
1.08
(0.91–1.29)

Incidence: Incidence: Incidence: Incidence: 1.02 Incidence: 1.17

1.28 1.01 0.78 (0.70–1.50) Urinary/bladder: 0.08 (0.69–1.70) (1.03–1.33)
(1.04–1.57) (0.80–1.27) (0.55–1.12) Head/neck: 1.26 (0.73–2.19)
Mortality: Leukemia: 2.18 (0.95–5.03)
1.46 Lymphoma: 0.91 (0.42–1.98)
(1.07–2.00) Mesothelioma: 1.52 (0.66–3.52)

NR NR NR Adenomatous NR NR
polyps
recurrence: 0.86
(0.51–1.45)

NR NR NR Adenomatous NR NR
polyps
recurrence: 0.16
0.04–0.46)

Note: ATBC indicates Alpha-Tocopherol Beta-Carotene; g, gram; mg, milligram; n, number; FU, followup; NR, not reported;
CARET, Beta-Carotene and Retinol Efficacy Trial.

7
 Routine Vitamin Supplementation to Prevent Cancer

found no effect on the relative risk for recurrence was greater among men with more pack-years of
in subjects randomized to beta-carotene vs placebo smoking, differing from the lung cancer result.
(adjusted RR, 1.01; 95% CI, 0.85 to 1.20), while
How valid is this result likely to be? The
an Australian trial12 was terminated after 2 years
examination of multiple individual cancers as
because of a trend toward an increased incidence
secondary endpoints raises the possibility that this
of recurrent polyps in the beta-carotene group
is a chance finding. Comparability between groups
(unadjusted odds ratio, 1.5; 95% CI, 0.9–2.5).
in the baseline risk for prostate cancer was only
partially ascertainable; this is a concern because
Breast Cancer baseline differences in risk cannot be ruled out in a
The Women’s Health Study found no effect of trial that used geographically defined blocks as the
beta-carotene supplementation on breast cancer unit of randomization. Geographic differences in
(RR, 1.01; 95% CI, 0.81–1.24).10 rates of prostate procedures, such as transurethral
resection of the prostate, could also result in
All-cause Mortality spurious differences in incidence. This ascertainment
Beta-carotene supplementation had no effect on bias could also affect the likelihood that a death was
all-cause mortality in the ATBC (RR, 1.08; 95% CI, attributed to prostate cancer. These biases cannot be
1.01–1.16),9 the Physicians’ Health Study (RR, 1.01; ruled out, but, if they are present, one would expect
95% CI, 0.93–1.10),8 the Women’s Health Study them to affect the results of beta-carotene as well
(RR, 1.07; 95% CI, 0.74–1.56),10 and a skin cancer as vitamin E. As noted earlier, beta-carotene had
prevention study (RR, 1.03; 95% CI, 0.82–1.30).20 no protective effect on prostate cancer incidence
or mortality in ATBC; in fact, there was a trend
toward harm.
Vitamin E
The only randomized controlled trial data on
vitamin E supplementation and cancer risk come
Colon Cancer and Polyps
from the ATBC trial, which included only male In 2 reports from the ATBC trial, vitamin E
smokers. No studies in women have been supplementation did not reduce colon cancer
completed. incidence significantly (RR, 0.78; 95% CI,
0.55–1.09).19 Vitamin E supplementation increased
the risk for adenomatous polyps (RR, 1.66; 95%
Lung Cancer CI, 1.19–2.32).23 This finding might have been
In ATBC, vitamin E had no effect on the due to differences in rates of detection because the
primary endpoint, lung cancer incidence (RR, 0.98; patients diagnosed to have polyps who received
95% CI, 0.86–1.12). vitamin E supplement were more likely to have
pre-diagnosis rectal bleeding and intestinal pain,
Prostate Cancer symptoms which may have led to higher rates of
colonoscopy.
In ATBC, the effect of vitamin E supplemen-
tation on incidence and mortality from several
cancers was examined as part of a planned Other Cancers
secondary endpoint analysis.21 Patients who took In ATBC, vitamin E supplementation was not
supplemental vitamin E had a lower incidence of associated with a significant difference in the
prostate cancer than those who did not (RR, 0.66; incidence of stomach cancer (70/14,564 in those
95% CI, 0.44–0.94; number of cases 99/14,564 receiving vitamin E compared with 56/14,569;
compared with 151/14,569) and also had lower P=0.21).21 It had no effect on pancreatic cancer
mortality from prostate cancer (RR, 0.59; 95% CI, incidence (RR, 1.34; 95% CI, 0.88–2.05) or
0.35–0.99).16, 22 Protection against prostate cancer mortality.24

8
 Routine Vitamin Supplementation to Prevent Cancer

All-cause Mortality that supplemental beta-carotene appears to have

The relative risk for all-cause mortality in the no effect on mortality or cancer incidence in the
vitamin E supplemented arm in ATBC was 1.02 general population.
(95% CI, 0.95–1.09). Historically, the results of randomized trials of
vitamin supplements have not always confirmed
Vitamin Combinations those of epidemiological cohort studies. Even for
Three of the studies described above, CARET, well-designed, well-conducted cohort studies, it is
ATBC, and the Polyp Prevention Study,11 and 2 never possible to be certain that the results do not
additional trials of polyp prevention, randomized at reflect the influence of unrecognized confounders.
least 1 arm to a vitamin combination. The results Uncertainty about the adequacy of control for the
from CARET (beta-carotene plus vitamin A) were “healthy user effect” and other potential confounders
discussed above. In ATBC, there was no benefit makes it difficult to decide how much weight to
in the combined antioxidant (beta-carotene plus place on consistent results from cohort studies.
vitamin E) arm for any cancer, and there were no Limitations of the randomized trials may also be
interactions between beta-carotene and vitamin E. responsible for the discrepancy between trials and
observational studies. The timeframe for prevention
Colonic Polyps of chronic diseases may be longer than the follow-
Studies of the efficacy of vitamin combinations up period in the trials, or the trials may have
to prevent colonic neoplasia had mixed results. examined the efficacy of supplements at the wrong
The Polyp Prevention Study, a US multi-center time in the natural history of the disease. The trials
trial, used a factorial design to evaluate beta- of beta-carotene, for example, recruited subjects at
carotene (described above) and vitamin C plus high risk for developing lung cancer in order to
vitamin E.11 After 4 years on study (3 years between increase statistical power. The subjects had smoked
colonoscopies), the relative risk for polyp recurrence tobacco or been exposed to asbestos for many years
for the vitamin C plus vitamin E supplement vs before the efficacy of supplements was tested. This
placebo was 1.08 (95% CI, 0.91–1.29). raises the possibility that the cohort studies observed
a benefit from using beta-carotene earlier in the
Two other trials examined the effect of vitamin course of carcinogenesis, at a period that the trials
combinations on colon polyps. In Canada, a have not examined.
randomized trial compared vitamin C plus vitamin
E to lactulose and placebo.25 After 2 years of The trials in nonsmokers may have the opposite
follow-up, there was no significant effect on polyp problem: the subjects included in the trials may
recurrence in the group randomized to vitamins C have been at lower risk than those observed to
plus E compared with placebo (RR, 0.86; 95% CI, benefit in the cohort studies. In some of the
0.51–1.45). A small Italian trial evaluated a daily epidemiological studies, supplements were
dose of vitamin A plus vitamin C plus vitamin E associated with a reduced risk for cancer only in
vs placebo in 150 participants.26 After a 3-year the subgroup of subjects who had low baseline
follow-up, the relative risk for polyp recurrence in intake or serum levels of antioxidant vitamins.5, 27,
28 As noted earlier, a similar observation was made
the intervention group compared with the control
group was 0.16 (95% CI, 0.04–0.46). in the Physicians’ Health Study, in which beta-
carotene supplementation was associated with a
reduced incidence of prostate cancer in subjects
Discussion who were relatively deficient in beta-carotene
The main findings of this review are summarized at baseline.17, 18 The trials may have limited
in Table 2. The strongest finding is that beta- the chance of finding a benefit by focusing on
carotene supplements and combinations including relatively well-nourished subjects selected by
beta-carotene are harmful in smokers and others at profession rather than by baseline nutritional
high risk for lung cancer. Another strong finding is measures.

9
 Routine Vitamin Supplementation to Prevent Cancer

Table 2. Summary of results of randomized trials

Vitamin(s) Outcome(s) Strength of evidence, comment

Beta-carotene or Increased incidence and mortality Consistent results from good
beta-carotene from lung cancer in smokers randomized trials
plus vitamin A
Increased incidence and mortality from One trial suggests harm
mesothelioma in high-risk group

No effect on lung cancer in the Consistent results from good

general population randomized trials

No benefit for prostate cancer in Mixed results, requiring additional study

the general population

No benefit for colon cancer in the Consistent results from good

general population randomized trials

No effect on the incidence of Results from two trials suggest

colonic polyps in subjects at high no benefit. (One suggests harm.)
risk for polyps

No effect on breast cancer in the Results from one good trial,

general population requiring additional study

No effect on all-cause mortality Consistent results from good

in the general population randomized trials

Vitamin E No effect on lung cancer, colon cancer, Results from one good trial,
cancer mortality, total mortality in smokers requiring additional study

Reduced incidence and mortality Results from one good trial,

from prostate cancer and stomach confirmatory trial underway
cancer in smokers

Effects in nonsmokers No data from trials. Trials for prostate

and breast cancer underway.

Vitamins C + E No effect on polyp recurrence Consistent results from two trials

Vitamins C + E + A Reduction in polyp recurrence Results from one fair-to-poor trial

Recommendations for cancer (see Appendix I). The cohort studies also
raise the possibility that vitamin A reduces the
Future Research risk for colon cancer in women.
Among the vitamin-cancer combinations that
have evidence from randomized trials, the most The cohort studies evaluating B vitamin
promising finding was that smokers who took supplementation in relation to breast cancer and
vitamin E had a lower incidence of prostate cancer colon cancer also show some evidence of benefit.
and lower mortality from prostate cancer than A randomized controlled trial of the effect of
those who did not. A large US trial is underway to long-term B vitamin supplementation on polyp
confirm the finding that vitamin E might reduce recurrence is currently underway.
mortality from prostate cancer.
Three other trials are also in progress. The
Among the vitamin-cancer combinations that do Physicians’ Health Study II has re-randomized many
not yet have evidence from randomized trials, the of the original Physicians’ Health Study participants
strongest evidence from observational studies is for and has recruited additional participants to a 2x2x2
a possible beneficial effect of vitamin A for breast factorial randomized controlled trial comparing

10
 Routine Vitamin Supplementation to Prevent Cancer

vitamin C, vitamin E, and a multivitamin MPH, and Kim Peterson, MS, of the Oregon
(100% of RDA type) with placebo.29 The Women’s Health & Science University Evidence-based
Health Study, focusing on health professionals, is Practice Center, for their contributions to this
a factorial study of beta-carotene, vitamin E, and project. The corresponding author confirms that
aspirin; the beta-carotene arm has been closed, the above acknowledgments include everyone who
but the vitamin E and aspirin arms continue. has contributed significantly to this work.
Finally, the “Supplementation en Vitamines et
Mineraux Antioxydants” study is a population- References
based, randomized trial of over 12,000 subjects. The
1. Cancer Facts and Figures. In: Atlanta, GA:
SU.VI.MAX study is designed to test the efficacy of American Cancer Society; 2003.
a daily supplementation with antioxidant vitamins
(vitamin C, 120 mg; vitamin E, 30 mg; and beta- 2. Evans P, Halliwell B. Micronutrients:
carotene, 6 mg) and minerals (selenium, 100 microg; oxidant/antioxidant status. Br J Nutr.
and zinc, 20 mg) at nutritional doses in reducing 2001;85(Suppl 2):S67–74.
mortality from cancers and cardiovascular diseases.30, 31 3. Choi SW, Mason JB. Folate and carcinogenesis: an
integrated scheme. J Nutr. 2000;130(2):129–132.
Epidemiological cohort studies will continue to be
extremely important in providing guidance regarding 4. Hansen LA, Sigman CC, Andreola F, et al.
the role of vitamin supplementation in the Retinoids in chemoprevention and differentiation
prevention of chronic disease. The largest established therapy. Carcinogenesis. 2000;21(7):1271–1279.
cohorts (Nurses’ Health Study, Health Professionals’ 5. Steinmetz KA, Potter JD. Vegetables, fruit, and
Follow-up Study, Iowa Women’s Study, and Leisure cancer prevention: a review. J Am Dietetic Assoc.
World Study) are now reaching a stage of maturity 1996;96:1027–1039.
in which they can provide information on risks and
benefits associated with behaviors taking place early 6. Harris R, Helfand M, Woolf SH, et al. Methods
in the carcinogenesis process. A problem that will of the third U.S. Preventive Services Task Force;
a review of the process. Am J Prev Med.
continue to plague the epidemiological studies,
2001;20(3S):21–35.
however, is the degree to which supplement users
differ in other ways from non-users, ways that may 7. Jadad AR, Moore RA, Carroll D, et al. Assessing
not be fully accounted for in the multivariate the quality of reports of randomized clinical trials.
analyses. Attempts by scientists to analyze the large Control Clin Trials. 1996;17:1–12.
cohort studies in ways that replicate, to the extent 8. Hennekens CH, Buring JE, Manson JE, et al.
possible, clinical trial designs would be extremely Lack of effect of long-term supplementation with
useful in elucidating the sources of the differences beta-carotene on the incidence of malignant
in findings between clinical trials and cohort studies. neoplasms and cardiovascular disease. New Engl
Understanding the sources of these differences will J Med. 1996;334:1145–1149.
permit us to better use the cohort study data and
9. The Alpha-Tocopherol Beta-Carotene Cancer
to better design long-term clinical trials. Prevention Study Group. The effect of vitamin E
and beta-carotene on the incidence of lung cancer
and other cancers in male smokers. N Engl J Med.
Acknowledgments 1994;330:1029–1035.
The study on which this summary is based was
conducted by the Oregon Health & Science University 10. Lee IM, Cook NR, Manson JE, et al. Beta-carotene
Evidence-based Practice Center, under contract to supplementation and incidence of cancer and
cardiovascular disease: the Women’s Health Study.
the Agency for Healthcare Research and Quality
J Natl Cancer Inst. 1999;91:2102–2106.
(Contract No. 290-97-0018, Task Order No. 2).
11. Greenberg ER, Baron JA, Tosteson TD, et al.
The authors wish to thank Janet Allan, PhD, A clinical trial of antioxidant vitamins to prevent
RN, and Steven Woolf, MD, MPH, of the U.S. colorectal adenoma. Polyp Prevention Study Group.
Preventive Services Task Force and Susan Carson, N Engl J Med. 1994;331:141–147.

11
 Routine Vitamin Supplementation to Prevent Cancer

12. MacLennan R, Macrae F, Bain C, et al. 22. Hartman TJ, Albanes D, Pietinen P, et al. The
Randomized trial of intake of fat, fiber, and beta association between baseline vitamin E, selenium,
carotene to prevent colorectal adenomas. The and prostate cancer in the alpha-tocopherol,
Australian Polyp Prevention Project. J Natl Cancer beta-carotene cancer prevention study. Cancer
Inst. 1995;87:1760–1766. Epidemiol Biomarkers Prev.1998;7:335–340.
13. Alaimo K, McDowell MA, Briefel RF, et al. Dietary 23. Malila N, Virtamo J, Virtanen M, et al. The
intake of vitamins, minerals, and fiber of persons effect of alpha-tocopherol and beta-carotene
ages 2 months and over in the United States: Third supplementation on colorectal adenomas in
National Health and Nutrition Examination Survey. middle-aged male smokers. Cancer Epidemiol
Phase 1, 1988–91: National Center for Health Biomarkers Prev. 1999;8:489–493.
Statistics; 1994.
24. Rautalahti MT, Virtamo JR, Taylor PR, et al. The
14. Omenn GS, Goodman GE, Thornquist MD, et al. effects of supplementation with alpha-tocopherol
Effects of a combination of beta carotene and and beta-carotene on the incidence and mortality
vitamin A on lung cancer and cardiovascular of carcinoma of the pancreas in a randomized,
disease. N Engl J Med. 1996;334:1150–1155. controlled trial. Cancer. 1999;86:37–42.

15. de Klerk NH, Musk AW, Ambrosini GL, et al. 25. McKeown-Eyssen G, Holloway C, Jazmaji V, et al.
Vitamin A and cancer prevention II: comparison A randomized trial of vitamins C and E in the
of the effects of retinol and beta-carotene. Int J prevention of recurrence of colorectal polyps.
Cancer. 1998;75:362–367. Cancer Res. 1988;48(16):4701–4705.

16. Heinonen OP, Albanes D, Virtamo J, et al. Prostate 26. Ponz de Leon M, Roncucci L. Chemoprevention
cancer and supplementation with alpha-tocopherol of colorectal tumors: role of lactulose and of other
and beta-carotene: incidence and mortality in a agents. Scand J Gastroenterol Suppl. 1997;222:72–75.
controlled trial. J Natl Cancer Inst. 1998;90:440–446. 27. Zhang S, Hunter DJ, Hankinson SE, et al. Dietary
17. Stampfer MJ, Cook NR, Hennekens CH. Effects of carotenoids and vitamins A, C, and E and risk of
beta-carotene supplementation on total and prostate breast cancer. J Nat Cancer Inst. 1999;91:547–556.
cancer incidence among randomized participants with 28. Zhang S, Hunter DJ, Hankinson SE, et al. A
low baseline plasma levels: the Physicians’ Health prospective study of folate intake and the risk of
Study. Proceedings of the Annual Meeting of the breast cancer. JAMA. 1999;281(17):1632–1637.
American Society of Clinical Oncology. 1997.
29. Christen WG, Gaziano JM, Hennekens CH. Design
18. Cook NR, Stampfer MJ, Ma J, et al. Beta-carotene of Physicians’ Health Study II—a randomized trial of
supplementation for patients with low baseline beta-carotene, vitamins E and C, and multivitamins,
levels and decreased risks of total and prostate in prevention of cancer, cardiovascular disease, and
carcinoma. Cancer. 1999;86:1783–1792. eye disease, and review of results of completed trials.
Ann Epidemiol. 2000;10:125–134.
19. Albanes D, Malila N, Taylor PR, et al. Effects of
supplemental alpha-tocopherol and beta-carotene 30. Hercberg S, Preziosi P, Briancon S, et al.
on colorectal cancer: results from a controlled trial A primary prevention trial using nutritional
(Finland). Cancer Causes Control. 2000;11:197–205. doses of antioxidant vitamins and minerals in
cardiovascular diseases and cancers in a general
20. Greenberg ER, Baron JA, Karagas MR, et al. population: the SU.VI.MAX study—design,
Mortality associated with low plasma concentration methods, and participant characteristics.
of beta-carotene and the effect of oral Supplementation en Vitamines et Mineraux
supplementation. JAMA. 1996;275:699–703. Antioxydants. Control Clin Trials. 1998;19:336–351.
21. Albanes D, Heinonen OP, Huttunen JK, et al. 31. Hercberg S, Preziosi P, Galan P, et al. “The
Effects of alpha-tocopherol and beta-carotene SU.VI.MAX Study”: a primary prevention trial
supplements on cancer incidence in the Alpha- using nutritional doses of antioxidant vitamins
Tocopherol Beta-Carotene Cancer Prevention Study. and minerals in cardiovascular diseases and
Am J Clin Nutr. 1995;62:1427S-1430S. cancers. Supplementation en Vitamines et
Mineraux Antioxydants. Food Chemical Toxicol.
1999;37:925–930.

12
 Routine Vitamin Supplementation to Prevent Cancer

Appendix I. Longitudinal Cohort Studies

The results of observational cohort studies that 1. The Netherlands Cohort Study.1,2 This study
met the inclusion criteria are summarized in was designed to examine the association between
Appendix Table 1. Major features of each study intake of vitamins C and E, retinol, and beta-
are summarized below. Additional information carotene and risk of breast cancer. Women
about most studies is available in: Morris CD and from the general population aged 50–69 at
Carson S. Vitamin supplementation to prevent baseline were eligible for enrollment. Of the
cardiovascular disease, summary of the evidence 62,573 women enrolled, 96% continued for
for the U.S. Preventive Services Task Force available 3.3 years (median) of follow-up. The study
at http://www.ahrq.gov/clinic/uspstfix.htm. found no relationship between the use of
vitamin C supplements and breast cancer.

Appendix Table I. Summary of cohort studies evidence

Multi-vitamins
Beta- containing
Cancer Vitamin A carotene Vitamin C Vitamin E folic acid
Lung No effect in men No data No effect in Trend toward No data
Cancer and women (LWS) men and benefit but NS
women (LWS) (LWS, HPFS)

Breast Four studies show No data No effect 4 studies No effect

Cancer NS RR estimates shown in 4 show no overall (NHS)
of 0.7–1.0; pooled studies effect but reduced
analysis suggests (pooled OR .98; incidence
possible benefit 95% CI in women
(RR 0.81; 95% 0.85–1.14) who consumed
CI 0.62–0.97). Harm shown > 15 g/d of
in 1 study alcohol (NHS)
(OR 1.46)

Prostate No effect (LWS) No data No effect (LWS) No data

Cancer

Colon Benefit in women No data No effect Lower incidence

Cancer (LWS, IWHS) (LWS, IWHS) observed after
15 years in
women
(RR, 0.25; 95%
CI, 0.13 to 0.51)
(NHS)

Bladder No effect (LWS) No data Benefit (LWS) No data

Cancer

Note: NS indicates not statistically significant; RR, relative risk; CI, confidence interval; LWS, Leisure World Study; NHS, Nurses,
Health Study; IWHS, Iowa Women’s Health Study; HPFS, Health Professionals, Follow-up Study.

13
 Routine Vitamin Supplementation to Prevent Cancer

2. Nurses’ Health Study I.3,4,5,6,7 This study was who were aged 55–69 years at baseline and did
conducted in the United States and designed to not have a history of cancer were eligible for
determine whether vitamins A, C, E, folic acid, participation following completion of a dietary
and multivitamin supplements are related to questionnaire. Of the 34,486 qualified study
risk of colorectal adenomas. Female nurses aged subjects, 84% continued for 5 years of follow-
34–59 at baseline were eligible for enrollment. up participation.
Of the 89,494 women enrolled, 96%
7. National Health and Nutritional
continued for 8 years of follow-up
Examination Survey I.16 This U.S. study was
participation.
designed to examine the relation between
3. Canadian National Breast Screening Study.8 vitamin C intake and cancer and all-cause
This study was conducted in Canada and mortality. Noninstitutionalized U.S. adults
designed to examine the relationships of aged 25–74 years at baseline who were
vitamins A, C, and E to risk of breast cancer. nutritionally examined during 1971–1974 as
Women aged 40–59 at baseline who were part of an epidemiological follow-up study
involved in an ongoing breast cancer screening (NHANES I) were eligible for participation.
study were eligible for enrollment. Of the Of the 14,407 qualified study subjects, 93%
56,837 qualified study subjects, 96% continued for 10 years (median) of follow-up
continued for 5 years of follow-up participation.
participation.
8. Established Populations for Epidemiologic
4. Leisure World Study.9,10 This U.S. study was Studies of the Elderly.17 This study was
designed to examine the relationships of conducted in the United States and designed to
vitamins A, C, and E to risk of lung, colon, examine vitamin E and vitamin C supplement
bladder, prostate, and breast cancers. Males and use in relation to mortality risk and whether
females living in the Leisure World retirement vitamin C enhanced the effects of vitamin E.
community were eligible for enrollment. The Residents of 4 East Coast communities who
average age of the 70,159 qualified study were older than 65 at baseline (range 67–105)
subjects (24,218 men and 45,941 women) was and who were involved in the Established
mid-70s. Four- to 8-year follow-up rates were Populations for Epidemiologic Studies of the
99% for vital status and 85% for incident Elderly in 1984–1993 were eligible for
cancers. Statistical adjustment was made for age enrollment. Of the 11,178 qualified study
and smoking but not for other confounders. subjects, 100% continued for 6 years of active
participation and an additional 2–3 years for
5. Health Professionals’ Follow-up Study.11,12
mortality follow-up. Adjustment was made for
This study was conducted in the United States
alcohol use, smoking history, aspirin use, and
and designed to examine the relationships of
medical conditions.
folic acid, beta-carotene and vitamins C and E
to risk of colon cancer. Male dentists, 9. Cancer Prevention Study II.18 This study was
osteopaths, optometrists, podiatrists, conducted in the United States and designed to
pharmacists, and veterinarians aged 40–75 were determine the relation between multivitamin
eligible for participation. Of the 43,738 use and cancer and all-cause mortality.
qualified study subjects, 96% continued for 4 American Cancer Society volunteers recruited
years of follow-up participation. men and women aged 30 years and older for
participation. The percentage of the 1,063,023
6. Iowa Women’s Health Study.13,14,15 This
(453,962 men and 609,061 women) eligible
study, conducted in the United States, was
study subjects who continued for the 7-year
designed to investigate whether high intakes of
follow-up participation period is unclear.
antioxidant micronutrients (vitamins A, C, and
E, and beta-carotene) protect against colon and 10. Substudy of Polyp Prevention Study.19
breast cancers. Women living in Iowa in 1986 Conducted in the United States, this study was

14
 Routine Vitamin Supplementation to Prevent Cancer

designed to determine the relation between 5. Hunter DJ, Manson JE, Colditz GA, et al. A
folate and multivitamin use and adenomatous prospective study of the intake of vitamins C, E,
polyp recurrence. Adults who were involved and A and the risk of breast cancer. N Engl J Med.
in the Polyp Prevention Study were eligible 1993;329(4):234–240.
for enrollment. Eligible study subjects were 6. Rimm EB, Willett WC, Hu FB, et al. Folate and
79% male and an average age of 60 years. vitamin B6 from diet and supplements in relation
The percentage of the 709 participants that to risk of coronary heart disease among women.
completed the 4-year follow-up colonoscopy JAMA. 1998;279(5):359–364.
is unclear.
7. Zhang S, Hunter DJ, Hankinson SE, et al. A
11. Nurses’ Health Study and Health prospective study of folate intake and the risk of
Professionals’ Follow-Up Study.4 This analysis breast cancer. JAMA. 1999;281(17):1632–1637.
was designed to measure the association of 8. Rohan TE, Howe GR, Friedenreich CM, et al.
folate and methionine to risk for colorectal Dietary fiber, vitamins A, C, and E, and risk of
adenoma. Female nurses aged 34–59 from the breast cancer: a cohort study. Cancer Causes Control.
Nurses’ Health Study started in 1976. Male 1993;4(1):29–37.
dentists, osteopaths, optometrists, podiatrists,
9. Paganini-Hill A, Chao A, Ross RK, et al. Vitamin
pharmacists, and veterinarians aged 40–75
A, beta-carotene, and the risk of cancer: a
from the Health Professionals Follow-up Study, prospective study. J Natl Cancer Inst.
all of whom had undergone an endoscopy, were 1987;79(3):443–448.
eligible for participation, started in 1986.
Combined 4-year follow-up rates for the 9,490 10. Shibata A, Paganini-Hill A, Ross RK, et al. Intake
male subjects and 10-year follow-up rates for of vegetables, fruits, beta-carotene, vitamin C and
the 15,984 females subjects were 96% for vitamin supplements and cancer incidence among
the elderly: a prospective study. Br J Cancer.
standardized questionnaire completion and
1992;66(4):673–679.
91% for response to letter following
adenomatous polyp recurrence. 11. Rimm EB, Stampfer MJ, Ascherio A, et al. Vitamin
E consumption and the risk of coronary heart disease
in men. N Engl J Med. 1993;328(20):1450–1456.
References 12. Chan JM, Stampfer MJ, Ma J, et al. Supplemental
1. Botterweck AA, van den Brandt PA, Goldbohm vitamin E intake and prostate cancer risk in a large
RA. Vitamins, carotenoids, dietary fiber, and the cohort of men in the United States. Cancer
risk of gastric carcinoma: results from a prospective Epidemiol Biomarkers Prev. 1999;8:893–899.
study after 6.3 years of follow-up. Cancer. 2000;
88(4):737–748. 13. Bostick RM, Potter JD, McKenzie DR, et al.
Reduced risk of colon cancer with high intake
2. Verhoeven DT, Assen N, Goldbohm RA, et al. of vitamin E: the Iowa Women’s Health Study.
Vitamins C and E, retinol, beta carotene and Cancer Res. 1993;53(18):4230–4237.
dietary fibre in relation to breast cancer risk:
a prospective cohort study. Br J Cancer. 14. Kushi LH, Fee RM, Sellers TA, et al. Intake of
1997;75(1):149–155. vitamins A, C, and E and postmenopausal breast
cancer. The Iowa Women’s Health Study. Am J
3. Giovannucci E, Stampfer MJ, Colditz GA, et al. Epidemiol. 1996;144(2):165–174.
Multivitamin use, folate, and colon cancer in
women in the Nurses’ Health Study. Ann Intern 15. Zheng W, Sellers TA, Doyle TJ, et al. Retinol,
Med. 1998;29(7):517–524. antioxidant vitamins, and cancers of the upper
digestive tract in a prospective cohort study
4. Giovannucci E, Stampfer MJ, Colditz GA, et al. of postmenopausal women. Am J Epidemiol.
Folate, methionine, and alcohol intake and risk 1995;142(9):955–960.
of colorectal adenoma. J Natl Cancer Inst.
1993;85(11):875–884.

15
 Routine Vitamin Supplementation to Prevent Cancer

16. Enstrom JE, Kanim LE, Klein MA. Vitamin C 18. Watkins ML, Erickson JD, Thun MJ, et al.
intake and mortality among a sample of the Multivitamin use and mortality in a large population
United States population. Epidemiology. study. Am J Epidemiol. 2002;152:149–162.
1992;3(3):194–202.
19. Baron JA, Sandler RS, Haile RW, et al. Folate
17. Losonczy KG, Harris TB, Havlik RJ. Vitamin E intake, alcohol consumption, cigarette smoking,
and vitamin C supplement use and risk of all-cause and risk of colorectal adenomas. J Natl Cancer Inst.
and coronary heart disease mortality in older 1998;90(1):57–62.
persons: the Established Populations for
Epidemiologic Studies of the Elderly. Am J Clin
Nutr. 1996;64(2):190–196.

16
 Routine Vitamin Supplementation to Prevent Cancer

Appendix II. Excluded Trials

Appendix II. Excluded trials

Study Population or nutrients Reason for exclusion

Mackerras, 19991 Women with CIN;
beta-carotene, vitamin C Wrong outcomes

Manson, 19952 WACS; Combination Wrong outcomes

Childers, 19953 Patients with CIN; Folate Wrong outcomes

Hercberg, 19994 Primary prevention; vitamin C, Results not available yet

Hercberg, 19985 vitamin E, and beta-carotene,
SU.VI.MAX Study 6 mg

Lonn, 19966 Vitamin E Wrong outcomes

SECURE

DeMaio, 19927 Post-PTCA vitamin E Wrong outcomes

Christen, 20008 Alternate day beta-carotene, Results not available yet

Physicians Health Study II alternate day vitamin E,
daily vitamin C, and a daily
multivitamin in the prevention
of total and prostate cancer

Butterworth, 19929 Patients with cervical Wrong outcomes

dysplasia, folic acid

Hofstad, 199210 Patients with polyps: calcium, Wrong outcomes (polyp growth)
vitamins A, C, and E, and selenium

Stephens, 199611 Vitamin E Wrong outcomes

CHAOS

Collaborative Group of Vitamin E Wrong outcomes

the Primary Prevention Project12

McLarty, 199213 Vitamin A Wrong outcomes

Miller, 199714 Vitamin E,C, and beta-carotene Wrong outcomes

Paganelli, 199215 Vitamins A,C, and E Wrong outcomes

Tsubono, 199716 Patients with atrophic gastritis Results not available yet

Gruppo Italiano per lo Vitamin E Wrong outcomes

Studio della Sopravvivenza
nell’Infarto miocardico, 199917

Hoogwerf, 200018 Vitamin E Wrong outcomes

Lonn, 19966
Yusuf, 200019

Munoz, 198520 Vitamin A (esophageal dysplasia) Wrong outcomes, population

17
 Routine Vitamin Supplementation to Prevent Cancer

Appendix II. Excluded trials (continued)

Study Population or Nutrients Reason for Exclusion

Stich, 198821 Beta-carotene, vitamin A, and Wrong outcomes
Stich, 198822 beta-carotene+vitamin A (leucoplakia)

Zaridze, 199323 Vitamin A, beta-carotene, Wrong outcomes

and vitamin E (leucoplakia)

Wang, 199424 Retinol and zinc; riboflavin and Wrong interventions and
Taylor, 199425 niacin; vitamin C and molybdenum; possibly wrong outcomes
Dawsey, 199426 and beta-carotene, vitamin E, (gastric dysplasia and gastric
Li, 199327 and selenium cancer and esophageal cancer)

Heart Protection Study, Vitamin E, vitamin C, and beta-carotene Wrong outcomes

199928

Wald, 200129 Folic acid Wrong outcomes

Skin Cancer Prevention Beta-carotene; skin cancer recurrence Wrong outcomes

Study, 199030 (nonmelanoma skin cancer only)

Nambour Study31, 32 Beta-carotene; skin cancer Wrong outcomes

(primary prevention) (nonmelanoma skin cancer only)

Southwest Skin Cancer Retinol, isotretinoin; skin cancer Wrong outcomes

Prevention Study Group33 recurrence (nonmelanoma skin cancer only)

Note: CIN indicates cervical intraephithelial neoplasia; SU.VI.MAX, Supplementation en Vitamines et Mineraux Antioxydants;
SECURE, Study to Evaluate Carotid Ultrasound changes in patients treated with Ramipril and vitamin E; post-PTCA, post-
percutaneous transluminal coronary angioplasty; CHAOS, Cambridge Heart Antioxidant Study.

References primary prevention trial using nutritional doses of

antioxidant vitamins and minerals in cardiovascular
1. Mackerras D, Irwig L, Simpson JM, Weisberg E, diseases and cancers. Supplementation on
Cardona M, Webster F, et al. Randomized Vitamines et Mineraux Antioxydants. Food
double-blind trial of beta-carotene and vitamin C Chemical Toxicol. 1999;37:925–930.
in women with minor cervical abnormalities.
Br J Cancer. 1999;79:1448–1453. 5. Hercberg S, Preziosi P, Briancon S, Galan P, Triol I,
Malvy D, et al. A primary prevention trial using
2. Manson JE, Gaziano JM, Spelsberg A, Ridker PM, nutritional doses of antioxidant vitamins and
Cook NR, Buring JE, et al. A secondary prevention minerals in cardiovascular diseases and cancers
trial of antioxidant vitamins and cardiovascular in a general population: the SU.VI.MAX study—
disease in women. Rationale, design, and methods. design, methods, and participant characteristics.
The WACS Research Group. Ann Epidemiol. Supplementation en Vitamines et Mineraux
1995;5:261–269. Antioxydants. Control Clin Trials. 1998;19:336–351.
3. Childers JM, Chu J, Voigt LF, Feigl P, Tamimi HK, 6. Lonn EM, Yusuf S, Doris CI, Sabine MJ, Dzavik
Ranklin EW, et al. Chemoprevention of cervical V, Hutchison K, et al. Study design and baseline
cancer with folic acid: a phase III Southwest characteristics of the study to evaluate carotid
Oncology Group Intergroup study. Cancer ultrasound changes in patients treated with
Epidemiol Biomarkers Prev. 1995;4:155–159. ramipril and vitamin E: SECURE. Am J Cardiol.
4. Hercberg S, Preziosi P, Galan P, Faure H, Arnaud J, 1996;78:914–919.
Duport N, et al. “The SU.VI.MAX Study”: a

18
 Routine Vitamin Supplementation to Prevent Cancer

7. DeMaio SJ, King SB, di Lembo NJ, Roubin GS, 16. Tsubono Y, Okubo S, Hayashi M, Kakizoe T,
Hearn JA, Bhagavan HN, et al. Vitamin E Tsugane S. A randomized controlled trial for
supplementation, plasma lipids and incidence chemoprevention of gastric cancer in high-risk
of restenosis after percutaneous transluminal Japanese population; study design, feasibility and
coronary angioplasty (PTCA). J Am Coll Nutr. protocol modification. Jpn J Cancer Res.
1992;11:68–73. 1997;88:344–349.
8. Christen WG, Gaziano JM, Hennekens CH. 17. Gruppo Italiano per lo Studio della Sopravvivenza
Design of Physicians’ Health Study II—a nell’Infarto miocardico. Dietary supplementation
randomized trial of beta-carotene, vitamins E and with n-3 polyunsaturated fatty acids and vitamin E
C, and multivitamins, in prevention of cancer, after myocardial infarction: results of the GISSI-
cardiovascular disease, and eye disease, and review Prevenzione trial. Gruppo Italiano per lo Studio
of results of completed trials. Ann Epidemiol. della Sopravvivenza nell’Infarto miocardico. Lancet.
2000;10:125–134. 1999;354:447–455.
9. Butterworth CEJ, Hatch KD, Soong SJ, Cole P, 18. Hoogwerf BJ, Young JB. The HOPE study.
Tamura T, Sauberlich HE, et al. Oral folic acid Ramipril lowered cardiovascular risk, but vitamin E
supplementation for cervical dysplasia: a clinical did not. Cleve Clin J Med. 2000;67:287–293.
intervention trial. Am J Obstet Gynecol.
1992;166:803–809. 19. Yusuf S, Dagenais G, Pogue J, Bosch J, Sleight P.
Vitamin E supplementation and cardiovascular
10. Hofstad B, Vatn M, Hoff G, Larsen S, Osnes M. events in high-risk patients. The Heart Outcomes
Growth of colorectal polyps: design of a Prevention Evaluation Study Investigators. N Engl
prospective, randomized, placebo-controlled J Med. 2000;342:154–160.
intervention study in patients with colorectal
polyps [published erratum appears in Eur J Cancer 20. Munoz N, Wahrendorf J, Bang LJ, Crespi M,
Prev 1993 Mar;2(2):189]. Eur J Cancer Prev. Thurnham DI, Day NE, et al. No effect of
1992;1:415–422. riboflavine, retinol, and zinc on prevalence of
precancerous lesions of oesophagus. Randomised
11. Stephens NG, Parsons A, Schofield PM, Kelly F, double-blind intervention study in high-risk
Cheeseman K, Mitchinson MJ. Randomised population of China. Lancet. 1985;2:111–114.
controlled trial of vitamin E in patients with
coronary disease: Cambridge Heart Antioxidant 21. Stich HF, Hornby AP, Mathew B,
Study (CHAOS). Lancet. 1996;347:781–786. Sankaranarayanan R, Nair MK. Response of oral
leukoplakias to the administration of vitamin A.
12. Collaborative Group of the Primary Prevention Cancer Lett. 1988;40:93–101.
Project. Low-dose aspirin and vitamin E in people
at cardiovascular risk: a randomised trial in general 22. Stich HF, Rosin MP, Hornby AP, Mathew B,
practice. Collaborative Group of the Primary Sankaranarayanan R, Nair MK. Remission of
Prevention Project. Lancet. 2001;357:89–95. oral leukoplakias and micronuclei in tobacco/betel
quid chewers treated with beta-carotene and with
13. McLarty JW. An intervention trial in high-risk beta-carotene plus vitamin A. Int J Cancer.
asbestos-exposed persons. Adv Exp Med Biology. 1988;42:195–199.
1992;320:141–149.
23. Zaridze D, Evstifeeva T, Boyle P. Chemoprevention
14. Miller ER, Appel LJ, Levander OA, Levine DM. of oral leukoplakia and chronic esophagitis in an
The effect of antioxidant vitamin supplementation area of high incidence of oral and esophageal
on traditional cardiovascular risk factors. cancer. Ann Epidemiol. 1993;3:225–234.
J Cardiovasc Risk. 1997;4:19–24.
24. Wang GQ, Dawsey SM, Li JY, Taylor PR, Li B,
15. Paganelli GM, Biasco G, Brandi G, Santucci R, Blot WJ, et al. Effects of vitamin/mineral
Gizzi G, Villani V, et al. Effect of vitamin A, C, supplementation on the prevalence of histological
and E supplementation on rectal cell proliferation dysplasia and early cancer of the esophagus and
in patients with colorectal adenomas. J Natl Cancer stomach: results from the General Population Trial
Inst. 1992;84:47–51. in Linxian, China. Cancer Epidemiol Biomarkers
Prev. 1994;3:161–166.

19
 Routine Vitamin Supplementation to Prevent Cancer

25. Taylor PR, Li B, Dawsey SM, Li JY, Yang CS, 30. Greenberg ER, Baron JA, Stukel TA, Stevens MM,
Guo W, et al. Prevention of esophageal cancer: Mandel JS, Spencer SK, et al. A clinical trial of
the nutrition intervention trials in Linxian, China. beta carotene to prevent basal-cell and squamous-cell
Linxian Nutrition Intervention Trials Study Group. cancers of the skin. The Skin Cancer Prevention
Cancer Res. 1994;54:2029s-2031s. Study Group [published erratum appears in N Engl
J Med 1991 Oct 31;325(18):1324]. N Engl J Med.
26. Dawsey SM, Wang GQ, Taylor PR, Li JY, 1990;323:789–795.
Blot WJ, Li B, et al. Effects of vitamin/mineral
supplementation on the prevalence of histological 31. Green A, Williams G, Neale R, Hart V, Leslie D,
dysplasia and early cancer of the esophagus and Parsons P, et al. Daily sunscreen application and
stomach: results from the Dysplasia Trial in beta-carotene supplementation in prevention of
Linxian, China. Cancer Epidemiol Biomarkers Prev. basal-cell and squamous-cell carcinomas of the
1994;3:167–172. skin: a randomised controlled trial. Lancet.
1999;354:723–729.
27. Li JY, Li B, Blot WJ, Taylor PR. [Preliminary
report on the results of nutrition prevention 32. Ambler JS, Hirst LW, Clarke CV, Green AC.
trials of cancer and other common diseases The Nambour study of ocular disease. I. Design,
among residents in Linxian, China.] [Chinese.] study population and methodology. Ophthalmic
Chung-Hua Chung Liu Tsa Chih [Chinese Epidemiol. 1995;2:137–144.
Journal of Oncology]. 1993;15:165–181.
33. Levine N, Moon TE, Cartmel B, Bangert JL,
28. MRC/BHF Heart Protection Study of Rodney S, Dong Q, et al. Trial of retinol and
cholesterol-lowering therapy and of antioxidant isotretinoin in skin cancer prevention: a randomized,
vitamin supplementation in a wide range of double-blind, controlled trial. Southwest Skin
patients at increased risk of coronary heart disease Cancer Prevention Study Group. Cancer Epidemiol
death: early safety and efficacy experience. Biomarkers Prev. 1997;6:957–961.
Eur Heart J. 1999;20:725–741.
29. Wald DS, Bishop L, Wald NJ, Law M, Hennessy
E, Weir D, et al. Randomized trial of folic acid
supplementation and serum homocysteine levels.
Arch Intern Med. 2001;161:695–700.

20
 Routine Vitamin Supplementation to Prevent Cancer

Appendix III. Additional Information on

Randomized Controlled Trials of Vitamin
Supplementation to Prevent Cancer
(Supplement to Table 1)
Appendix III. Additional information on randomized controlled trials of vitamin supplementation
to prevent cancer (Supplement to Table 1)

Beta-Carotene

Study, Treatment Control # Enrolled/ Other nutrients Compliance Jadad

Year (n) (n) Exclusion criteria # screened supplemented method, rate score
Hennekens 11,036 11,035 History of cancer, 22,071 enrolled Aspirin; 4% on Pill count; 4
19968 MI, stroke, placebo took 78% at
Physicians’ cerebral ischemia BC or vitamin A year 12
Health Study

Alpha- 13,602 13,669 Prior MI, severe angina, 42,957 Co-intervention Pill count 99% 5
Tocopherol proven malignancy, screened; with vitamin E
Beta- chronic renal 29,246 enrolled;
Carotene insufficiency, cirrhosis, 29,133
(ATBC) Study alcoholism, randomized;
Group, 19949 anticoagulant therapy, 22,269 included
any current use of vitamin
A, vitamin E, BC, other
serious medical illness

Greenberg BC, 184; 187 FAP, history of cancer, 2,092 screened; See Table 1 82% at year 4 5
199411 vit C+E, 205; malabsorption, 981 signed;
Polyp all, 175 renal calculi, 864 randomized
Prevention thrombophlebitis
Study

MacLennan BC, 53; 48 Chronic IBD, GI tract 2,780 cx; 1304 See Table 1 No data 3
199512 BC+low fat, resection, FAP, potential; 559
Australian 51; cancer history, renal, eligible; 424
Polyp BC+C+ liver, GB disease randomized
Prevention bran, 47;
Study low fat+
bran+BC, 50

I-M Lee BC, 19,937 History of cancer, 65,169 eligible; Vitamin E After 2 years, 4
199910 19,939 coronary heart disease, 39,876 87% of active
Women’s or cerebrovascular randomized group reported
Health Study disease taking at least
2/3 of capsules

21
 Routine Vitamin Supplementation to Prevent Cancer

Appendix III. Additional information on randomized controlled trials of vitamin supplementation

to prevent cancer (Supplement to Table 1) (continued)

Beta-Carotene (continued)

Study, Treatment Control # Enrolled/ Other nutrients Compliance Jadad

Year (n) (n) Exclusion criteria # screened supplemented method, rate score

DeKlerk BC, 512; none 1,203 joined; Pill counts at 3

199815 retinyl 1,024 1 year, 100%.
Mesothelioma palmitate, randomized Process
Prevention 512 discontinued
Study

Vitamin E
Vitamin E See above
ATBC Study9

Multivitamin and antioxidant combinations

Greenberg See above
199411
Polyp
Prevention
Study

Omenn 2,044 2,016 Limited vitamin A use 4,060 workers Retinol and BC Weighed 4
199614 asbestos asbestos and no BC supplement exposed to co-intervention; capsules or
CARET workers; workers; use asbestos; 2% took additional self-report in
7,376 6,878 14,254 heavy BC; 1% took 15%
smokers smokers smokers (44% vitamin A
female)
randomized

McKeown- 96 89 Unwilling to stop 200 signed; Intervention group 95% based on 5

Eyssen supplementation; 185 randomized more likely to have urinary ascorbate
198825 no adenomatous polyps taken supplement in intervention
(n=15) prior to group; slightly
randomization fewer completed
in placebo group

Ponz de 70 Lactulose 255 randomized 3 follow-up 3

Leon 199726 61; no colonoscopies:
Italian Polyp treatment, 6–8 months,
Prevention 78 12–18 months,
Study 24–36 months

Note: n indicates number; ATBC, Alpha-Tocopherol Beta-Carotene; MI, myocardial infarction; BC, beta-carotene; FAP, familial adenomatous
polyposis; GB, gall bladder; CARET, Beta-Carotene and Retinol Efficacy Trial.

AHRQ Pub. No. 03–523C

June 2003

22