Nine RCTs (over 107,000 participants) were included.
Beta-carotene with or without vitamin A (5 RCTs).
Lung cancer in smokers: consistent results from RCTs showed that beta-carotene with or without vitamin A increased the incidence of lung cancer and mortality from lung cancer in smokers. The relative risks (RRs) for lung cancer were 1.19 (95% confidence interval, CI: 1.03, 1.35) (1 RCT with 29,133 male smokers) and 1.28 (95% CI: 1.04, 1.57, P=0.02) (1 RCT with 18,000 asbestos workers and heavy smokers). The RR for lung cancer mortality (1 RCT with 18,000 participants) was 1.46 (95% CI: 1.07, 2.00).
Mesothelioma in high-risk group: one RCT (1,024 asbestos workers) of patients receiving beta-carotene versus vitamin A found that beta-carotene increased mesothelioma and mortality from mesothelioma (RR for incidence 0.24, 95% CI: 0.07, 0.86).
Lung cancer in the general population: consistent results from good RCTs showed no effect of beta-carotene on lung cancer in the general population. The RRs were 0.93 (95% CI: 0.69, 1.25) (1 RCT with 22,071 participants) and 1.43 (95% CI: 0.82, 2.48) (1 RCT with 39,876 participants that was terminated early).
Prostate cancer in the general population: the results from two RCTs were mixed. Both RCTs found no significant difference between beta-carotene and control for all patients randomised, but one RCT found beta-carotene reduced prostate cancer in a subgroup.
Colon cancer in the general population: consistent results from good RCTs showed no benefit from beta-carotene. The RRs were 1.05 (95% CI: 0.75, 1.47) (1 RCT), 0.96 (95% CI: 0.78, 1.18) (1 RCT with 22,071 participants) and 0.99 (95% CI: 0.62, 1.60) (1 RCT with 39,876 participants).
Recurrence of colonic polyps: the results from two RCTs suggested no benefit from beta-carotene on polyp recurrence. The RRs were 1.01 (95% CI: 0.85, 1.20) (1 RCT with 864 participants) and 1.5 (95% CI: 0.9, 2.5) (1 RCT with 424 participants that was terminated early).
Breast cancer in the general population: one good RCT (39,876 participants) found no effect of beta-carotene on breast cancer. The RR was 1.01 (95% CI: 0.81, 1.24).
All-cause mortality in the general population: consistent results from good RCTs showed no effect of beta-carotene on all-cause mortality. The RRs were 1.08 (95% CI: 1.01, 1.16) (1 RCT with 29,133 participants), 1.01 (95% CI: 0.93, 1.10) (1 RCT with 22,071 participants), 1.07 (95% CI: 0.74, 1.56) (1 RCT with 39,876 participants) and 1.03 (95% CI: 0.82, 1.30) (1 RCT of skin cancer prevention).
Vitamin E (1 trial).
One good RCT (29,133 male smokers) found no effect of vitamin E on lung cancer (RR 0.98, 95% CI: 0.86, 1.12), colon cancer (RR 0.78, 95% CI: 0.55, 1.09) or total mortality (RR 1.02, 95% CI: 0.95, 1.09) in smokers. Vitamin E was also found to reduce the incidence of prostate cancer (RR 0.66, 95% CI: 0.44, 0.94) and prostate cancer mortality (RR 0.59, 95% CI: 0.35, 0.99).
No studies in non-smokers were identified.
Vitamins C plus E.
Consistent results from two RCTs showed no effect on polyp recurrence. The RR was 0.86 (95% CI: 0.51, 1.45) (1 RCT with 185 participants) at 2 years and 1.08 (95% CI: 0.91, 1.29) at 4 years (1 RCT with 864 participants).
Vitamins C plus E plus A.
One fair-to-poor RCT (150 people) found that vitamin supplementation reduced polyp recurrence. The RR was 0.16 (95% CI: 0.04, 0.46) at 3 years.