TABLE 2.
Association between randomly assigned vitamin E supplementation and the risks of total cancers, site-specific cancers, and mortality in the PHS II1
| Intervention |
Posttrial observation3 |
Overall4 |
|||||
| Outcome | No. in analysis2 | Active/placebo | HR (95% CI) | Active/placebo | HR (95% CI) | Active/placebo | HR (95% CI) |
| No. of events | No. of events | No. of events | |||||
| Total cancers | 14,641 | 1001/970 | 1.04 (0.96, 1.14) | 381/390 | 0.98 (0.85, 1.13) | 1342/1327 | 1.02 (0.95, 1.10) |
| Total epithelial cell cancers5 | 14,641 | 892/887 | 1.02 (0.93, 1.12) | 344/347 | 1.00 (0.86, 1.16) | 1198/1204 | 1.01 (0.93, 1.09) |
| Prostate cancer | 13,980 | 497/520 | 0.96 (0.85, 1.09) | 183/173 | 1.06 (0.86, 1.30) | 680/693 | 0.99 (0.89, 1.10) |
| Prostate cancer deaths | 14,641 | 40/40 | 1.03 (0.66, 1.60) | 37/31 | 1.27 (0.78, 2.04) | 77/71 | 1.13 (0.82, 1.56) |
| Colorectal cancer | 14,519 | 76/88 | 0.88 (0.64, 1.19) | 24/22 | 1.11 (0.62, 1.97) | 100/110 | 0.92 (0.70, 1.21) |
| Colorectal cancer deaths | 14,641 | 22/33 | 0.69 (0.40, 1.18) | 13/8 | 1.67 (0.69, 4.03) | 35/41 | 0.88 (0.56, 1.38) |
| Lung cancer | 14,610 | 50/59 | 0.86 (0.59, 1.26) | 27/26 | 1.06 (0.62, 1.82) | 77/85 | 0.92 (0.68, 1.26) |
| Lung cancer deaths | 14,641 | 46/48 | 0.98 (0.65, 1.47) | 22/22 | 1.04 (0.57, 1.87) | 68/70 | 1.00 (0.71, 1.39) |
| Bladder cancer | 14,570 | 38/33 | 1.16 (0.73, 1.85) | 14/13 | 1.10 (0.52, 2.34) | 52/46 | 1.15 (0.77, 1.70) |
| Bladder cancer deaths | 14,641 | 9/14 | 0.66 (0.28, 1.52) | 5/5 | 1.08 (0.31, 3.75) | 14/19 | 0.76 (0.38, 1.52) |
| Pancreatic cancer | 14,638 | 32/25 | 1.30 (0.77, 2.20) | 11/11 | 1.02 (0.44, 2.35) | 43/36 | 1.22 (0.78, 1.90) |
| Pancreatic cancer deaths | 14,641 | 31/23 | 1.38 (0.80, 2.36) | 11/13 | 0.87 (0.39, 1.93) | 42/36 | 1.20 (0.77, 1.87) |
| Lymphoma | 14,595 | 74/61 | 1.23 (0.88, 1.73) | 22/31 | 0.72 (0.41, 1.24) | 96/92 | 1.06 (0.79, 1.41) |
| Lymphoma deaths | 14,641 | 25/21 | 1.23 (0.69, 2.20) | 9/16 | 0.58 (0.26, 1.31) | 34/37 | 0.95 (0.60, 1.51) |
| Leukemia | 14,612 | 46/34 | 1.38 (0.88, 2.14) | 15/15 | 1.01 (0.50, 2.07) | 61/49 | 1.26 (0.87, 1.84) |
| Leukemia deaths | 14,641 | 22/12 | 1.87 (0.93, 3.79) | 12/13 | 0.95 (0.43, 2.08) | 34/25 | 1.38 (0.83, 2.32) |
| Melanoma | 14,483 | 75/63 | 1.20 (0.86, 1.68) | 32/34 | 0.95 (0.59, 1.54) | 107/97 | 1.11 (0.85, 1.46) |
| Melanoma deaths | 14,641 | 7/7 | 1.04 (0.36, 2.95) | 3/2 | 1.57 (0.26, 9.40) | 10/9 | 1.15 (0.47, 2.83) |
| Total mortality | 14,641 | 843/823 | 1.07 (0.97, 1.18) | 540/551 | 1.03 (0.92, 1.16) | 1383/1374 | 1.06 (0.98, 1.14) |
| Cancer mortality | 14,641 | 295/265 | 1.14 (0.97, 1.35) | 151/148 | 1.05 (0.84, 1.32) | 446/413 | 1.11 (0.97, 1.27) |
HRs (95% CIs) were obtained from Cox proportional hazards models that adjusted for age, the PHS cohort (original PHS I participant, new PHS II participant), and randomized treatment assignment (β-carotene, multivitamin, and either vitamin E or vitamin C) and stratified on baseline cancer. PHS, Physicians’ Health Study.
For total cancers, site-specific mortality, total mortality, and cancer mortality, analyses included all 14,641 participants. For the incidence of site-specific cancers, analyses were restricted to men without that site-specific cancer at baseline.
For analyses of site-specific cancers during posttrial observation, participants were excluded if they developed cancer of the same site that was confirmed or unrefuted by the end of intervention.
The sum of the number of total cancers during the intervention and posttrial observation periods may exceed the number of total cancers overall because of multiple cancers in the same individual.
Epithelial cell cancers were limited to carcinomas, which included all cancers except for lymphoma and leukemia.