Main Menu Health Library Antioxidant Model
SELENIUM and CANCER
Abstract: The use of a daily 100 mg supplement of Selenium for a duration of 10 or more years can reduce risk of most types of cancer by at least 40%. Although women may not benefit from selenium during their pre-menopausal years, and part of selenium's potential benefit may be provided by other antioxidant vitamins, Selenium may in combination with Vitamin E provide an even much higher level of benefit. Because Selenium also can reduce risk of cardiovascular diseases, health-interested persons should consider taking a daily 100 mg supplement of Selenium as a potentially important part of their program of health care.
The Health Establishment has invested billions of dollars into the so-called war on Cancer. Yet except for the benefits from a decline in cigarette smoking, very little other real reduction in this hated disease has been accomplished. Yet there is a simple pill that taken everyday by every adult might eventually drop the national cancer rates of all kinds of cancer in half. This is not some wild speculation in a health advertisement for some dubious potion. It is a summarized result from 31 seriously conducted and expensive research comparisons in 14 different research studies - all of those found published that had acceptable margins of error.
This best kept of health 'Secrets' is that risk of cancer can be reduced by taking of a simple 100 mg pill of Selenium each day. And risk of cancer might even be reduced by five or more times by using a combination of dietary supplements of Vitamin E plus Selenium! There are two important caveats, however. The reduction in cancer probably will not take place immediately. It will take 5 years of Selenium supplement taking for a benefit to become clearly visible, 10 years for benefit to be substantial, and 20 years for full benefit to be achieved. And the benefits may accrue mostly to men and post-menopausal women.
Levels of cancer incidence in various geographical areas clearly are related to Selenium levels in local crops. Thus Selenium long has been considered as a possible anti-cancer agent. The extensive Medline data base of medical research articles includes more than 2,000 different mostly biochemical type studies relating Selenium to cancer. Table S provides the results of the 31 risk ratio comparisons of Selenium with risks of various kinds of cancer found in studies directly relating risk of cancer to Selenium for populations of men and women. The nature of this research differs from that conducted for most other vitamins and minerals because most measured risks of cancer were related Selenium in blood or in toenails. Only one study found examined directly the risks as related to actual dietary intake.
This one important study of diet by Clark represented by comparisons #1, 10, and 17 in Table S was a randomized clinical study of the use of 200 mg Selenium supplements administered for 4.5 years and followed up for 1.9 more years. The study included about 1300 patients all of whom had suffered skin cancer but no other cancer. Thus the studied groups presumably had an above average risk of cancer per se. Selenium had no effect on recurrence rate of skin cancers. But use of the 200 mg supplements substantially slowed the development of other and much more serious types of cancer. The risk ratio for Selenium users vs. non-users for all types of internal cancer was 0.50 (range=0.31-0.90) for all kinds cancer. For specific types of cancer for which an adequate numbers of events were available the risks were 0.54 (0.30-0.98) for lung cancer, and 0.37 (0.18-0.71) for prostate cancer. The number of cancer deaths in the Selenium supplemented group were only half of those in the placebo group. These are very large, consistent, and highly significant reductions in risk.
The Clark study results are of particular interest because they were obtained over the short mean or average time to event of only about 3.2 years. Other clinical studies of antioxidants and cancer for this time period usually could not find any effect because the biochemical processes involved operate over very long time periods. It seems likely that the selected populations that had skin cancer were able to develop other cancers without the usual long time lag usually needed for measurement of a cancer risk. If an average population had been used, it seems unlikely that the selenium would benefit risk of cancer within this short time period. However, it remains possible that Selenium might protect via some other mechanism that can provide a direct anti-cancer effect in short time.
Most other research evaluated Selenium from its concentration in serum or blood. Concentration of vitamins and mineral factors in the blood usually are reasonably well correlated with dietary intake. But this is a 'snapshot' measurement at one time, and we have no records in any of the studies of the usual concentration of Selenium during the much longer past time periods usually required for cancer to develop. Table S does list some rough estimates of a probable amount of Selenium in diet needed to produce the differences in amounts in blood. But based on information provided for use of some other supplements it seems likely the differences at time usually reflect use in a past 7-10 year period. With this qualification, results of the research still remain impressive.
Seven studies of the effect of Selenium and all types of cancer identified risk ratios varying from 0.11 to 0.87, with all risk ratios statistically consistent. The average risk ratio of this group is a low value of 0.41 with a likelihood of null effect less than 1 in 1000. For specific diseases - and considering study results with a desired margin of error of less than 1.00 - risks for lung, prostrate, and breast cancer averaged 0.37, 0.70, and 0.57 respectively. Of particular interest were observations that a combination of Selenium with Vitamin E produced additive benefits. Study #5 found a risk ratio of only about 0.09, a risk reduction of 11 times for this combination, and the large study of Willet also noted a higher benefit for the vitamin E - Selenium combination. Thus both the clinical study and all studies based on Selenium in blood showed that Selenium reduced risk of cancer by an average of half or more.
The principal 'Contrary' research on Selenium is that from the large Women's Health study on about 63,000 nurses. Selenium was measured by amounts in toenails that have been fairly well correlated with amounts of Selenium in blood. Results of this study are noted for Hunter as #21 for breast cancer and Garland as #8 and #13 for other types of cancer. No benefit for Selenium for either breast or other cancers was found, and these results are clearly statistically inconsistent with the well confirmed and highly significant results of other research. The question here is "Why are these results on the Nurses so different? Four possibilities emerge. First, are the results based on toenail measurements useful? Second, did the nurses as a selected population use much more other antioxidants in diet and supplements than usual? Third, is there a difference for the effect of Selenium on women vs. men? and fourth, this population involved mostly pre-menopausal women whereas other studies involved older women. Did the estrogen rich environment of pre-menopausal women that increases cancer development at about 3% each year remove the potential benefit of antioxidants? A possible involvement of more than one or even all of these theories could contribute to the difference in results which appears well established.
Other study results found no difference between the effect of Selenium on men vs. women. Other studies did find the usually measured effect when using the toenails basis. But the risk reduction from added Selenium obtained by the Nurses probably was reduced by their use of a better than average diet that included somewhat more than usual antioxidants. An average amount of Vitamins C, E, and Beta Carotene in the Nurse's diet was significantly higher than that in the Life Ahead average diet. A Life Ahead computation of the expected advantage for adding a daily 100 mg of Selenium to a diet that matched the Vitamin C, E, and Beta carotene cited in the Nurses diet produced an expected risk ratio for all cancer of only 0.87. Thus this becomes a probable part reason for the different result in the Nurses study.
The fourth above theory - that pre-menopausal women may not obtain much if any reduction of cancer from antioxidants also appears as a possibility. Estrogen provides an added antioxidant environment. Study #11 that found a 0.50 risk ratio for women was done entirely on post-menopausal women, and most women in the Clark study of average age 63 were post-menopausal. Other research appears to have been done on all or mostly post-menopausal women. This mechanism also is further confirmed by negative results from the Women's Health study for vitamins and cancer, and is discussed further there and for the Life Ahead antioxidant model. The Nurses study on cancer and Vitamin E did find a 10% lesser reduction in cancer risk for pre-menopausal than for post-menopausal women. A combination of the higher antioxidants in their diets and fraction of pre-menopausal women involved thus could explain much of the difference in result for the Nurses vs. that from all other populations researched. This problem should be studied further because the long range potential for reducing cancer on our population with Selenium based on all other research to date is enormous.
Life Ahead provides a conservative valuation of the benefits of Selenium in reducing risk of cancer. Its effect for men is taken as duration of use related over 10 years maximum at a risk ratio of 0.9954 per mg-year. This produces a maximum risk ratio of 0.63 for 10 years or more use of 100 mg of Selenium. A maximum of 100 mg/day is accepted for computed benefit from supplements, and no effect of Selenium or other anti-oxidant supplements on cancer is computed for women during their pre-menopausal years. Thus benefits for women below about age 60 may be minimal. The antioxidant model also shows that if adequate amounts of Vitamin E, C, and A are present, no further advantage will be computed for use of Selenium. The program suggests an addition of 100 mg to a usual dietary amount of 90 mg/day, or a target value for best health of 190 mg/day. The RDA for Selenium is cited as 70 mg/day, but it is not clear just what this was based on. This value is much too low for best protection against either heart disease or cancer. And it remains possible that larger amounts of Selenium supplement - at least to 200 mg/day - might provide further benefit.
Highest sources of Selenium in foods are in junk foods, meats, chicken, and fish. A listing of Selenium values for foods in order of amount can be viewed from the Life Ahead food library option in the sort menu display for any diet entry. It may not be practical to design diets sufficiently high in Selenium that do not produce other undesirable nutrients. Thus taking a supplement of at least 100 mg/day would appear to be a most useful action for any health-interested person. Its advantage for coronary disease appears likely for both men and women, there is a large indicated reduction in cancer risk for men, and there is a probable a benefit in reducing cancer risk for post-menopausal women. Although Life Ahead now credits Well-Days for up to 100 mg. of added Selenium, further benefits from up to 200 mg are quite possible, and no adverse problems have been found for use of 200 mg/day f Selenium.
The present Life Ahead basis that assumes Selenium may act mostly as another antioxidant and in combination with the other vitamin antioxidants may be too conservative. The very large reductions in risk noted above for combinations of Selenium and Vitamin E - even if far from adequately confirmed - suggest that Selenium may operate to benefit risk of cancer in part by some alternate mechanism. Thus at this time it seems prudent for a health-interested person to take a 100 mg supplement of Selenium each day even if sufficient other antioxidants are available to produce minimum risk via the present Life ahead method.
Table S
Research on Selenium and Cancer
Study and Reference Sex Risk Ratio Error Margin Basis Est Selenium Diff
High vs. Low 5%-95% in Diet, mg
1. Clark, LC M & W 0.50 0.31-0.80 Supplements 200
JAMA 276:1957 Treatment 4.5 years
in Clinical Trial
2. Knekt,P J Natl M 0.41 P<0.001 Blood 100
Cancer Inst 82:864 Pop 5ths
M 0.11 P<0.001 Pop 10ths 140
W 0.86 P=0.6 Pop 5ths 100
W 0.69 0.47-1.02 Pop 10ths 140
3. Kok, FJ Am J Blood
Epidem 125:12 M 0.37 0.16-0.83 Pop 5ths 100
Only 69 cases for M+W W "No Diff" Pop 5ths 100
4. Fex, J Nutr M 0.26 p<0.05 Blood
Cancer 10:221 Pop 5ths 100
5. Salonen, JT Blood
Br Med J 290:417 M & W 0.17 0.03-0.83 Pop 3rds 80
Memo: Low Selenium + Low Vitamin E: Risk ratio 0.088 (11.4 times)
6. Salonen, JT Am J M & W 0.32 0.15-0.67 Blood 70
Epidemiol 120:342 <45 ug vs higher
7. Willet, C Blood 100
Lancet 2:130 M & W 0.50 0.30-0.91 Pop 5ths
Memo: Also higher benefit with Vitamin E
8. Garland, M J Natl Toenails 100?
Cancer Inst 87:497 W 1.24 0.93-1.65 Pop 5ths
(Includes results on Breast Cancer)
Study and Reference Sex Risk Ratio Error Margin Basis Est Selenium Diff
High vs. Low 5%-95% in Diet, mg
9. Knekt, P Am J M & W 0.44 0.21-0.89 Blood 80
Epidemiol 148:975 Pop 3rds
10. Clark, LC J M & W 0.54 0.30-0.98 Supplements 200
JAMA 276:1957 avg Treatment 4.5 years
in Clinical Trial
Cancer Res 53:4860 Pop 5ths Age 55-69
M 0.50 0.30-0.82 Pop 5ths 100
W 0.40 0.13-1.24 Pop 4ths 80
12. Knekt,P J Natl M 0.30 P=0.001 Blood 100
Cancer Inst 82:864 Pop 5ths
13. Garland, M J Nat W 0.46 0.15-1.44 Toenails 100?
Cancer Inst 87:497 Pop 5ths (probably less)
14. Goodman, GE Cancer
Epidemiol, Biomarkers M & W 1.20 0.77-1.88 Blood
Prev 10:1069 M 1.53 0.83-1.88 Pop 4ths 80
W 0.76 0.29-2.01
Note: Margins of error here are above the level usually acceptable for use in Life Ahead
15. Brooks, JD Blood 80
J Urol 166;2034 M 0.24 0.08-0.77 Pop 4ths
16. Goodman, GE Cancer Blood
Epidemiol, Biomarkers M 1.00 0.65-1.60 Pop 4ths 80
Prev 10:1069
17. Clark, LC J M 0.37 0.18-0.71 Supplements 200
JAMA 276:1957 Treatment 4.5 years
in Clinical Trial
18. Knekt, P J Natl M 0.81 P=0.52 Blood 100
Cancer Inst 82:864 Pop 5ths
Study and Reference Sex Risk Ratio Error Margin Basis Est Selenium Diff
High vs. Low 5%-95% in Diet, mg
19. Knekt, P J Natl W 0.64 P=0.45 Blood 100
Cancer Inst 82:864 Pop 5ths
20. Van’t Veer, P Am J Pop 4ths
Epidemiol 131:987 W 0.63 0.29-1.25 Diet 80
0.50 0.23-1.11 Blood same
0.91 0.42-1.67 Toenails probably less
21. Hunter, DJ W 1.10 0.70-1.72 Toenails 100?
JAMA 264:1128 Pop 5ths (probably less)