| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Epidemiology and Surveillance Research, American Cancer Society, Atlanta, Georgia 30329-4251
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This analysis focuses on vitamin C or E intake specifically from individual vitamin C or E supplements, rather than intake from diet, multivitamins, or from all sources combined. Most individual vitamin C or E supplements contain doses several times greater than those typically obtained from diet or multivitamins.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The vital status of study participants was determined through December 31, 1996 using two approaches. American Cancer Society volunteers made personal inquiries in September 1984, 1986, and 1988 to determine whether the participants they had enrolled were alive or dead and to record the date and place of all deaths. Reported deaths were verified by obtaining death certificates. Automated linkage using the National Death Index (13) extended follow-up of the entire cohort through December 31, 1996 and identified deaths among the 21,704 participants lost to follow-up between 1982 and 1988. At the completion of follow-up in December 1996, 237,452 participants had died (20.0%), 944,313 were alive (79.7%), and 2,892 (0.2%) had follow-up truncated on September 1, 1988 because of insufficient data for National Death Index linkage. Death certificates or codes for cause of death were obtained for 98.6% of all deaths. The underlying cause of death was coded from death certificates according to the ninth revision of the International Classification of Disease (ICD-9; Ref. 14 ). Colorectal cancer deaths were defined as ICD-9 codes 153.0–153.9 (colon cancer) or 154.0–154.9 (rectal cancer).
We excluded participants who reported a history of cancer other than nonmelanoma skin cancer (n = 82,349) or did not provide complete information on frequency and duration of use for all four vitamin supplements (vitamin A, vitamin C, vitamin E, and multivitamins) listed on the questionnaire (n = 158,180). We also excluded participants who reported less than weekly use of one or more of the vitamin supplements (n = 110,684) because it is unclear whether such irregular use at enrollment is likely to have adequately represented long-term patterns of vitamin use. In addition, we excluded 17,628 participants who reported identical frequency and duration for all four vitamin supplements because it is possible that many of these participants were reporting components of their multivitamin on the questions intended to measure use of individual vitamin supplements. Finally, we excluded 103,925 participants with missing information on diet, education, or BMI. Analyses are based on the remaining 711,891 participants (334,125 men and 377,766 women).
Ascertainment of Vitamin Supplement Use.
Vitamin supplement use was ascertained from the 1982 baseline
questionnaire, which included a section asking about duration and
frequency of current use of four vitamin supplements (vitamin A,
vitamin C, vitamin E, and multivitamins). Specifically, participants
were asked to fill in two boxes for each vitamin, with the first box
reporting the number of times in the last month they had used this
vitamin, and the second box reporting the number of years of use. In
this analysis, we considered participants reporting use of a particular
vitamin 25 or more times during the past month to be "daily users."
No information was collected on the dose or brand of vitamin
supplements, use of any other dietary supplements, or any past vitamin
supplement use that had been discontinued before study enrollment.
Statistical Analysis.
We used Cox proportional hazards modeling to examine the association of
vitamin C and E supplement use with colorectal cancer mortality while
adjusting for other potential risk factors. The time-axis used was
follow-up time since enrollment in 1982. Age adjustment was
accomplished by stratifying on exact year of age at enrollment within
each Cox model (15)
. All Cox models were also adjusted for
educational level, BMI, use of aspirin and estrogen replacement
therapy, and consumption of two food groups (vegetables and high-fiber
grain foods) associated with decreased risk of colorectal cancer
mortality in this cohort (16)
. All covariates, other than
age, were modeled as dummy variables using the categories shown in
Table 1
. Women who left all of the questions on postmenopausal estrogen use
blank were categorized as never users. Food consumption variables were
derived from items on the questionnaire that asked about the frequency
of consumption of 32 common food items. The derivation of food group
variables from this questionnaire has been described previously
(16)
. Of particular relevance to this analysis, the
vegetable variable was derived by summing the numbers of days per week
that each participant reported eating each of the seven vegetable items
on the questionnaire (carrots, tomatoes, potatoes, squash/corn, green
leafy vegetables, raw vegetables, and cabbage/broccoli/brussels
sprouts). Similarly, the high-fiber grain foods variable was derived by
summing reported consumption of three food items (bran/corn muffins,
brown rice/whole wheat/barley, and oatmeal/shredded wheat/bran
cereals). In combined sex models, we also included interaction terms
between sex and each BMI category to account for the somewhat different
association of BMI with colorectal cancer mortality by sex in this
cohort (17)
. Race, marital status, cigarette smoking,
cigar and pipe smoking, alcohol use, exercise level, consumption of
citrus fruits/juices, consumption of low fat meats (fish and chicken),
consumption of high fat meats (red meat), family history of colorectal
cancer, and personal history of colorectal polyps, diabetes, heart
disease, or hypertension were also examined as potential confounders.
However, we did not adjust for these factors in final models because
such adjustment had virtually no effect on our results. All models also
included dummy variables for each combination of frequency (<25 times
per month, 
25 times per month) and duration of use (<10 years, 10
years) of multivitamins, vitamin A, vitamin C, and vitamin E.
|
We did not examine results stratified by dietary intake of vitamin C or E for two reasons. First, our dietary data may not be sufficiently detailed to provide accurate estimates of these micronutrients. Second, for an analysis stratified by dietary intake of vitamin C or E to be meaningful, we would first need to exclude users of multivitamins, which contain considerable amounts of vitamin C and E. After this exclusion, our statistical power to investigate the effects of vitamin C or E would be limited because many vitamin C or E supplement users also used multivitamins.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Table 1
compares baseline characteristics of participants who were
daily users of individual vitamin C or E supplements with baseline
characteristics of participants who reported no use of either vitamin C
or E supplements. The great majority of participants were white and
middle-aged or elderly, regardless of vitamin use. However, daily users
of vitamin C or E supplements were slightly more likely than nonusers
to be white. Compared with nonusers, daily vitamin C or E users were
much more likely to use multivitamins and slightly more likely to be
well educated, to have low BMI, to report frequent consumption of
vegetables and high-fiber grain foods, and to use aspirin regularly.
Male daily vitamin C or E users were slightly less likely to be current
cigarette smokers, and female daily vitamin C or E users were somewhat
more likely to be former cigarette smokers. Female daily vitamin C or E
users were also more likely than nonusers to be current or former users
of estrogen replacement therapy. In this cohort, as in other
epidemiological studies, aspirin use (19)
, estrogen
replacement therapy (20)
, physical activity,
and consumption of vegetables and high-fiber grains (16)
are associated with reduced risk of colon cancer mortality, whereas
high BMI is associated with increased risk (17)
.
Overall, we found no evidence of reduced risk of colorectal cancer
mortality associated with daily use of individual vitamin C or E
supplements, even for long-term use (Table 2)
. However, there was a suggestion of decreased risk for men with daily
vitamin C use of 10 or more years at enrollment (RR = 0.75; 95%
CI, 0.57–1.00), but not for women (RR = 1.05; 95% CI,
0.80–1.38). There were few participants who reported less than daily
vitamin supplement use (<25 uses per month). Therefore, results for
less-than-daily vitamin supplement use were not informative and are not
shown. Results were similar when we excluded participants reporting
weight loss of 10 or more pounds during the year before study
enrollment or when we excluded the first 2 years of follow-up (data not
shown). RRs adjusted only for age and sex were somewhat lower than the
multivariate adjusted RRs shown in Table 2
[age- and sex-adjusted
RR = 0.77 (95% CI, 0.66–0.90) for 10 years of vitamin C use;
age- and sex-adjusted RR = 0.86 (95% CI, 0.71–1.04) for 10
years of vitamin E use].
|
shows multivariate-adjusted RRs for colorectal cancer mortality
associated with vitamin C and vitamin E supplement use by attained age,
multivitamin use, follow-up year, and tumor subsite (colon or rectum).
Long-term use (10 or more years) of vitamin C was associated with
substantially decreased risk of colorectal cancer mortality before age
65 years (RR = 0.48; 95% CI, 0.28–0.81), but not at age 65 years
or after (RR = 1.02; 95% CI, 0.83–1.26; P =
0.009 for interaction by age group). Long-term vitamin C use was also
associated with decreased risk of rectal cancer mortality at all ages
(RR = 0.40; 95% CI, 0.20–0.80) but was not associated with
decreased risk of colon cancer mortality (RR = 0.98; 95% CI,
0.80–1.20; P = 0.01 for difference by subsite). There
was some suggestion that long-term vitamin E use was associated with
decreased risk of colorectal cancer mortality in participants who were
current cigarette smokers at enrollment (RR = 0.60; 95% CI,
0.31–1.16) but not in never or former cigarette smokers, although the
interaction between cigarette smoking and long-term vitamin E use was
not statistically significant (P = 0.13). Otherwise, we
found little evidence of differences in the association between
colorectal cancer mortality and vitamin C or E supplement use by age,
multivitamin use, follow-up year, or subsite (Table 3)
or by education
level, BMI, cigarette smoking status, or alcohol use (results not
shown). In addition, we found little evidence of interaction between
daily vitamin C and daily vitamin E use, regardless of whether or not
users of multivitamins (which contain lower doses of vitamin C and E)
were included or excluded.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our results concerning vitamin C supplement use (RR = 0.89; 95%
CI, 0.73–1.09 for 10 years of use) are generally consistent with
results from two cohort studies of supplemental vitamin C and colon
cancer incidence, each of which included about 200 cases. In the Iowa
Women’s Health Study cohort (follow-up from 1986–1990; Ref.
10
), high average daily intake of supplemental vitamin C
(predominantly from individual vitamin C supplements) at study baseline
was not associated with colon cancer risk after adjustment for
potential confounders, including vitamin E intake. In the California
Leisure World retirement community cohort (follow-up from 1981–1989),
intake of any supplemental vitamin C (from either multivitamins or
individual vitamin C supplements) at baseline was not associated with
colon cancer risk in men but was associated with modestly decreased
risk in women (RR = 0.67; 95% CI, 0.45–0.99; Ref.
9
). In contrast, two population-based case-control
studies, a study of colon cancer incidence in the Seattle area with
about 400 cases (21)
and a study of colorectal cancer
incidence in Australia with about 700 cases (22)
, found
high average daily intake of supplemental vitamin C (likely
predominantly from individual vitamin C supplements) to be associated
with decreased risk in both sexes. It is possible that the reduced risk
associated with vitamin C supplement use observed in some of these
studies may be due at least in part to confounding by use of other
dietary supplements because only the Iowa Women’s Health Study
analysis (10)
presented risk estimates adjusted for use of
any other supplemental vitamins.
In subanalyses, we found that long-term vitamin C supplement use was associated with decreased risk of colorectal cancer mortality in men and in both sexes before age 65 years. Results from these subanalyses should be interpreted cautiously. No previous study has suggested that vitamin C supplement use is associated with a stronger reduction in risk for men than for women or for younger rather than older people.
In additional subanalyses, we found that long-term vitamin C supplement use was associated with substantially decreased risk of rectal cancer mortality but not colon cancer mortality. The Australian case-control study found significantly decreased risk of both rectal and colon cancer incidence associated with vitamin C use (22) . No other study has examined the effect of vitamin supplements of any type specifically on rectal cancer. We know of no reason to expect a specific effect of vitamin C on rectal rather than colon cancer. When interpreting subsite results, it should be acknowledged that death certificates often misclassify deaths due to rectal cancer as being due to colon cancer and (somewhat less frequently) misclassify colon cancer deaths as being due to rectal cancer (23) . However, such misclassification would be expected to result in an underestimate of any effects that are limited to either the colon or rectum.
We found no association between vitamin E supplement use and risk of
colorectal cancer mortality (RR = 1.08; 95% CI, 0.85–1.38 for
10 years of use). Results from previous epidemiological studies have
been inconsistent. Two of four studies of colon cancer incidence
(9, 10, 11
, 21)
, the Iowa Women’s Health Study cohort
(10)
and the Seattle area case-control study
(21)
, found substantial reductions in risk of colon cancer
associated with high daily intake of supplemental vitamin E
(predominantly from individual vitamin E supplements). In the Seattle
case-control study (21)
, high average daily intake of
supplemental vitamin E over a 10-year period before the reference date
was associated with an OR of 0.4 (95% CI, 0.3–0.7). Similarly, in the
Iowa Women’s Health Study cohort (10)
, high daily intake
of supplemental vitamin E at study baseline was associated with a RR of
0.5 (95% CI, 0.3–0.9). In contrast, there was no association between
vitamin E supplement use and colon cancer incidence in the California
Leisure World cohort (supplement use measured at study baseline; Ref.
9
) or in a large United States multisite case-control
study with nearly 2000 cases (supplement use measured at reference
date; Ref. 24
). A randomized trial in Finnish male smokers
(11)
reported a RR of 0.78 (95% CI, 0.55–1.09) for
colorectal cancer incidence after an average of 6 years of
supplementation and follow-up, but these results may be of limited
relevance due to the short duration of exposure and the use of
relatively low doses of vitamin E (50 IU/day).
Our results for both vitamin C and E supplement use are consistent with results from randomized trials examining the occurrence of new colorectal polyps after colonoscopy. Although an Italian trial found combined vitamin A, C, and E supplementation was associated with decreased risk of colorectal polyps (25) , a Canadian trial (26) and a large United States trial (27) found no reduction in risk after combined supplementation with vitamin C and vitamin E.
There are several possible explanations that should be considered that might explain why we found little association between vitamin C or E supplement use and either colon or colorectal cancer, whereas some previous studies have found substantially decreased risk. First, we examined mortality rather than incidence. In theory, poorer survival from colorectal cancer among vitamin supplement users than among nonusers could account for the lack of association observed in our mortality study, although we know of no reason to expect such a difference. Second, any association between vitamin C or E supplement use and colorectal cancer could be stronger among younger people and therefore less evident in our older study population, in which the median age at death from colorectal cancer was 72 years. This hypothesis is consistent with the reduced risk of colorectal cancer mortality before age 65 years associated with vitamin C supplement use in our study but is less consistent with our results for vitamin E supplement use. Third, there could have been important differences in the dose or duration of vitamin C or E supplement use in our cohort compared with other study populations. However, we know of no reason to believe that vitamin supplement dose varied substantially between our nationwide study population and the other United States study populations, which were studied during similar time periods. Differences in duration of vitamin supplement use also seem to be an unlikely explanation for the lack of association in our study, given that we specifically examined long-term use and found little association. Finally, the differences in results between studies could be due to chance. However, given the large size of our study, it is unlikely that chance alone could have obscured a strong effect of vitamin C or E supplement use.
As in any observational study, the effect of potential confounding factors needs to be considered. This is particularly true for analyses of vitamin supplement use because vitamin users may be more likely to have health conscious behaviors. Although we were able to adjust (or determine that adjustment was unnecessary) for many potential confounding factors, we had no information on colorectal cancer screening or use of dietary supplements other than vitamins A, C, and E or multivitamins. In addition, we had only relatively crude measures of diet and physical activity. However, any confounding from these factors would be expected to bias our results toward finding vitamin supplement use to be associated with reduced of risk of colorectal cancer mortality; we observed little evidence of such a reduction in risk.
One limitation of this analysis was that we had no information about past use of vitamin supplements for participants who were former vitamin supplement users at baseline. Our referent group therefore likely included some former vitamin C and E supplement users. Our results could be biased toward the null if a large proportion of our participants were former vitamin C or E users who had used these supplements for a long enough period of time to affect the risk of colorectal cancer mortality. However, such bias is not likely to have obscured a strong effect of vitamin C or E supplement use.
A second limitation was that we had no updated information on vitamin use after study baseline. If vitamin C or E supplement use affects later stages of colorectal carcinogenesis, the lack of updated vitamin use information could have biased our results toward the null during later years of the 14-year follow-up period. However, we found no association between either vitamin C or E supplement use and colorectal cancer mortality during the first 7 years of follow-up, when the lack of updated vitamin use information would not be expected to substantially effect results.
An additional limitation of this analysis is that we had no information on vitamin dose. However, it is likely that the doses of vitamin C and E obtained from individual supplements in our United States study population were similar to those reported during the same time period in the Nurse’s Health Study, which also included participants from throughout the United States. In 1982 (the year the CPS-II questionnaire was administered), 80% of the participants in the Nurse’s Health Study who were taking vitamin C supplements (and knew their dose) reported a dose of 400 mg or more. Similarly, 88% of nurses who were taking vitamin E supplements (and knew their dose) reported a dose of 200 IU or more.3 Regardless of the exact supplement doses used, it is likely that most of our individual vitamin C or E supplement users had total vitamin C and E intakes several times higher than those receiving vitamin C or E only from diet and/or multivitamins. Mean daily dietary intake in the United States is approximately 100 mg of vitamin C and 10–15 IU of vitamin E (28, 29, 30) , and multivitamins typically contain 60 mg of vitamin C and 30 IU of vitamin E.
Important strengths of this analysis are its prospective design, large size, and the availability of information on duration of supplement use. The large size of this study allowed us to examine the effect of vitamin supplements stratified by potentially important effect modifiers, such as age, multivitamin use, and follow-up time. In addition, we were able to obtain fairly precise estimates for long-term use of each type of vitamin supplement, even when adjusting for use of other vitamin supplements. No previous prospective study has examined the association between long-term use of vitamin C and E supplements and colorectal cancer.
In conclusion, our results do not support a substantial effect of even relatively long-term use of vitamin C or vitamin E supplements on risk of colorectal cancer mortality. However, our data cannot rule out effects in populations with poor nutrition or effects that are modest or limited to specific subgroups.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
1 To whom requests for reprints should be
addressed, at Department of Epidemiology and Surveillance Research,
American Cancer Society, National Home Office, 1599 Clifton Road NE,
Atlanta, GA 30329-4251. 
2 The abbreviations used are: CPS-II, Cancer
Prevention Study II; RR, rate ratio; CI, confidence interval; BMI, body
mass index; OR, odds ratio.
3 Meir Stampfer, personal communication.
Received 4/19/00; revised 10/12/00; accepted 10/17/00.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-tocopherol on fecal mutagenicity. Mutat. Res., 102: 27-37, 1982.[Medline]
-tocopherol and ß-carotene on colorectal cancer: results from a controlled trial (Finland). Cancer Causes Control, 11: 197-205, 2000.[Medline]
This article has been cited by other articles:
|
|
 |
|
  M. A. Murtaugh, K.-n. Ma, J. Benson, K. Curtin, B. Caan, and M. L. Slattery Antioxidants, Carotenoids, and Risk of Rectal Cancer Am. J. Epidemiol., January 1, 2004; 159(1): 32 - 41. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. K. Kim, S. Sasazuki, S. Sasaki, S. Okubo, M. Hayashi, and S. Tsugane Effect of Five-Year Supplementation of Vitamin C on Serum Vitamin C Concentration and Consumption of Vegetables and Fruits in Middle-Aged Japanese: A Randomized Controlled Trial J. Am. Coll. Nutr., June 1, 2003; 22(3): 208 - 216. [Abstract] [Full Text]
|
|
|
|
 |
|
 M. L. Neuhouser Dietary Supplement Use by American Women: Challenges in Assessing Patterns of Use, Motives and Costs J. Nutr., June 1, 2003; 133(6): 1992S - 1996. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Foote, S. P. Murphy, L. R. Wilkens, J. H. Hankin, B. E. Henderson, and L. N. Kolonel Factors Associated with Dietary Supplement Use among Healthy Adults of Five Ethnicities: The Multiethnic Cohort Study Am. J. Epidemiol., May 15, 2003; 157(10): 888 - 897. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. J. Padayatty, A. Katz, Y. Wang, P. Eck, O. Kwon, J.-H. Lee, S. Chen, C. Corpe, A. Dutta, S. K Dutta, et al. Vitamin C as an Antioxidant: Evaluation of Its Role in Disease Prevention J. Am. Coll. Nutr., February 1, 2003; 22(1): 18 - 35. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. J Lenton, A. T Sane, H. Therriault, A. M Cantin, H. Payette, and J R. Wagner Vitamin C augments lymphocyte glutathione in subjects with ascorbate deficiency Am. J. Clinical Nutrition, January 1, 2003; 77(1): 189 - 195. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Wu, W. C. Willett, J. M. Chan, C. S. Fuchs, G. A. Colditz, E. B. Rimm, and E. L. Giovannucci A Prospective Study on Supplemental Vitamin E Intake and Risk of Colon Cancer in Women and Men Cancer Epidemiol. Biomarkers Prev., November 1, 2002; 11(11): 1298 - 1304. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. C. Willett and M. J. Stampfer What Vitamins Should I Be Taking, Doctor? N. Engl. J. Med., December 20, 2001; 345(25): 1819 - 1824. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
