Folic acid (folate) deficiency (erythrocyte folate <140 ng/ml or plasma folate <3 ng/ml) is one of the most common nutrient deficiencies in developed countries and, prior to the fortification of foods with folate, was estimated to affect approximately 10% of the U.S. population.1 Folate is crucial for normal DNA methylation, synthesis and repair, and deficiency of this nutrient is hypothesized to lead to cancer through disruption of these processes.2 Methylation of DNA may block gene expression, and abnormalities in DNA methlyation contribute to the loss of normal controls on protooncogene expression.2 The production of S-adenosylmethionine, the primary methyl donor in the body,3 is dependent on both folate and methionine. When methionine levels are low, more folate is utilized as methyl-tetrahydrofolate to form methionine. This may lower the level of methylene-tetrahydrofolate, which is required for DNA synthesis.2, 4 There is some evidence to suggest that relatively high dietary folate intake might be associated with reduced colorectal cancer risk,5 although few cohort studies have examined this association according to methionine consumption.6–8 Therefore, we examined dietary intake of folate and methionine in relation to colorectal cancer risk in a large cohort of women.
Folate is crucial for normal DNA methylation, synthesis and repair, and deficiency of this nutrient is hypothesized to lead to cancer through disruption of these processes. There is some evidence to suggest that relatively high dietary folate intake might be associated with reduced colorectal cancer risk, especially among individuals with low methionine intake. A case-cohort analysis was undertaken within the cohort of 56,837 women who were enrolled in the Canadian National Breast Screening Study and who completed a self-administered dietary questionnaire. During follow-up to the end of 1993, a total of 389 women were diagnosed with colorectal cancer, identified by linkage to the Canadian Cancer Database. For comparative purposes, a subcohort of 5,681 women was randomly selected from the full dietary cohort at baseline. After exclusions for various reasons, the analyses were based on 295 cases and 5,334 non-cases. Folate intake was inversely associated with colorectal cancer risk (IRR = 0.6, 95% CI = 0.4–1.1, p for trend = 0.25). The inverse association was essentially similar among individuals with low and high methionine intake, and was similar for colon and rectal cancers when those endpoints were analyzed separately. Among individuals with low methionine intake, folate intake did not appear to lower the risk of rectal cancer, a finding that may be due, in part, to the low number of cases in the subgroup analysis. Overall, our data lend some support to the hypothesis that high folate intake is associated with a reduced risk of colorectal cancer. © 2001 Wiley-Liss, Inc.
MATERIAL AND METHODS
The design of our study has been described in detail elsewhere.9 In brief, the investigation was conducted using data from women in the Canadian National Breast Screening Study (NBSS).10 The NBSS is a multicenter randomized controlled trial of mammographic screening for breast cancer in 89,835 women aged 40–59 years at recruitment. Participants were recruited between 1980 and 1985 by various means, including personal invitation by letter, group mailings to employees of large institutions and to members of professional associations, advertisements in newspapers and public service announcements on radio and television. On enrollment in the NBSS, all participants completed a questionnaire that sought data on demographic characteristics, lifestyles, menstrual and reproductive history and use of oral contraceptives and replacement estrogens.
Starting in 1982, a modified version of a previously validated self-administered quantitative food-frequency questionnaire11 was distributed to all new attendees at all screening centers, and to women returning to the screening centers for rescreening. By the time the dietary questionnaire was introduced, some women had already been enrolled in the study and were not seen again at the screening centers. A total of 56,837 women returned completed dietary questionnaires, which asked about the frequency of consumption and usual portion size of 86 food items, and from which nutrient data were derived. Photographs of various portion sizes were included in the questionnaire to assist participants in quantifying intake. A comparison between the self-administered questionnaire and a full interviewer-administered questionnaire, which has been subjected to both validity and reliability testing12 and used in a number of epidemiologic studies,13 revealed that the 2 methods gave estimates of intake of the major macronutrients and dietary fiber that were moderately to strongly correlated with each other (reported correlation coefficients ranged from 0.47 to 0.72).11 Although folate intake was not specifically evaluated, the reported correlation for vitamin C (which was highly correlated with folate in our data) was 0.64.
Data from the self-administered questionnaire were used to estimate daily intake of calories and nutrients, using a nutrient database developed by modifying and extending food composition tables from the United States Department of Agriculture to include typical Canadian foods. The nutrient database has been described in detail elsewhere.13 Nutrients were adjusted for total energy using the residual method.14 Values for folate intake presented here are for intake from dietary sources alone, since data on the folate content of vitamin supplements were not available. However, only 6.2% of the individuals were multivitamin users at assessment. The major sources of dietary folate are liver, green leafy vegetables and whole grains, and the major sources of dietary methionine are meat, milk and grains.
The present investigation was undertaken by performing a case-cohort analysis within the dietary cohort of 56,837 women. For this purpose, a subcohort was constructed by selecting a random sample of 5,681 women from the dietary cohort at baseline. A total of 389 incident colorectal cancer cases were observed during an average of 10.4 years of follow-up, identified by linkage to the National Canadian Cancer Database, and deaths were ascertained by linkage to the National Mortality Database. Analyses were based on the 295 incident case patients (198 colon, 97 rectal) and 5,379 women (including 45 of the case patients) in the subcohort with dietary information and without extreme energy values (± 3 SD of logged energy intake). Poisson regression was used to generate incidence rate ratios (IRR), and robust standard errors were calculated.15 Subjects were followed from their date of enrollment (between 1980 and 1985) until date of diagnosis, date of death or the end of follow-up (December 31, 1993). For tests of trend, median values of each level of categorized variables were fitted in risk models as successive integers.16
The average ages at diagnosis of colon and rectal cancer were 58.3 and 57.7 years, respectively. Folate intake was inversely associated with current smoking and fat intake (r = −0.36) and was positively associated with age, body mass index (BMI; kg/m2), vigorous physical activity, percentage with post-secondary education, alcohol consumption and intakes of dietary fiber (r = 0.53), calcium (r = 0.29) and vitamin C (r = 0.76) (Table I). Energy intake and multivitamin use varied little over strata of folate intake.
|Characteristics||Quintile level of folate intake|
|Person-years of follow-up||114,673||114,292||113,074||113,968||114,072|
|Folate intake (median μg/d ± SD)||204 ± 27||256 ± 12||296 ± 12||338.1 ± 14||416 ± 87|
|Age at baseline (median yr)||48||49||49||50||50|
|Body mass index (median kg/m2)||23.8||23.8||23.8||24.2||24.1|
|Smoking (% current)||28.5||20.5||17.9||17.0||18.3|
|Physical activity (% exercise heavily)||19.8||23.5||23.4||24.9||26.4|
|Education (% post secondary)||18.6||26.7||27.9||33.5||32.6|
|Multivitamin use (% at assessment)||7.0||6.6||5.7||5.6||6.3|
|Dietary intake (median)1|
|Total dietary fiber (g/d)||14.8||17.9||19.9||21.7||25.4|
|Vitamin C (mg/d)||86||132||160||192||239|
Folate intake was weakly inversely associated with colorectal cancer risk in age-adjusted models (IRR for the highest compared with the lowest quintile level = 0.8, 95% CI = 0.5–1.5, p for trend = 0.75). After multivariate adjustment, we observed a stronger inverse association, there being a 40% lower risk among women in the highest compared with the lowest quintile level (Table II). Additional adjustment for vitamin C intake strengthened the inverse association between folate intake and colorectal cancer risk (IRR for the highest compared with the lowest quintile level = 0.5, 95% CI = 0.3–1.1, p for trend = 0.06). Further adjustment for intakes of dietary fiber, calcium, alcohol, total fat, use of oral contraceptives, use of exogenous hormones, parity, age at first birth and height did not alter the rate ratio estimates.
|Study population||Cancer site|
|Folate intake (μg/d)||(n = 295)||(n = 198)||(n = 97)|
|≤233||1.0 (referent)||1.0 (referent)||1.0 (referent)|
|>233–276||0.7 (0.4–1.1)||0.5 (0.3–1.0)||1.1 (0.6–2.3)|
|>276–316||0.9 (0.6–1.4)||0.8 (0.5–1.4)||1.2 (0.6–2.5)|
|>316–367||0.8 (0.5–1.3)||0.8 (0.4–1.4)||0.9 (0.4–2.0)|
|>367||0.6 (0.4–1.1)||0.6 (0.3–1.1)||0.7 (0.3–1.8)|
|p for trend||.25||.41||.36|
|Methionine <2.0 g/d2|
|Folate intake (μg/d)||(n = 147)||(n = 96)||(n = 51)|
|≤233||1.0 (referent)||1.0 (referent)||1.0 (referent)|
|>233–276||0.6 (0.3–1.1)||0.4 (0.2–0.9)||1.3 (0.5–3.2)|
|>276–316||0.9 (0.5–1.6)||0.7 (0.3–1.4)||1.5 (0.6–3.9)|
|>316–367||0.8 (0.4–1.6)||0.7 (0.1–1.6)||1.0 (0.4–2.9)|
|>367||0.6 (0.3–1.3)||0.4 (0.2–1.0)||1.2 (0.4–4.1)|
|p for trend||.37||.26||.99|
|Methionine ≥ 2.0 g/d2|
|Folate intake (μg/d)||(n = 148)||(n = 102)||(n = 46)|
|≤233||1.0 (referent)||1.0 (referent)||1.0 (referent)|
|>233–276||0.7 (0.4–1.8)||0.7 (0.3–1.7)||0.8 (0.2–2.7)|
|>276–316||0.9 (0.5–2.8)||1.0 (0.5–2.3)||0.9 (0.3–3.1)|
|>316–367||0.9 (0.4–2.6)||1.0 (0.4–2.4)||0.7 (0.2–2.4)|
|>367||0.7 (0.3–2.2)||0.9 (0.4–2.3)||0.4 (0.1–2.0)|
|p for trend||.55||.97||.21|
In both age- and multivariate-adjusted models, methionine intake was only weakly inversely associated with colorectal cancer risk. The multivariate-adjusted IRR for the highest (median 2.6 g/d) compared with lowest (median 1.6 g/d) quintile level was 0.8 (95% CI 0.4–1.3, p for trend = 0.43). The inverse association with dietary folate intake was essentially similar among individuals with low and high methionine intake, respectively, and was similar for colon and rectal cancers when these endpoints were analyzed separately (Table II).
The association between folate intake and colon cancer risk did not vary over strata of alcohol consumption (not shown). Excluding multivitamin users (n = 350) or cases that occurred during the first year of follow-up did not alter the results for dietary folate.
An inverse association between folate intake and colon cancer risk was recently observed in a large cohort of American nurses.7 In that study, the association was clearly evident only among women with low methionine intake and only for colon cancer. In addition, 86.3% of women in the highest folate quintile level used multivitamins, which contain micronutrients other than folate (e.g., vitamin D) that may be inversely associated with colorectal cancer risk.17 Multivitamin use in our cohort was low (6.2%) and did not vary across categories of folate intake. Thus, high intake of folate from dietary sources alone may be sufficient to lower the risk of colon cancer.
Few other cohort studies have addressed the association between folate intake and colorectal cancer risk.6–8, 18, 19 Four cohort studies of dietary folate found inverse associations with colon cancer risk,6–8, 18 as did a cohort study of serum folate.19 In the studies that examined the association between folate and colon cancer by methionine intake, the results appeared stronger among men and women with low methionine intake. Of the 3 cohort studies that examined rectal cancer,7, 18, 19 all found either weaker results than for colon cancer19 or no association.18 Thus, the literature is less clear for rectal cancers, possibly because of etiologic differences between colon and rectal cancers or perhaps due to the low statistical power of investigations that have examined rectal cancer risk. The results of case-control studies of folate intake and colorectal cancer risk have varied, although one study that examined the association according to methionine intake20 found evidence of a stronger association among individuals with low methionine intake.
Among the strengths of our study was the large sample size of our cohort of women and the relatively long-term follow-up. The completeness of follow-up of the cohort reduces the likelihood that our results reflect bias due to differential follow-up of exposed compared with unexposed women. It is also unlikely that undiagnosed early stages of colorectal cancer altered folate intake since we observed similar associations after excluding cases that occurred during the first year of follow-up.
Our study was limited by measurement error in our data. Because nondifferential misclassification can attenuate any association that might exist,16 we cannot rule out the possibility that the inverse association between folate intake and colorectal cancer risk would be stronger in the absence of such measurement error. Although foods fortified with folate are now an important source of folate in Canada, the fortification of foods with folate was not commenced in Canada until after the end of follow-up and was therefore not a source of misclassification in our data.
Although we adjusted our estimates for a wide range of potentially confounding variables, uncontrolled confounding from dietary (or other) factors cannot be excluded. In our main analyses, we did not adjust our rate ratios for the potentially confounding effect of vitamin C intake. Univariately, vitamin C intake was positively associated with colorectal cancer risk (although not with statistical significance) and, therefore, adjustment for this variable strengthened the inverse association that we observed with folate intake. (This was especially evident among individuals with low methionine intake.) However, given the strong correlation between folate and vitamin C intake in our data (r = 0.76), the rate ratios observed after mutual adjustment might be the result of collinearity. Therefore, we did not adjust our estimates for vitamin C. Vitamin C intake was not an appreciable source of confounding in a cohort study of colorectal adenoma.21
In conclusion, our results suggest an inverse association between dietary folate intake and colon cancer risk, although dose-risk trends were not very clear. Indeed, our data are not inconsistent with there being an increased risk of colon cancer primarily among individuals with low folate intake, as modest reductions in risk were observed in most intake categories above the lowest. Although the bioavailability of polyglutamyl folate from foods is lower than that of folic acid or pteroyl-monoglutamic acid from supplements,22 folate from dietary sources alone may be important in colorectal cancer prevention.