Article Text

This article has a correction. Please see:

Download PDFPDF

Dietary folate, alcohol and B vitamins in relation to LINE-1 hypomethylation in colon cancer
  1. Eva S Schernhammer1,2,3,
  2. Edward Giovannucci2,
  3. Takako Kawasaki4,
  4. Bernard Rosner1,
  5. Charles S Fuchs1,4,
  6. Shuji Ogino2,4,5
  1. 1Channing Laboratory, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts, USA
  2. 2Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts, USA
  3. 3Ludwig Boltzmann-Institute for Applied Cancer Research, KFJ-Spital, Vienna, Austria and Applied Cancer Research – Institution for Translational Research Vienna (ACR–ITR Vienna), Austria
  4. 4Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts, USA
  5. 5Department of Pathology, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts, USA
  1. Correspondence to Dr Eva S Schernhammer, Channing Laboratory, 181 Longwood Avenue, Boston, MA 02115, USA; eva.schernhammer{at}


Background and aims Although critical for methylation reactions, how dietary folate and B vitamins affect global DNA methylation level in colorectal cancers is currently unknown. Long interspersed nucleotide element-1 (LINE-1) is an emerging indicator of genome-wide DNA methylation level that has previously been linked to colon cancer survival.

Methods We examined the association between dietary intake of folate, alcohol and B vitamins and LINE-1 hypomethylation in 609 incident colon cancers, utilising the database of two independent prospective cohort studies.

Results Participants with ≥400 μg folate intake per day were significantly less likely to develop LINE-1 hypomethylated colon cancers than those reporting <200 μg of folate intake per day (RR=0.57, 95% CI=0.36 to 0.91 for <55% LINE-1 methylated colon tumours; RR=0.74, 95% CI=0.51 to 1.06 for 55–64% LINE-1 methylated colon tumours; and RR=1.08, 95% CI=0.66 to 1.75 for ≥65% LINE-1 methylated tumours; Pinteraction=0.01). By contrast, high alcohol consumption conferred a higher risk of LINE-1 hypomethylated cancers (≥15 g alcohol per day versus none, RR=1.67, 95% CI=1.04 to 2.67 for <55% LINE1 methylated tumours; and RR=1.55, 95% CI=1.10 to 2.18 for 55–64% LINE-1 methylated tumours) but had no association with ≥65% LINE-1 methylated tumours (RR=1.06, 95% CI=0.69 to 1.62). High intakes of vitamin B6, B12 or methionine were not significantly associated with colon cancers, regardless of LINE-1 methylation level.

Conclusion The influence of dietary folate intake and alcohol consumption on colon cancer risk differs significantly according to tumoral LINE-1 methylation level.

  • Methylgroup donors
  • folate
  • vitamin B6
  • colorectal cancer
  • DNA methylation
  • colorectal cancer

Statistics from


DNA methylation is an important epigenetic mechanism that plays a major role in gene silencing, imprinting and repression of endogenous retroviruses.1–3 Genome-wide DNA hypomethylation is believed to play an important role in genomic instability and carcinogenesis.4–9 Several studies indicate a relation between global DNA hypomethylation and chromosomal instability in tumour cells.5 8–12 Moreover, global DNA hypomethylation as determined by repetitive nucleotide elements such as long interspersed nucleotide element-1 (LINE-1) methylation level is inversely correlated with microsatellite instability (MSI) and the CpG island methylator phenotype (CIMP13). In a prior (and to our knowledge, the first) large-scale survival study of 643 colon cancer patients, LINE-1 hypomethylation was associated with poor prognosis.14

Folic acid and related B vitamins (one-carbon nutrients) are essential for DNA methylation and nucleotide biosynthesis; it is therefore plausible that chronic folate deficiency may influence global DNA methylation level. Adequate dietary intake of these nutrients has previously been related to a lower colon cancer risk,15–17 whereas alcohol consumption increases colorectal cancer risk,18 likely through its anti-folate effects.19 Whether one-carbon nutrient intake differentially affects subtypes of colon cancer stratified by global DNA methylation level has not been studied. We therefore assessed whether the influence of folate and B vitamin intake on colon cancer risk differed according to LINE-1 methylation level in two prospective cohort studies in which folate intake has been inversely associated with the risk of colon cancer.18 20

Materials and methods

Study subjects and covariate assessment

Two independent prospective cohort studies, the Nurses' Health Study (121 701 women followed since 197621) and the Health Professionals Follow-up Study (51 529 men followed since 198622), formed the study population. Information on potential risk factors and newly diagnosed cases of cancer was updated biennially. Dietary intake of various nutrients including folate, vitamin B6, B12 and methionine as well as daily alcohol consumption were assessed by self-administered semiquantitative food frequency questionnaires (SFFQ23 24) All nutrient contributions including those from supplements were added to the specific nutrient intake from foods to calculate a daily intake for each participant.23 We assumed an ethanol content of 13.1 g for a 12-ounce (38 dl) can or bottle of beer, 11.0 g for a 4-ounce (12 dl) glass of wine, and 14.0 g for a standard portion of spirits. After excluding participants who did not complete the baseline dietary questionnaire, or who reported a baseline history of cancer (except non-melanoma skin cancer), inflammatory bowel disease, hereditary nonpolyposis colorectal cancer, or a familial polyposis syndrome, 88 691 women and 47 365 men were eligible for analysis.

Ascertainment of colon cancer and deaths

We included colon cancers reported in the Nurses' Health Study and the Health Professionals Follow-up Study on biennial questionnaires between the return of the 1980 or 1986 questionnaires, respectively, and 1 June 2002. With permission from study participants, colon cancer was confirmed through physicians' reviews of the participants' medical records. If permission was denied, we attempted to confirm the self-reported cancer with an additional letter or phone call. We also searched the National Death Index to identify deaths among nonrespondents. The computerised National Death Index is a highly sensitive method for identifying deaths in these cohorts.25 For all deaths attributable to colon cancer, we requested permission from family members (subject to state regulation) to review the medical records. We collected paraffin-embedded tissue blocks from hospitals where colon cancer patients underwent resections of primary tumours.22 In a previous analysis of these cohorts, folate intake was inversely associated with the risk of colon cancer but had no influence on the risk of rectal cancer20; as a result, we did not include incident rectal cancer among the study participants in this analysis. Like rectal cancer cases, cases of colon cancer for which we were unable to assess LINE-1 methylation level were censored from the analyses at their date of diagnosis and were not included as endpoints.

Quantification of tumoral LINE-1 methylation levels

Based on the availability of adequate tissue specimens, we analysed 606 colon cancers for LINE-1 methylation level. Characteristics of those cancers for whom we did and did not analyse for molecular markers have previously been found to be very similar.22 In order to accurately quantify relatively high methylation levels, we utilised Pyrosequencing technology using the PyroMark kit and the PSQ HS 96 System (Qiagen, Valencia, CA) as previously described.13 The nucleotide dispensation order was: ACT CAG TGT GTC AGT CAG TTA GTC TG. Complete conversion of cytosine at a non-CpG site ensured successful bisulfite conversion. The percentage of C relative to the sum of the amounts of C and T at each CpG site was calculated. The average of the relative amounts of C in the four CpG sites was used as overall LINE-1 methylation level in a given sample. Pyrosequencing to measure LINE-1 methylation has been previously validated and shown to be a good indicator of cellular 5-methylcytocine level.13 26 27

Statistical analyses

We used a previously described method of competing risk analysis utilising duplication method Cox regression to compare the specific association of baseline intake of folate and other nutrients with colon cancer risk according to three categories of LINE-1 methylation level (<55%; 55–64%; and ≥65%).28 29 We assessed the statistical significance of the difference between the risk estimates according to tumour type using a likelihood ratio test comparing the model that allowed for separate associations of folate and other nutrients according to LINE-1 methylation level with a model that assumed a common association. To represent interaction effects between dietary folate intake and LINE-1 methylation level, we created models with an indicator variable for LINE-1 methylation level in three categories as well as a product term of this indicator variable and dietary folate intake (continuously), and reported the Wald test statistic of this product term. Established or suspected risk factors for colon cancer were included in the multivariate models, as described at the bottom of table 2. We used SAS version 9.1.3 for all analyses. Tissue collection and analyses were approved by the Harvard School of Public Health and Brigham and Women's Hospital Institutional Review Boards.


Among all 88 691 women and 47 363 men included in these analyses, those with a baseline folate intake of <200 μg/day were slightly more likely to eat meat and to smoke and less likely to exercise or use multivitamins (table 1).

Table 1

Age and age-standardised baseline characteristics of the Nurses' Health Study and Health Professionals Follow-up cohort*

We documented 609 incident cases of colon cancer accessible for LINE-1 methylation data during 2 563 086 person-years. Of these, the LINE-1 methylation levels of 148 (24.3%) tumours were <55%, 265 (43.5%) were 55–64%, and 196 (32.2%) were ≥65%. LINE-1 methylation levels were approximately normally distributed (mean, 61.4%; median, 62.3%; standard deviation, 9.4). Median time interval between baseline folate intake and the diagnosis of incident colon cancer in our analyses was 17.2 years.

As in our previous studies,18 20 30 31 we identified an inverse association between folate and vitamin B6 intake and colon cancer risk among all cases in this study. The multivariate relative risk of colon cancer was 0.76 (95% CI, 0.59 to 0.99) for total daily folate intake of ≥400 μg compared to <200 μg folate (table 2). The influence of total folate intake differed according to LINE-1 methylation; comparing extreme categories of folate intake (≥400 μg/day vs <200 μg/day), the RR was 0.57 (95% CI, 0.36 to 0.91) for <55% LINE-1 methylated tumours, 0.74 (95% CI, 0.51 to 1.06) for 55–64% LINE-1 methylated tumours, and 1.08 (95% CI, 0.66 to 1.75) for ≥65% LINE-1 methylated tumours (Pinteraction=0.01). In analyses restricted to folate from dietary sources, these RRs were generally similar albeit slightly weaker (data not shown). Similarly, using folate intake updated until up to 12 years before cancer diagnosis did not materially alter our results.

Table 2

Relative risk of baseline folate intake and colon cancer according to LINE-1 methylation level among 88 691 women from the Nurses' Health Study (NHS) and 47 363 men from the Health Professionals Follow-up Study (HPFS)

Next, we examined the influence of vitamin B6 intake according to LINE-1 methylation (table 3). The benefit of vitamin B6 intake also appeared confined to <65% LINE-1 methylated tumours, though none of these associations was statistically significant. Of note, for both folate and B6 intake, risk was principally elevated among participants in the lowest category, whereas the risk did not appear to decline substantially beyond the second category of exposure.

Table 3

Risk of colon cancer according to baseline quintiles of one-carbon nutrient intake by tumoral LINE-1 methylation level among 88 691 women and 47 363 men (NHS and HPFS combined)

The influence of intake of vitamin B12 and methionine (in quintiles) on colon cancer did not appear to differ by LINE-1 status (table 3).

We further evaluated the influence of alcohol consumption on colon cancer risk according to LINE-1 methylation level. For daily alcohol consumption (table 3), the overall increased risk of colon cancer with ≥15 g alcohol compared to no alcohol consumption (RR, 1.41; 95% CI, 1.11 to 1.79) appeared to be essentially restricted to <65% LINE-1 methylated tumours. Comparing extreme categories of alcohol consumption, the RR was 1.67 (95% CI, 1.04 to 2.67; Ptrend=0.02) for <55% LINE-1 methylated tumours, 1.55 (95% CI, 1.10 to 2.18; Ptrend=0.03) for 55–64%, and 1.06 (95% CI, 0.69 to 1.62; Ptrend=0.87) for ≥65% LINE-1 methylated tumours (Pinteraction=0.13).

Because previous analyses have suggested that the association between alcohol consumption and colon carcinogenesis is increased in individuals with inadequate folate intake,18 we examined dietary contrasts of total folate availability and daily alcohol consumption. The RR in participants with <299 μg folate intake/day and >5 g alcohol consumption/day (ie, dual depleted folate status) was 1.85 (95% 1.12 to 3.03) for <55% LINE-1 methylated tumours and 1.76 (95% CI, 1.17 to 2.64) for 55–64% LINE-1 methylated tumours, when compared to participants with ≥300 μg folate intake and <5 g alcohol per day, whereas this RR was 1.04 (95% CI, 0.64 to 1.69) for ≥65% LINE-1 methylated tumours.


In this large prospective cohort study, we found that low folate and, to a lesser degree, vitamin B6 intake and excess alcohol consumption were associated with increased risk of colon cancers with LINE-1 hypomethylation. The elevation in risk was principally limited to participants with the lowest levels of folate and vitamin B6 intake, and no additional risk reduction was observed for intake beyond the second lowest category of consumption. Specifically, higher doses of either vitamin did not appear to confer any additional benefit. By contrast, the increased risk with alcohol consumption appeared to follow a linear dose–response. Overall, our data support a possible aetiological link between deficiency in some one-carbon nutrients and genome-wide DNA hypomethylation during colorectal carcinogenesis.

To our knowledge, no prior study has assessed the influence of one-carbon nutrients on colon cancer risk according to LINE-1 methylation level, and only two previous studies have examined colon cancer survival according to LINE-1 methylation level. The larger study was based on data from our own cohorts, reporting LINE-1 hypomethylation to be an independent predictor of shorter survival in colon cancer patients.14 Another much smaller study (with only 93 tumours) also identified a trend (albeit non-significant) towards poor survival in DNA-hypomethylated tumours.32

We have previously shown evidence supporting that folate prevents p53 mutational events in colorectal carcinogenesis, but we did not observe any influence of folate on p53 wild-type tumours.33 The processes underlying aggressive tumour behaviour in LINE-1 hypomethylated tumours are currently unknown. Possible mechanisms include activation of retroviruses at transposons, which may cause genomic instability, in particular chromosomal instability.34 35

It is plausible that chronic folate deficiency may be associated with genome-wide DNA hypomethylation, given the importance of folate in DNA methylation and synthesis. Recent experimental data show a significant reduction in global DNA methylation level in colonic epithelial cells of mice with folate deficient diet.36 A prior study explored the association between folate and other methyl donors and colon cancer subtypes.37 While the overall inverse association between folate and colon cancer did not differ significantly according to MSI status, there was the suggestion of a slightly stronger association between folate and MSI-high colon tumours (RR, 0.79, 95% CI, 0.60 to 1.03 for microsatellite stable/MSI-low colon cancers and RR, 0.61; 95% CI, 0.37 to 1.02 for MSI-high colon cancers). Our current findings of a stronger association between low folate and LINE-1 hypomethylation are in line with a prior report that LINE-1 hypomethylation is inversely associated with MSI in these cohorts.13 Further, as previously described, survival was poorer among colon cancer patients with deplete prediagnostic plasma folate in our cohorts.38 If, as suggested by our data, LINE-1 hypomethylated colon tumours occur more frequently in folate deplete individuals, this provides compelling mechanistic support for the association between folate depletion and poor colon cancer survival.

In a recent report, Figueiredo et al39 showed that folate supplementation did not alter LINE-1 methylation levels in normal colorectal mucosa. Together with our data, this could suggest that folate levels may not be relevant in terms of LINE-1 methylation in normal mucosa, with relatively normal cellular kinetics, but once neoplasia develops some factor, possibly the increase in cellular proliferation, may reveal the relationship between folate and LINE-1 methylation.

Our study has several important strengths. First, because we collected detailed, updated information on a number of dietary and lifestyle covariates relevant to colon carcinogenesis over up to 22 years of follow-up and with high follow-up rates, we were able to examine long-term exposures to one-carbon nutrients and to take into consideration important confounding factors. Second, our study is prospective, eliminating concerns about differential recall bias, particularly with regard to our dietary assessments. Any remaining bias from exposure misclassification would thus be nondifferential by nature, biasing our results towards the null. Third, our molecular characterisation of colon cancer has proven very reliable, resulting in a number of interesting epidemiologic observations relating to colon cancer and tissue biomarkers.13 14 22 37

Limitations of note relate to folate fortification, which became mandatory in the USA in 1998.40 We did obtain multiple assessments of one-carbon nutrient intakes prior to fortification. In addition, since the development of colon cancer likely requires some induction period before the onset of a clinically apparent tumour, it is unlikely that the post-fortification folate exposure would substantially influence colon cancer risk through 2002. Another potential limitation is that we were unable to obtain tumour tissue from all cases of confirmed colon cancer in the two cohorts. However, risk factors in cases unavailable for tissue analysis did not appreciably differ from those in cases with tumour tissue available.

In conclusion, we show that the reduced risk of colon cancer associated with replete folate status varies by LINE-1 methylation level, an indicator of global DNA methylation status. Thus, genome-wide DNA hypomethylation may be one mechanism by which folate affects colon cancer risk and survival, but additional studies are needed to further elucidate these preventive effects.


We thank the participants of the Nurses' Health Study and Health Professionals Follow-up Study for their cooperation and participation, and hospitals and pathology departments throughout the US for generously providing us with tumour tissue materials. The authors are grateful to Gregory Kirkner for technical and administrative assistance.



  • The content of this paper is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.

  • Funding This work is supported by National Institutes of Health research grants CA70817, CA87969, CA55075, CA42812, CA58684, CA90598, CA122826, the Bennett Family Fund and Entertainment Industry Foundation, and the Entertainment Industry Foundation National Colorectal Cancer Research Alliance (NCCRA).

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the BWH.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Digest
    Emad El-Omar Severine Vermeire Alexander Gerbes
  • Corrections
    BMJ Publishing Group Ltd and British Society of Gastroenterology