Biomarkers of B vitamins, alcohol intake, methylenetetrahydrofolate reductase (MTHFR) polymorphisms and breast cancer risk in the European Prospective Investigation into Cancer and Nutrition (EPIC) study

  • Topic: Breast Cancer
  • Institution: International Agency for Research on Cancer (IARC)
  • Country: France
  • Status: Completed

Scientific abstract

(View plain language abstract)

Background

Breast cancer (BC) is the most common cancer in women worldwide. BC etiology is complex and several factors appear to play a role, including dietary factors. Among those, there is convincing evidence for a relationship between alcohol intake and BC risk; however, nutrients such as folate, a water-soluble B-vitamin with an important role in the synthesis, repair and methylation of DNA, could also play a role.

Hypothesis & Objectives

The overall aim of the study was to investigate the relationship between dietary folate, plasma levels of B vitamins, and BC risk using data from the European Prospective Investigation into Cancer and Nutrition (EPIC). More specifically, major aims were as follows:

  1. To analyse plasma levels of folate and B-vitamins in pre-diagnostic plasma samples from 3,858 breast cancer cases, and 3,858 matched controls who are part of the large European cohort EPIC.
  2. To conduct multivariable analysis of the associations between dietary folate, plasma B vitamin levels and the risk of BC, with stratification on potential modifying effect factors (estrogen receptor status, alcohol intake).
  3. To analyse MTHFR polymorphisms (Methylenetetrahydrofolate reductase, a key enzyme residing at a critical metabolic branch point in folate metabolism polymorphisms) in the population study, to study the modifying effects of MTHFR polymorphisms on the associations B vitamins-BC risk.

Costs for sample extraction from biobanks and for the majority of biomarker analyses have been previously obtained, the funds received from WCRF allowed the completion of biomarker analyses, and supported all statistical analyses on associations.

Settings & Methods

The EPIC project is a multi-centre prospective study that was set up to investigate relationships of cancer risk with nutrition and metabolic risk factors. So far, 3,858 cases with known oestrogen and progesterone receptor status (998 premenopausal, 2,860 postmenopausal) and biological samples available have been identified. For each case subject, one matched control was randomly chosen among cohort women without BC. Plasma B vitamins have been measured through validated methods, and genotyping of MTHFR polymorphisms was performed by Taqman. Dietary folate cohort-wide was estimated from food frequency questionnaires.

Cox proportional hazards regression models were used to quantify the association between dietary intake of folate and BC risk in the overall cohort. BC tumours were classified by receptor status. Sub-group analyses were performed by menopausal status and alcohol intake. Conditional logistic regression models were used to assess the associations between plasma levels of vitamins B and BC risk. Relative risk (RR) estimates and their 95% confidence intervals (CI) were estimated, adjusting for relevant confounders where necessary. Multivariable linear regression models were used to assess the main determinants of log natural transformed folate and vitamin B12 plasma concentrations.

Results

An inverse association was observed between dietary folate and BC risk overall, and especially in premenopausal women. A 15% reduction in BC risk was observed comparing the highest to the lowest dietary folate tertiles in women having a high (>20g/day) alcohol intake (HRT3-T1=0.85, 95%CI: 0.75-0.97), although the formal interaction test was not statistically significant Determinants of plasma levels of B vitamins were identified. Analyses on risk of BC with plasma levels of B vitamins are on-going.

Impact

We provided evidence that higher dietary folate intake estimated through dietary questionnaires is associated with a lower risk of hormonal receptor negative breast cancer in premenopausal women.

Plain language abstract

Background

There is limited evidence on the association between folate and the risk of breast cancer (BC).

Aims & Objectives

We aimed to investigate the relationship between dietary folate, plasma levels of B vitamins, and BC risk according to some genetic polymorphisms, hormonal receptor status, and alcohol intake, using data from the European Prospective Investigation into Cancer and Nutrition (EPIC).

How the study was carried out

A total of 367,993 women aged 35-70 years were recruited in 10 European countries. During a median follow-up of 11.5 years, 11,575 women with BC were identified. Dietary folate intake was estimated using the updated EPIC Nutrient Data Base (ENDB). A sub-analysis was conducted among 3,858 cases with known oestrogen and progesterone receptor status (998 premenopausal, 2,860 postmenopausal). For each case subject, one matched control was randomly chosen among cohort women without BC. Control subjects were matched to cases for center, age, menopausal status, date of recruitment and fasting status. Within this group, plasma levels of folate and vitamin B12 were estimated through validated microbiological methods, and genotyping of MTHFR polymorphisms was performed by Taqman.

Cox proportional hazards regression models were used to quantify the association between dietary intake and BC risk on the overall cohort. Sub-group analyses were performed by receptor status, menopausal status and alcohol intake.

Statistical analyses of the association between plasma levels of B vitamins and BC risk according to MTHFR polymorphisms, alcohol intake and hormonal receptor status are on-going at IARC. Relative risk (RR) estimates and their 95% confidence intervals (CI) will be evaluated using a conditional logistic regression model, adjusting for relevant confounders where necessary.

Key findings & conclusions

Median (p25-p75) folate intake across countries was 286 (232 – 352) µg/day, with the highest intakes observed in the UK and France. Median (p25-p75) intakes of other B vitamins were: B2 1.68 (1.36-2.15) mg/day, B6 1.70 (1.45-1.99) mg/day, B12 5.47 (4.05-7.30) µg/day.

An inverse association was observed between dietary folate and BC risk (Hazard ratio comparing top vs. bottom quintile, HRQ5-Q1=0.92, 95%CI: 0.83-1.01, ptrend=0.037). A borderline significant decreasing trend in risk for folate intakes higher than 300 µg/day, with a levelling-off at 400 µg/day, was observed.

In premenopausal, women, we observed a significant decreasing trend in the risk of oestrogen receptor-negative BC (HRQ5-Q1=0.66, 95%CI: 0.45-0.96, ptrend=0.042) and progesterone receptor-negative BC (HRQ5-Q1=0.70, 95%CI: 0.51-0.97, ptrend=0.021). No associations were found in postmenopausal women.

A 15% reduction in BC risk was observed comparing the highest to the lowest dietary folate tertiles in women having a high (>20g/day) alcohol intake (HRT3-T1=0.85, 95%CI: 0.75-0.97), although the formal interaction test was not statistically significant (pinteraction=0.086).

These data indicated that higher dietary folate intake is associated with a lower risk of hormonal receptor negative BC in premenopausal women. These data will be further confirmed by exploring the association between biomarkers of B vitamins and BC risk, and refined by investigating the potential modifying effect of MTHFR polymorphisms.

Grant publications