Serum concentrations of fatty acids and colorectal adenoma risk: a case-control study in Japan.

BACKGROUND: Epidemiologic studies of n-3 fatty acids (FAs) and risk of colorectal cancer have generated inconsistent results, and relations with precursor colorectal adenomas (CRA) have not been evaluated in detail. We here focused on possible associations of serum FAs with CRA in the Japanese population. METHODS: We conducted a case-control study of 203 asymptomatic CRA cases (148 men, 55 women) and 179 healthy controls (67 men, 112 women) during 1997-2003 in Nagoya, Japan. Baseline information was obtained using a lifestyle questionnaire and serum FA levels were measured by gas chromatography. RESULTS: A non-significant inverse association with CRA was observed for eicosapentaenoic acid (EPA) among women. Moreover, the concentrations of docosahexaenoeic acid (DHA), a major component of n-3 highly-unsaturated FAs (HUFAs), were significantly lower in cases in both sexes. In addition, serum concentrations of total FAs, saturated FAs (SFAs) and mono-unsaturated FAs (MUFAs) had strong positive links with CRA risk. In contrast, arachidonic acid (AA) and DHA were inversely related, with 66% and 59% risk reduction, respectively. Ratios of SFAs/n-3 PUFAs and SFAs/n-3 HUFAs exhibited significant positive relations with CRA risk but there was no clear link with n-6 PUFAs/n-3 PUFAs. CONCLUSIONS: Our findings suggest a promoting influence of SFAs and MUFAs along with a protective effect of DHA on CRA risk. However, further research is needed to investigate the observed discrepancy with the generally accepted roles of the AA cascade in carcinogenesis.

Risk of endometrial cancer mortality by ever-use of sex hormones and other factors in Japan.

OBJECTIVES: To examine associations of ever-use of sex hormones (EUSH) and other factors with endometrial cancer (EC) mortality through a nation-wide Japan Collaborative Cohort Study. METHODS: A total of 63,541 women aged 40-79 years, enrolled in 1988-90 from 45 municipalities of Japan, were followed until 2003 to record their vital status. Using baseline data, the Cox proportional hazard model (age adjusted and multivariate) was used to estimate the hazard ratio (HR) and 95% confidence interval (CI) for EC mortality by selected factors, including EUSH. Bivariate analysis was also conducted to establish associations between EUSH and other factors. RESULTS: The mortality rate from EC was 2.6 per 100,000 person-years during the mean follow-up period of 13.3 years. Prevalence rate of EUSH was 5.2%. Significantly increased risk of EC mortality was found for EUSH with both age adjusted (HR=6.43, 95%CI=2.10-19.67) and multivariate (HR=5.33; 95%CI=1.51-18.82) analyses. Bivariate analysis indicated that history of diabetes mellitus, smoking, drinking, and age at first delivery were positively associated with EUSH, whereas age, number of delivery, number of pregnancy, and age at menarche demonstrated inverse links. CONCLUSIONS: Our results imply that EUSH may increase the risk of EC mortality among Japanese women. However, further studies with more deaths are needed to validate the results.