Abstract LB-395: Alcohol intake and risk of non-Hodgkin's lymphoma: The Multiethnic Cohort

Abstract

Abstract Evidence from case-control studies suggests an inverse association between alcohol intake and risk of non-Hodgkin's lymphoma (NHL), which might be due to immunemodulatory effects of alcohol. Results from prospective studies are sparse and less consistent. Therefore, we examined the relationship between intake of alcohol and NHL risk within the prospective Multiethnic Cohort (MEC). Between 1993 and 1996, residents of Hawaii and the county of Los Angeles, aged 45-75 years, entered the cohort by completing a questionnaire on demographics, lifestyle and diet. Portion size and consumption frequency of beer, wine, and liquor was assessed. The study population comprised 193,041 participants of Caucasian, African American, Japanese American, Latino and Native Hawaiian ancestry. NHL cases were identified via record linkage with cancer registries. Hazard ratios (HR) and 95% confidence intervals (CI) of NHL or subtypes associated with alcohol intake were calculated using Cox regression with age as underlying time metric. Models were adjusted for age at baseline, sex, ethnicity, body mass index and education. Furthermore, we tested for effect modification by sex, ethnicity, and smoking status. During a mean follow-up time of 9.1 years, 939 incident cases of NHL occurred, including 311 cases of diffuse large B-cell lymphoma (DLBCL), 152 of follicular lymphoma (FL), and 198 of small lymphocytic lymphoma/chronic lymphocytic leukemia (SLL/CLL). Overall, 39% of men and 62% of women were classified as non-drinkers, while 30% of men and 10% of women reported > 1 drink/day. We found no association between alcohol intake and overall risk of NHL. As compared to non-drinkers, the respective HRs for participants with ≤ 1 and > 1 drink/day were 1.02 (95% CI 0.88-1.19) and 1.03 (95% CI 0.86-1.24) in multivariate adjusted models. We found no interaction between alcohol intake and sex (pinteraction = 0.63) or smoking (pinteraction = 0.40). Stratified by ethnicity, HRs tend to increase with alcohol intake in Caucasians, while in Latinos those with > 1 drink/day had a lower risk of NHL than non-drinkers, but the test for interaction of alcohol intake and ethnicity was not significant (pinteraction = 0.14). However, among NHL subtypes, participants with > 1 drink/day had a HR of 0.69 (95% CI 0.49-0.98) for DLBCL, of 1.93 (95% CI 1.24-3.00) for FL, and of 0.86 (95% CI 0.59-1.26) for SLL/CLL as compared to non-drinkers. Our findings regarding DLBCL support previous studies reporting an inverse association between alcohol intake and NHL risk, but the absence of an overall effect and the positive association with FL disagree with published evidence. The small number of NHL cases per subtype, the ethnic heterogeneity of our study population, and the large proportion of non-drinkers should be kept in mind when interpreting these findings. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr LB-395.