516Association of dietary diversity with all-cause mortality by body mass index in Japanese older adults

Background Dietary diversity is thought to reduce risk of malnutrition although it is also linked to obesity. We examined whether dietary diversity is associated with all-cause mortality in Japanese elderly by body mass index (BMI). Methods The current study included 2,944 people aged 64-65 years who participated in the NISSIN project in 1996-2005. Dietary diversity was measured by the Food Variety Score (FVS) which calculates frequency of all food items consumed daily using a self-administered food frequency questionnaire. Participants were divided into tertiles according to their FVS (1st: low, 2nd: middle, 3rd: high). Multivariate adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated using the Cox proportional hazard regression model. In stratified analysis, BMI was used to divide the participants into 3 groups: lean (BMI <20), normal (BMI 20–24.9), and overweight/obese (BMI ≥ 25). Results Overall, 454 (30.7%) men and 222 (15.2%) women died over the median follow-up period of 16.6 years. No significant association between FVS and all-cause mortality was observed overall. However, when grouped by BMI, in the lean, multivariate adjusted HRs were 0.32 (CI: 0.17-0.59) in the middle FVS and 0.41 (CI: 0.19 -0.91) in the high FVS, compared with the low FVS. No significant association was found for normal and overweight/obese. Conclusions These findings indicate dietary diversity should be promoted for lean older adults. Key messages BMI should be considered when promoting dietary diversity.