Cross-sectional and longitudinal associations between empirically derived dietary patterns and frailty among older men: The Concord Health and Ageing in Men Project.

Abstract

Diet may be associated with frailty. We aimed to evaluate the associations between empirically derived dietary patterns and frailty in older men. Prospective cohort study. The Concord Health and Ageing in Men Project, Sydney, Australia. 785 community-dwelling older Australian men aged 75 years and older. Men underwent dietary assessment using a validated dietitian-administered diet history questionnaire. Factor analysis identified three dietary patterns. Multinomial logistic regression was conducted between frailty and dietary patterns for cross-sectional analyses and longitudinal analyses over a 3-year follow-up. Frailty was defined by the Fried frailty phenotype. Of the 785 men, pre-frailty was prevalent in 47.1% (n = 370), and frailty in 8.3% (n = 65). In fully adjusted cross-sectional analyses, the top tertile and a higher 'vegetables-legumes-seafood' dietary pattern score were associated with reduced prevalence of frailty (OR 0.34 [95% CI: 0.12, 0.93, P = .036]) and OR 0.50 [95% CI: 0.30, 0.83, P = .007] respectively). The top tertile of the 'discretionary-starchy vegetables-processed meats' dietary pattern was also associated cross-sectionally with increased prevalence of pre-frailty (OR 1.75 [95% CI: 1.08, 2.83, P = .022]). Of the 296 robust men in fully adjusted longitudinal analyses, the incidence of pre-frailty was 52.4% (n = 155), and frailty was 5.4% (n = 16) over a 3-year follow-up. The middle tertile of the 'vegetables-legumes-seafood' dietary pattern had a non-significant trend towards reduced incident pre-frailty (OR 0.52 [95% CI: 0.27, 1.00, P = .050]). Consumption of a 'vegetables-legumes-seafood' dietary pattern appears to be less favoured by frail older men.