BACKGROUND
Frailty occurs in 10-15% of community-living older adults and inflammation is a key determinant of frailty. Though diet is a modulator of inflammation, there are few prospective studies elucidating the role of diet-associated inflammation on frailty.
OBJECTIVE
To determine whether a pro-inflammatory diet was associated with increased odds of frailty in adults from the Framingham Heart Study (FHS).
DESIGN AND METHODS
This study is nested in a prospective cohort that included individuals without frailty. Diet was assessed in 1998-2001 using a valid food frequency questionnaire (FFQ) and frailty was measured in 2011-2014. FFQ-derived energy-adjusted dietary inflammatory index (E-DII®) scores were computed, with higher E-DII scores indicating a more pro-inflammatory diet. Frailty was defined as fulfilling ≥3 of 5 Fried Phenotype criteria. Information on potential mediators, serum interleukin-6 and C-reactive protein was obtained in 1998-2001. Logistic regression estimated odds ratios (OR) and 95% confidence intervals (95% CI) for E-DII (as continuous and in quartiles) and frailty onset adjusting for relevant confounders.
RESULTS
Of 1,701 individuals without frailty at baseline (mean age = 58 years, SD = 8, range: 33-81; 55% female), 224 developed frailty (13% incidence) over ∼12 years. Mean E-DII score was -1.95 (SD = 2.20; range: -6.71 to +5.40). After adjusting for relevant confounders, a one-unit higher E-DII score was associated with 16% increased odds of developing frailty (95% CI = 1.07, 1.25). In categorical analyses, participants in the highest (pro-inflammatory) vs. lowest quartile of E-DII had >2-fold increased odds of frailty (ORquartile4vs.1 = 2.22, 95% CI = 1.37, 3.60, Ptrend<0.01). Interleukin-6 and C-reactive protein were not major contributors in the pathway.
CONCLUSIONS
In this cohort of middle-aged and older adults, a pro-inflammatory diet was associated with increased odds of frailty over ∼12 years of follow-up. Trials designed to increase consumption of anti-inflammatory foods for frailty prevention are warranted.Clinical registry number and website: Not applicable.