Growth Differentiation Factor 15 and Diet Quality Trajectory Interact to Determine Frailty Incidence among Middle-Aged Urban Adults.

ABSTRACT

BACKGROUND: Elevated plasma growth differentiation factor 15 (GDF15) and poor diet quality may be associated with increased frailty incidence, although their interactive associations have not been assessed in urban middle-aged adults. OBJECTIVES: We aimed to examine GDF15 and its interactive association with diet quality in relation to frailty incidence among a sample of middle-aged urban adults. METHODS: The relationship between GDF15 and diet quality trajectories in relation to incident frailty was examined in a longitudinal study of participants in the Healthy Aging in Neighborhoods of Diversity across the Life Span (2004-2017). Serum GDF15 concentration and frailty incidence were primary exposure and outcome, respectively. Group-based trajectory models were used to assess diet quality trajectories (≤3 visits/participant, N = 945, N' = 2247 observations) using the Healthy Eating Index 2010 version (HEI-2010), Dietary Inflammatory Index, and mean adequacy ratio (MAR). Cox proportional hazards models were used, testing interactive associations of GDF15 and diet quality trajectories with frail/prefrail incidence (N = 400 frailty-free at first visit, N' = 604 observations, n = 168 incident frail/prefrail). RESULTS: Both elevated GDF15 and lower diet quality trajectories were associated with a lower probability of remaining nonfrail (≤13 y follow-up). Among females, the "high diet quality" HEI-2010 trajectory had a hazard ratio (HR) of 0.15 [95% confidence interval (CI) 0.04, 0.54; P = 0.004; fully adjusted model] when compared with the "low diet quality" trajectory group. Among males only, there was an antagonistic interaction between lower HEI-2010 trajectory and elevated GDF15. Specifically, the HR for GDF15-frailty in the higher diet quality trajectory group (high/medium combined), and among males, was 2.69 (95% CI 1.06, 6.62; P = 0.032), whereas among the lower diet quality trajectory group, the HR was 0.94 (95% CI 0.49, 1.80; P = 0.86). Elevated GDF15 was independently associated with frailty among African American adults. CONCLUSIONS: Pending replication, we found an antagonistic interaction between GDF15 and HEI-2010 trajectory in relation to frailty incidence among males.