(P12-45-26) Ultra-Processed Food Consumption (UPF) and Metabolomic Biomarker Score of UPF Consumption With Frailty Trajectory and Mortality in Older Adults

Objectives: The aim of this study is to examine the association of UPF consumption, assessed using nova classification on data from food frequency questionnaire (FFQ) or metabolomic biomarkers, and the trajectory of frailty and risk of all-cause mortality in participants of the Baltimore Longitudinal Study of Aging (BLSA).

Methods: Using data from a prospective cohort study, dietary data from FFQ were used to assess UPF consumption based on the nova classification. These were expressed as the percentage of energy from UPF and categorized into quartiles (dUPF). Plasma metabolites were assessed with the Global Metabolon Assay, and a metabolite UPF score (mUPF) was constructed. Frailty was defined using a 44-item frailty index, with follow-up over 10 years. All-cause mortality was assessed through self-report and confirmed through death registries. The association between baseline UPF and frailty trajectory was evaluated using linear mixed-effects models, and its association with mortality was assessed using Cox proportional hazards regression.

Results: In 743 BLSA participants, both dUPF and mUPF were associated with frailty trajectory. For dUPF, compared to the lowest quartile (Q1), those in the third quartile had greater frailty throughout the follow-up period (b=0.02±0.001, p=0.009). For mUPF, compared to the lowest quartile (Q1), the three higher quartiles had higher FI throughout the follow-up period (Q2: b=0.02±0.01, p=0.004; Q3: b=0.01±0.01, p=0.033; Q4: b=0.03±0.01, p< 0.001). Similarly, both dUPF and mUPF were associated with all-cause mortality. For dUPF, significant differences were observed for Q3 (HR=1.02 [1.01-1.04]), while mUPF, significant differences were observed for Q2 (HR=1.02 [1.01-1.04]), Q3 (1.01 [1.01-1.03]), and Q4 (1.03 [1.02-1.05]).

Conclusions: Higher consumption of UPF evaluated using FFQ and metabolomics was significantly associated with a greater burden of frailty and increased risk of all-cause mortality. The mUPF was more robustly associated with aging outcomes, suggesting that an objectively measured biomarker may be useful for monitoring the risk of age-related decline and UPF consumption.

Funding Sources: This research was supported in part by the Intramural Research Program of the National Institutes of Health (NIH) and by NIA grants (R21AG085005, R01AG051752).