(P12-32-26) Prospective Associations of Almond Intake Metabolic Profiles With Depressive Symptoms Vary by the Presence of Almond-Responsive Gut Bacteria in Hispanics/Latinos

Objectives: To assess almond intake metabolic responses and their prospective association with depressive symptoms in the Hispanic Community Health Study/Study of Latinos, stratifying by almond-responsive gut bacteria abundance.

Methods: Including 3,771 adults without cardiovascular disease/cancer, we applied elastic net with leave-one-out cross-validation to baseline (2008-11) 24-hour dietary recall and metabolomics data to regress almond intake on 650 normalized serum metabolites. A signature score was computed: sum of selected metabolites across all validation iterations using elastic net coefficients as weights. Multivariable Poisson models estimated rate ratios (RRs; 95% CI) for depressive symptoms at follow-up (2017), defined as a score ≥10 on the Center for Epidemiologic Studies Depression Scale, per 1-SD increase in the signature. In a random subsample with shotgun metagenomics (n=1,220), models were stratified by low/high abundance of literature-based almond-responsive genera (Bacteroides, Parabacteroides, Alistipes, Akkermansia, Bifidobacterium, Lactobacillus, Blautia, Dorea, Ruminococcus, Faecalibacterium), defined by a score (sum of normalized abundances) dichotomized at the median.

Results: Almond intake (mean: 0.6g/day) explained 9% of variance in almond-responsive genera composition (p=0.001) and was associated with a 57-metabolite signature, whose most prevalent sub-pathways were food component/plant, tocopherol, fatty acid dicarboxylate (7% each), secondary bile acid and sulfur amino acid metabolism (5% each). Overall associations between the signature and depressive symptoms (938 cases) were not significant. In persons with high almond-responsive genera abundance, each SD increase in the signature was associated with a 26% lower rate of depressive symptoms (RR: 0.74; 95% CI: 0.59, 0.94), but with a 26% higher rate in those with low abundance (1.26; 1.02, 1.55).

Conclusions: Almond intake was associated with a metabolic profile enriched in plant, vitamin E, lipid oxidation, bile acid, and sulfur amino acid pathways. In Hispanics/Latinos with high abundance of almond nutrient-metabolizing genera, this profile was associated with lower depressive symptom rates, and conversely for the low abundance group. Replication in people with higher almond intake is warranted.

Funding Sources: NIH/Almond Board of California