The Journal of Nutrition · Oct 26, 2026

Associations Between Dietary Carotenoids and Cognitive Performance in Older Hispanic/Latino Adults are Stronger in Men and Smokers

By Krause CA, Woods SP, Litson K et al.

BACKGROUND: Dietary intake is associated with age-related cognitive decline and neurodegenerative disease. Carotenoids, fruit-and-vegetable-derived nutrients, support brain health through anti-oxidative and anti-inflammatory properties in non-Hispanic White samples. However, their relevance to ethnoracially diverse populations, including the rapidly growing and aging Hispanic/Latin American (H/L) population, remains unclear. We sought to elucidate the relationship between carotenoids, cognition, and inflammation in adult H/Ls. METHODS: Cross-sectional data from the National Health and Nutrition Examination Survey represented participants (Mage=67.5±5.9y) of Mexican American (n=2,078,580; 42% Spanish-preferring) and Other H/L heritage groups (n=2,244,252; 51% Spanish-preferring). Regression models adjusted for age, education, smoking status, BMI, acute inflammation, and total energy intake examined the relationship between carotenoid intakes (MTotal Carotenoid=9.47±9.05mg), and a cognitive composite score. Inverted mediation tested whether three systemic inflammation indices (i.e., systemic inflammation response index, SIRI; systemic immune-inflammation index, SII; and pan-immune-inflammation value, PIV) mediated these relationships. RESULTS: Regression results reflect sensitivity analyses that removed (n=61) extreme carotenoid intake values identified by Mahalanobis and Cook's distances. Regression models revealed that total carotenoids (β=0.26), α-carotene (β=0.22), β-cryptoxanthin (β=1.0), lutein+zeaxanthin (β=0.21), and lycopene (β=0.19) were significantly (all ps ≤ 0.04) associated with cognitive performance. Ethnicity/Language-stratified analyses demonstrated significant associations between total and individual carotenoids exclusively among English-preferring H/Ls. However, these associations did not remain significant after false discovery rate (FDR) correction (all FDR-adjusted ps ≥ 0.09). Significant interactions suggested associations between a-carotene and cognitive performance differed by sex (χ²(1)=7.02; p=0.029) and smoking (χ²(1)=8.33; p =0.018). Post-hoc analyses revealed stronger effects for men (β=0.34) and smokers (β=1.64), respectively (all ps ≤ 0.001). Results did not support mediation effects of the systemic inflammation indices. CONCLUSION: Findings revealed nuanced associations between carotenoids and cognition among H/Ls, varying by sex and smoking. Future diet-brain health work should prioritize culturally-informed, community-based approaches to address disparities at the intersection of culture, cognition, aging, and diet.

doi.org/10.1016/j.tjnut.2026.101904PubMed 42933645Populations