Joint association of dietary fatty acid patterns identified by Gaussian graphical models and cognitive impairment status with all-cause mortality in U.S. adults: A prospective cohort study

BACKGROUND: The association between dietary fatty acid patterns (DFAPs) and cognitive function jointly affecting long-term survival remains poorly understood. Network-based methods addressing conditional dependencies among fatty acids may reveal novel patterns, yet the joint association of DFAPs derived from Gaussian Graphical Models (GGMs) and cognitive status with all-cause mortality (ACM) risk has not been examined in nationally representative U.S. adults. METHODS: GGMs were applied to 24-hour recall data for 19 fatty acids from 2446 adults aged ≥18 years in the NHANES 2011-2014 cycles. Psychometric mild cognitive impairment (pMCI) was defined as a composite cognitive score >1 standard deviation below the mean. Survey-weighted Cox proportional hazards models estimated joint associations of DFAP quartiles and pMCI with ACM. RESULTS: : 0.67, 95% CI: 0.45-0.98). CONCLUSION: Greater adherence to the MLCFA pattern combined with pMCI may be associated with ACM risk reduction among U.S. females and adults without diabetes.