Associations of energy-adjusted dietary niacin equivalent density with cognitive function and incident dementia: evidence from a large population-based cohort

Niacin is an essential micronutrient involved in energy metabolism and neurological function; however, evidence linking energy-adjusted dietary niacin equivalent density (E-NE density) with cognitive performance and dementia risk remains limited and inconclusive. Our objective was to investigate the association between E-NE density, cognitive performance, and incident dementia, considering potential moderation by genetic susceptibility to dementia. Participants from the UK Biobank who completed at least one 24-hour dietary questionnaire and were free of dementia before the end date of the final dietary assessment cycle were included ( n = 207,043) and followed for a median of 10.49 years. During this period, 2,403 incident dementia cases were identified, including 1,046 cases of Alzheimer’s dementia and 439 cases of vascular dementia. Linear regression was employed to examine the association between E-NE density and cognitive performance. Cox proportional hazards models were used to estimate the association between E-NE density and dementia risk. Potential non-linear associations were further assessed, and interactions between E-NE density and polygenic risk score were evaluated. Higher E-NE density was associated with better cognitive performance across multiple domains in a non-linear manner, with inflection points approximately corresponding to quintiles 2–4 of E-NE density (P for non-linear < 0.001). Compared with moderate E-NE density (quintiles2-4), low E-NE density was associated with higher risks of all-cause dementia (HR,1.26; 95% CI, 1.13–1.41), Alzheimer’s dementia (HR, 1.22; 95% CI, 1.03–1.44), and vascular dementia (HR,1.32; 95% CI,1.30–1.69) in the fully adjusted model. Dose-response analyses suggested L-shaped non-linear associations of E-NE density with risks of all-cause dementia and Alzheimer’s dementia, with inflection points of 18.088 and 18.107 mg/1000 kcal, respectively, while vascular dementia showed an approximately linear inverse association. The association with Alzheimer’s dementia appeared to be modified by genetic risk, with moderate and high E-NE density being associated with lower Alzheimer’s dementia risk in participants with high genetic risk, but not among those with low genetic risk (p for interaction = 0.019). E-NE density was associated with cognitive performance and dementia risk, particularly among individuals with high genetic susceptibility. Further studies are needed to confirm these observational findings.