Abstract
BACKGROUND: Although several studies have investigated the association between cholesterol and dementia, few have examined cholesterol and decline across cognitive domains. We examined serum total and high-density lipoprotein (HDL) cholesterol, total-to-HDL ratio, and trajectories across cognitive domains. METHODS: Participants were 436 community-residing women (70-79 years old) in the Women's Health and Aging Study II; they were screened to be physically high-functioning and cognitively intact at baseline. Cognition and other health-related variables were assessed at five intervals spanning 9 years. Cognitive assessments included Trail Making Test Parts A (TMT-A) and B (TMT-B), Hopkins Verbal Learning Test-Revised, Purdue Pegboard, and Mini-Mental State Examination (MMSE). The association between baseline levels of serum lipids and cognitive trajectories were evaluated using Generalized Estimating Equations (GEE). Covariates included age, education, race, vascular disease, serum creatinine, depression, and lipid-lowering medications. RESULTS: In multivariate analyses, baseline higher total (p =.02) and HDL (p =.03) cholesterol were associated with better performance on the Purdue Pegboard. Using clinical cholesterol cutoffs, baseline serum total cholesterol levels >240 mg/dL were associated with the best performance (p =.02). Baseline lipids were not associated with any other cognitive tests; there were no Lipid x Time interactions. CONCLUSION: Higher baseline serum lipid levels predicted better performance over time on a measure of motor speed, but not memory or psychomotor or executive functioning in this population of elderly women. This association suggests that peripheral cholesterol levels, measured in late-life, may not be a good predictor of subsequent cognitive decline. Future research examining peripheral cholesterol over the life span and its relationship with cognition is needed.