The association of cholesterol levels with memory and memory change over a 14-year period in a US national cohort.

INTRODUCTION: The impact of cholesterol on late-life cognition remains controversial. We investigated the association of high-density lipoprotein cholesterol (HDL-C) and non-HDL-C with memory in a nationally representative cohort.

METHODS: Health and Retirement Study (HRS) participants ( = 13,258) aged 50+ (mean age: 67.2 years) followed from 2006 to 2020 provided cholesterol measures every 4 years and cognitive assessments biennially. Linear mixed models predicted memory scores using both baseline and time-updated cholesterol values.

RESULTS: Higher baseline HDL-C (mean: 53.9 mg/dL) predicted better memory scores (: 0.05, 95% confidence interval [CI]: 0.03 to 0.08), but not memory change. Baseline non-HDL-C (mean: 143 mg/dL) predicted poorer memory scores (: -0.01, 95% CI: -0.02 to 0.00), but not memory change. Time-updated HDL-C predicted better memory (: 0.02, 95% CI: 0.00 to 0.04), but non-HDL-C showed no such associations.

DISCUSSION: While higher peripheral HDL-C is linked to better memory, the small effect sizes and absence of associations of HDL-C and non-HD-CL with memory change suggests that peripheral cholesterol has a small effect on the variation of memory scores.

HIGHLIGHTS: Higher HDL-C levels predict better memory scores but not memory change across 14 years of follow-up.Baseline higher LDL-C levels predict poorer memory scores across time, but not memory change.The small effects and absence of consistent association between cholesterol levels and memory change suggest that cholesterol plays a minor role in cognitive decline.