Association of Long-Term Vascular Risk Factor Variability from Childhood and Midlife Cognitive Function: The Bogalusa Heart Study

Background Variability in vascular risk factor (VRF) levels in adulthood has been associated with higher dementia risk. Whether VRF variability from childhood influences midlife cognitive function (CF) is unclear. We aimed to evaluate the association of long-term VRF variability from childhood with midlife CF, independent of mean levels. Methods In this cohort study, participants in the Bogalusa Heart Study with ≥ 3 VRF (systolic blood pressure [SBP], fasting glucose, non-high-density lipoprotein cholesterol [non-HDL], and body mass index [BMI]) measurements from childhood and midlife CF assessment were included. VRF variability was measured using indices reflecting variability around the mean (SD, coefficient of variation [CV]) and around the age-adjusted slope (deviation from age-predicted values [DEV], residual SD [RSD]). CF was assessed using a global cognitive score (GCS) averaging eight standardized neuropsychological tests, and three cognitive domain performances. Associations between tertiles of VRF variability and CF were evaluated using multivariable linear regression with the lowest tertile as reference. Results Among 1,010 participants (mean age 48 years, 61% women, 34% Black, mean follow-up 39 years), being in highest tertiles of SBP and glucose variability were associated with poorer GCS (SBP by DEV: β = -0.10, 95% CI = -0.17, -0.02; glucose by CV: β = -0.12, 95% CI = -0.20, -0.04), after adjusting for covariates. Being in middle and highest tertiles of SBP variability (by DEV) and glucose variability (by SD, CV) were associated with poorer episodic memory and executive function, with effect sizes increasing across higher degrees of variability. Association between non-HDL variability and CF was non-linear: only being in the middle tertile was associated with poorer executive function (by CV, β = -0.12, 95% CI = -0.23, -0.01). Association between BMI variability and CF was bidirectional across degrees of variability: being in the middle tertile of variability was associated with better GCS (by CV, β = 0.09, 95% CI = 0.02, 0.16); whereas being in the highest tertile (by DEV) was associated with poorer executive function, although attenuated after adjusting for APOE ε4 status. Conclusions VRF variability from early life is a potential target for lowering risk of late-life cognitive dysfunction.