Abstract Body (Do not enter title and authors here): Background: Metabolic syndrome (MetS), a cluster of risk factors, has been linked to higher risk of dementia in older adults. However, the composite impact of long-term variability of metabolic parameters from childhood on midlife cognitive function is unclear.
Methods: We studied 1145 midlife participants of the Bogalusa Heart Study (age at enrollment: 12.2 ± 5.4 years, follow-up: 36.1 ± 3.4 years, age at last exam (midlife): 48.3 ± 5.0 years, 60.9% women, 33.7% Black participants) with ≥3 measurements of metabolic parameters (systolic blood pressure (SBP), fasting blood glucose (FBG), triglyceride (TG), high density lipoprotein (HDL), and body mass index (BMI)) across follow-up. MetS was defined by the National Cholesterol Education Program guidelines, but using BMI >30 kg/m² for abdominal obesity. Long-term variability of SBP, FBG, TG, HDL, and BMI from childhood were measured as the residual standard deviation (RSD). Repeated measurements of each parameter were fitted in random-effect models to obtain individual growth curves. Using the growth curves, RSD was obtained as the SD of the difference between each observed and predicted parameter level. A composite variability score (range, 0-10) was defined by 1) assigning 0 points to the lowest tertile of RSD and 2 points to the highest tertile for each parameter, then 2) summing the points for all 5 parameters. Midlife cognitive function was measured by 8 neuropsychological (NP) tests. NP test results were categorized into 3 distinct NP profiles (optimal, average, mixed-low) using cluster analysis. Associations between the composite variability score and NP profiles were analyzed using multinomial logistic regression models, adjusting for education status. Sensitivity analysis excluding participants with MetS at midlife were also performed.
Results: A total of 348 (30.4%) participants had MetS at midlife. The odds of having mixed-low cognitive function at midlife increased by 13.3% (OR 1.13, 95% CI 1.03, 1.24) for a 1-unit increase in the composite variability score, compared to optimal cognitive function, adjusted for education status. Sensitivity analysis showed consistent results in participants without MetS at midlife (OR 1.18, 95% CI 1.05, 1.33).
Conclusions: A graded association was observed between composite long-term variability in metabolic parameters from childhood and poorer midlife cognitive function. Lowering variability of metabolic parameters may be beneficial to midlife cognitive function.