Maintaining level of modifiable dementia risk scores is associated with better cognitive outcomes than increasing risk scores: A population-based prospective cohort study

BACKGROUND: A brain healthy lifestyle, consisting of good cardiometabolic health and being cognitively and socially active in midlife, is associated with a lower risk of cognitive decline years later. However, it is unclear whether lifestyle changes over time also affect the risk for mild cognitive impairment (MCI)/dementia, and rate of cognitive decline. OBJECTIVES: To investigate if lifestyle changes over time are associated with incident MCI/dementia risk and rate of cognitive decline. DESIGN: Population-based prospective cohort study SETTING: Personality and Total Health (PATH) Through Life Study cohort (Australia). PARTICIPANTS: 4,777 participants (50.4% women), recruited between 2000 and 2002, who were 40-44 and 60-64 years old at baseline, without a prevalent dementia diagnosis. Participants had to have cognitive outcome measures available after baseline. MEASUREMENTS: Various measurements (neurocognitive assessment, blood pressure) and survey responses (demographics, cognitive, social, and physical activity, smoking, alcohol consumption, body height and weight, depression, and previous diagnoses) were collected approximately every four years. A brain-healthy lifestyle was operationalized via two well-validated modifiable dementia risk scores, the LIfestyle for BRAin health (LIBRA) score and the modifiable part of the Australian National University Alzheimer's Disease Risk Index (ANU-ADRImod). Their change over time was estimated using latent growth curve modelling, and their association with cognition and incidence of MCI/dementia was investigated using parallel process modelling and Cox regression analysis. RESULTS: Within those aged 60-64 years at baseline (n=2,409), 211 cases of incident MCI/dementia were recorded over a median follow-up time of 12.2 years. On average, individuals' LIBRA and ANU-ADRImod increased (i.e., worsened) over time, but individuals whose scores increased one standard deviation (SD) less had a 19.0-24.6% lower risk for MCI/dementia (hazard ratio (95% confidence interval): LIBRAchange over time=0.754 (0.664-0.857), ANU-ADRImod, change over time=0.810 (0.71-0.915)), while controlling for the risk score at baseline and multiple potential confounders. Various associations between dementia risk score trajectories and cognitive performance trajectories were observed. CONCLUSIONS: Efforts to maintain a brain healthy lifestyle could reduce the risk for MCI or dementia, and slow cognitive decline.