Age, race, and education as moderators of post-stroke cognitive decline following dental care

Abstract Post-stroke cognitive decline (PSCD) poses a significant challenge to long-term recovery and quality of life following stroke, influenced by both fixed biological factors and modifiable health behaviors such as oral and dental care. In this data-driven exploratory analysis of the PREMIERS Phase II randomized trial (ClinicalTrials.gov NCT#02541032), we examined the moderating effects of clinical, biological, and demographic characteristics on the relationship between dental care and PSCD over a 12-month period. The study included 280 stroke/transient ischemic attack (TIA) survivors who received either intensive or standard dental care. Cognitive outcomes were assessed using the Montreal Cognitive Assessment (MoCA) at baseline and follow-up, with change in MoCA score as the primary outcome. Lasso regression was applied for empirically based feature selection of moderators, and bootstrapped multiple linear regression demonstrated that increased dental visits predicted relatively better cognitive outcomes in older adults (age interaction-term β = -0.664, p < 0.001), Black participants (race interaction-term β = -0.475, p < 0.05), and those with low-intermediate education levels (education interaction-term β = 0.413, p < 0.05). Exploratory graphs revealed that older adults, Black adults, and adults with low-intermediate education showed greater cognitive improvement with higher dental visit frequency, with the final model (including selected moderators) significantly predicting PSCD ( F (11, 268) = 10.51, p = 5.17 x 10 −16 ). These findings highlight the potential of equity-focused, precision-medicine interventions that incorporate dental care to mitigate PSCD in vulnerable stroke populations.