Abstract
OBJECTIVE: The Brain Care Score (BCS), previously validated to predict the risk of age-related brain diseases, incorporates 12 modifiable risk factors, including blood pressure and laboratory measurements. In the U.S., fewer than 50% of individuals recall these measurements, limiting the BCS's accessibility. This study aims to evaluate the predictive validity and discriminatory power of a BCS derived from participant responses for incident stroke. METHODS: We performed retrospective analyses using the prospectively collected COSMOS data. The BCS, ranging 0-19 points, was calculated from modifiable risk factors derived from participant responses only. A Cox proportional hazard model, adjusted for sex and age was used to estimate the association between a higher BCS, indicating better brain health, and incident stroke. RESULTS: Among 17 150 participants (median age 70.8, 59.6% female) free of stroke and TIA at baseline and with complete BCS data, 187 (1.1%) experienced a stroke during a mean follow-up duration of 3.6 (SD:0.7) years. A 5-point higher BCS was associated with a 36% lower stroke risk (HR:0.64 [95%CI:0.48-0.84], c-statistic:0.68). DISCUSSION: A BCS derived from participant responses showed similar predictive performance and discriminatory ability compared to previous validation studies that use physical and laboratory measurements. Future studies could consider incorporating the BCS derived from participant responses when physical and laboratory measurements are not readily available.