Predictors of cognitive resilience in the old-old: An observational study using real-world electronic health record data.

INTRODUCTION: Understanding factors associated with cognitive resilience is critical for developing interventions to preserve cognitive health in the old-old. Our aim is identifying clinical predictors of cognitive resilience in a large real-world cohort of adults aged 75+. METHODS: Longitudinal analysis of electronic health records from Johns Hopkins primary care clinics between January 1, 2014, and August 31, 2025. A minimum of two follow-up visits at least a year apart were required. To estimate the associations between predictors and cognitive resilience, we employed a Cox model with cognitive disorder diagnosis as the outcome. RESULTS: Of 43,178 patients, 6563 developed cognitive disorder during follow-up. Psychiatric conditions were strongly associated with decreased resilience. Antihypertensive use was associated with increased resilience. Black race and Medicaid enrollment were associated with decreased resilience. DISCUSSION: Most adults reaching 90 remained cognitively healthy. Resilience reflected physiological reserve and care engagement, while vascular, psychiatric, and social factors reduced resilience.