Abstract
INTRODUCTION: Family relationships and friendships are important for cognitive health, yet how their quality relates to dementia risk remains poorly understood. METHODS: This prospective cohort study included 104,093 UK Biobank participants aged 40 to 69 years who were dementia free within the first 4 years. Relationship satisfaction was categorized as satisfied or dissatisfied. Cox regression models assessed associations. RESULTS: Over 13.56 years, 1359 developed dementia. In the fully adjusted model, dissatisfied (vs. satisfied) family relationships were associated with elevated dementia risk (hazard ratio [HR]: 1.334, 95% confidence interval [CI]: 1.069-1.663), driven primarily by very dissatisfied participants (HR: 1.663). Friendship dissatisfaction showed no significant association with dementia risk (HR: 1.060, 95% CI: 0.752-1.494). Overall relationship dissatisfaction was associated with increased dementia risk (HR 1.265, 95% CI 1.035-1.546), though becoming non-significant after adjustment for additional psychological factors. DISCUSSION: Family relationship quality may represent a risk factor of dementia warranting consideration in prevention strategies.