INTRODUCTION: Chronic pain (CP) was associated with cognitive impairment in previous studies. However, the longitudinal association between CP and dementia remains under debate. We aimed to assess the prospective link between CP and long-term dementia risk in a population-based cohort of older participants, considering covariables linked to CP and cognitive functioning. METHODS: The study sample was selected from the PAQUID study, an ongoing cohort of older community-dwellers aged 65 years and over at baseline; Information regarding CP and analgesics consumption was collected using questionnaires. Dementia was clinically assessed every 2 years. The population was divided into 4 groups according to CP and analgesic drugs intake (CP+/A+, CP+/A-, CP-/A+, CP-/A-). An illness-death model was used to estimate the link between CP and incident dementia risk controlled for sex, educational level, comorbidities, depression, antidepressant drugs and analgesics. RESULTS: Five hundred ninety three participants (364 women) who completed a CP questionnaire, were included. They were followed-up over 24 years (mean follow-up: 11.3 years, SD 7.3). A total of 223 participants (32.5%) had CP, among them 88 (38.6%) took analgesic drugs. Compared to CP-/A- group, CP+/A+ participants had a higher risk of developing dementia in the univariate model (hazard ratio (HR) = 1.73, 95%CI:1.18-2.56; p = 0.0051). However, these results did not persist in the multivariate models (aHR = 1.23, 95%CI:0.88-1.73; p = 0.23). No significant risk for dementia were observed in CP-/A+ and CP+/A- (HR = 1.30, 95%CI:0.84-2.01; p = 0.23 and HR = 1.36, 95%CI:0.95-1.96; p = 0.09, respectively). CONCLUSION: Our results failed to show a significant relationship between the presence of CP and long-term dementia risk, suggesting that the cognitive decline associated with CP observed in the literature does not appear to be related to Alzheimer's disease or related disorders.< Réduire