INTRODUCTION: Hearing impairment is associated with dementia. We aimed to clarify the association between hearing impairment and future cognitive test performance measured by the Montreal Cognitive Assessment Scale (MoCA), adjusted for confounders, avoiding reverse causation through long follow-up. METHODS: We used the Norwegian population-based longitudinal cohort study, The Trøndelag Health Study (HUNT). At baseline, we invited all residents 20+ for an audiometric hearing assessment, and at 20+ years follow-up, we cognitively assessed all persons 70+ including MoCA adjusted for hearing impairment. We analyzed the association using linear regression. RESULTS: We included 6879 persons (mean 56.1 years, standard deviation 6.2). At follow-up, the MoCA score was -0.25 (95% confidence interval [CI] -0.35, -0.14), per 10 dB increase in hearing threshold and for persons <
85 years, -0.31 (95% CI -0.42, -0.20). DISCUSSION: This study finds a long-term association between hearing impairment and dose related reduced cognitive performance, particularly in those aged <
85. CLINICAL TRIAL REGISTRATION: ID NCT04284384, hearing impairment as a risk factor for dementia in older adults. HIGHLIGHTS: Hearing loss predicts long-term cognitive decline measured by MoCA over 20+ years. Long follow-up is crucial to avoid reverse causation in the hearing-cognition relationship. A 10 dB hearing threshold increase is linked to a 0.25-point reduction in MoCA score. Strongest cognitive decline associations are seen in people aged below 85 years. The association remained after excluding hearing-dependent tasks in the MoCA.