The longitudinal association between cognitive impairment and incident visual impairment in a multiethnic Asian population: a prospective cohort study

Background The relationship between baseline cognitive impairment (CI) and incident visual impairment (VI) in Asians is unclear. Objective To determine the associations between baseline CI with incident VI and visual acuity (VA) at 6-year follow-up in multiethnic Asians. Design Cohort. Setting Population-based. Subjects Two thousand three hundred and twenty-four adults aged >= 60 years from the Singapore Epidemiology of Eye Diseases Study (response rate 64%). Methods CI was defined using the validated Abbreviated Mental Test (AMT). VA was objectively measured using a LogMAR chart. Any incident VI was defined as having no VI (Snellen's VA better than or equal to 20/40) at baseline but present (VA worse than 20/40) at 6-year follow-up. VI severity was defined according to the International Classification of Diseases, 11(th) Revision. Associations were assessed using logistic and linear regression models. Results Of the 2,324 participants, 248 had CI at baseline. Presence of baseline CI was associated with more than twice the odds of any incident VI, incident mild and moderate-severe VI (OR [95% confidence interval]: 2.48 [1.55-3.90], 2.07 [1.17-3.55], and 2.61 [1.36-4.93], respectively) and worse VA (beta [95% confidence interval]: 0.026 [0.006-0.046]) at 6-year follow-up. The leading causes of incident VI were cataract and under-corrected refractive error. Conclusions Older adults with CI had more than double the odds of VI development and poorer VA than their cognitively intact counterparts, and most causes of incident VI were correctable. Strategies such as targeted vision screening and early intervention for early detection and management of vision loss in patients with cognitive decline are warranted.