Relationship of Visual Impairment and Poor Mobility With Mortality

Older adults with visual impairment (VI) have a greater risk of mortality, but the reasons are poorly understood. We have shown that older adults with VI are more likely to have poor mobility performance on the short physical performance battery (SPPB). In this analysis, we examined whether VI predicted mortality over 10 years and if this was related to poor baseline mobility (SPPB < 9).

We analyzed 2457 older adults (mean age 75.6 (±2.8) years, 38.5% black, 51.9% female) who completed vision testing at the year 3 visit in the Health, Aging and Body Composition study. Cox proportional hazards models for mortality were right-censored at 10 years and adjusted for demographic and clinical comorbidities. VI (visual acuity < 20/40 or log contrast sensitivity < 1.55 or stereoacuity > 85) and SPPB < 9 were tested as main predictors and their interaction was tested.

In separate multivariable models, VI (HR 1.511, 95% CI [1.335-1.709], p < 0.0001) and SPPB < 9 (HR 1.442, 95% CI [1.210-1.717], p < 0.0001) each predicted mortality. When including both poor mobility and vision variables as main effects, both poor mobility and impaired vision remained significant predictors of mortality in all models (all p < 0.001). When adding poor mobility (as a main effect and interaction with VI) and using unimpaired vision and mobility as the reference, those with only VI (HR 1.467, 95% CI [1.287-1.672], p < 0.0001) or only poor mobility (HR 1.380, 95% CI [0.963-1.979], p = 0.0792) had similar HRs, while those with both VI and poor mobility had an increased mortality risk (HR 2.035, 95% CI [1.643-2.522], p < 0.0001), but the interaction was not significant (p = 0.981).

Older adults with both VI and poor mobility are at an additive increased risk of mortality. Future interventions may want to target older adults with both VI and poor mobility to improve survival.