Assessing the burden of worsening self-reported vision in older Americans using the Health and Retirement Study

Background Visual disorders are estimated to affect about 3.6% of Americans by 2020. Vision impairment has a substantial impact on individuals functionally. Vision impairment is associated with many comorbidities, functional limitations, as well as higher caregiving needs. This study aims to assess the transition to self-reported vision impairment and the effect on an individual's functionality and ability to live independently. Methods The Health and Retirement Study is an ongoing, biennial survey of older Americans that collects an abundance of data on an individual's family, health, and financial circumstances. The explanatory variable of interest was a dichotomous variable of vision impaired or not vision impaired as defined by their self-reported vision status. Outcomes studied were functional limitations defined as difficulty performing activities of daily living (ADLs) and instrumental activities of daily living (IADLs), likelihood of move to nursing home, and both informal and formal caregiving needs. Descriptive univariate and bivariate statistics were performed to describe the demographics of the sample over the ten waves of data used, from 1995-2012. Multivariable logistic regressions controlling for sociodemographic characteristics, were performed to test the association between vision status and outcomes of interest for specific representative waves. Panel data methods of logistic regression were used to measure the association between a transition to impaired vision and the effect on functional limitations and nursing home residence while controlling for potential confounders and within-patient correlation. Specifically, mixed-effects logistic regressions with subject-level random-effects were conducted for the binary outcomes of nursing home residence, difficulty in performing at least one ADL, and difficulty in performing at least one IADL. Results The prevalence of overall vision impairment was, on average, about 6%. On average, about 5% and 7% of respondents had near and distal vision impairment, respectively. The overall prevalence of nursing home residence in our sample was approximately 2%. Prevalence of ADL and IADL limitations was stable across waves but ranged from 1%-20% for each of the six ADL and five IADL limitations. More respondents reported receiving informal caregiving versus formal caregiving; however the hours of caregiving received was similar for both formal and informal caregiving recipients. Multivariable logistic regression for three representative waves found numerically higher odds of a nursing home residence for vision impaired individuals compared to individuals with no vision impairment, though the difference was not significant. The odds ratios for 1998, 2006, and 2012 were 1.08 (95% CI: 0.64, 1.84), 1.09 (95% CI: 0.64, 1.86), and 1.22 (95% CI: 0.76, 1.97), respectively. Significant associations were found between vision impairment and functional limitations. Using a similar approach, the odds ratio for difficulty performing at least one ADL was 2.57 (95% CI: 2.16, 3.05), 2.37 (95% CI: 1.97, 2.84), and 2.31 (95% CI: 1.92, 2.77) in 1998, 2006, and 2012, respectively. The odds ratios for difficulty performing at least one IADL was 3.78 (95% CI: 3.15, 4.53), 3.94 (95% CI: 3.26, 4.76), and 3.49 (95% CI: 2.89, 4.22) in 1998, 2006, and 2012, respectively. In the mixed-effects logistic regression of the total study panel, a transition from no vision impairment to vision impairment was significantly associated with 1.37 times higher odds of a nursing home residence (95% CI: 1.05, 1.78), a 2.96 times higher odds of difficulty performing at least one ADL (95% CI: 2.71, 3.24), and a 4.02 times higher odds of difficulty performing at least one IADL (95% CI: 3.70, 4.37), after controlling for confounders. Discussion We found the estimated prevalence of visual impairment and functional limitations within our sample, and the association between the two, to remain relatively stable over time. Additionally, we found an age-adjusted transition to self-reported vision impairment within n in ividual to have significant detrimental effects on the ability to live independently and perform ADLs and IADLs without difficulty. Thus, prevention of this transition can substantially impact an individual's quality of life, and benefits derived from early detection and improved treatment of medical conditions that contribute to vision loss can have extensive value beyond simply improving vision.

Public health