Self-reported blindness was associated with higher odds of elevated loneliness scores in a cross-sectional analysis of more than 61,000 patients with diagnosed ocular conditions, while diabetic retinopathy showed a smaller association.
The study, published in JAMA Ophthalmology, included 61,349 patients from the National Institutes of Health All of Us cohort who completed the 8-item UCLA Loneliness Scale and had at least 1 diagnosed ocular condition. Patients without a history of eye care were excluded because they could have undiagnosed major eye disease. The mean age was 60 years, and 63% of patients were female.
Researchers evaluated whether major eye diseases, defined in the study as age-related macular degeneration, diabetic retinopathy, and glaucoma, were associated with high loneliness scores. The primary outcome was a UCLA Loneliness Scale score above the cohort median. Models were adjusted for age, gender, race and ethnicity, income, employment status, educational level, insurance status, and other eye diseases, with additional models also adjusting for self-reported blindness.
Self-reported blindness was the strongest positive finding. In adjusted analyses, patients who reported blindness had 1.61 times the odds of high loneliness scores compared with patients who did not report blindness.
Among the major eye diseases studied, diabetic retinopathy was the only condition associated with high loneliness scores after adjustment. In models adjusted for sociodemographic factors and other eye diseases, proliferative diabetic retinopathy was associated with 1.28 times the odds of high loneliness scores, and nonproliferative diabetic retinopathy was associated with 1.16 times the odds.
The researchers noted that the larger nonproliferative diabetic retinopathy cohort may have provided greater statistical power and that the lower bounds of the diabetic retinopathy confidence intervals were close to 1.0, raising the possibility of a small effect size or residual confounding. They also noted that they did not adjust for multiple analyses and that correction for multiple analyses could further reduce the statistical significance of the findings.
After additional adjustment for self-reported blindness, only nonproliferative diabetic retinopathy remained associated with high loneliness scores. The association with proliferative diabetic retinopathy no longer met statistical significance in that model.
In contrast, neither exudative nor nonexudative age-related macular degeneration was associated with high loneliness scores after adjustment. No glaucoma-related diagnosis, including primary open-angle glaucoma, primary angle-closure glaucoma, secondary glaucoma, unspecified or other glaucoma, or glaucoma suspect status, was associated with high loneliness scores in adjusted models.
Some unadjusted inverse associations for age-related macular degeneration and glaucoma-related diagnoses were no longer present after adjustment for sociodemographic factors and other eye diseases. The null adjusted glaucoma findings differed from prior prospective work cited by the researchers, in which baseline loneliness was associated with subsequent glaucoma incidence.
The researchers suggested that the diabetic retinopathy association may reflect factors beyond vision loss alone. They noted that diabetic retinopathy is more common among working-age adults than glaucoma or age-related macular degeneration and that systemic health outcomes of diabetes may co-occur with diabetic retinopathy and contribute to loneliness.
The study could not establish causality or directionality. Because the analysis was cross-sectional, it could not determine whether eye disease, blindness, systemic disease burden, or other factors preceded loneliness.
Several additional limitations should be considered. Blindness was assessed by self-report rather than structured visual acuity measures, and diagnosis-code classifications did not capture details such as symmetry of visual impairment or fellow-eye involvement. About 75% of the cohort was White, which may limit generalizability. Selection bias may also be present.
“In this large diverse cohort, loneliness was associated with self-reported blindness and DR but not with glaucoma or AMD,” wrote Zachary Ray Wakefield, BA, of the Department of Ophthalmology, Byers Eye Institute, Stanford University, and colleagues.
The researchers did not study or recommend loneliness screening; their report was limited to the observed associations described above. Any extension to screening or intervention in patients with diabetic retinopathy or self-reported vision loss would require dedicated study.
Disclosures: The researchers reported no conflicts of interest.
Source: JAMA Ophthalmology
