The association between sedentary behavior and eye diseases: a prospective study from the UK Biobank

Abstract

Abstract Background Sedentary behavior (SB) is a well-established risk factor for multiple chronic diseases. However, evidence regarding its association with common ocular disorders remains limited and inconsistent. This study investigated the prospective associations between SB and the risks of incident age-related macular degeneration (AMD), glaucoma, dry eye disease (DED), and retinal vein occlusion (RVO). Methods We conducted a prospective cohort analysis using data from the UK Biobank. Separate analytical cohorts were established for AMD ( n = 374,552), glaucoma ( n = 370,997), DED ( n = 374,398), and RVO ( n = 374,924). During a median follow-up of approximately 14.04 years, 7,266 AMD, 8,112 glaucoma, 452 DED, and 786 RVO cases were identified. Pre-existing ocular diseases were identified using ICD-10 records and self-reported diagnoses. Incident AMD, glaucoma, and RVO were identified from ICD-10 records, whereas incident DED was identified by self-report. Cox proportional hazards models, restricted cubic spline (RCS) analyses, interaction analyses, and sensitivity analyses were performed. Results In unadjusted models, higher SB was associated with increased risks of AMD, glaucoma, DED and RVO. After multivariable adjustment, only a borderline statistically significant association remained for AMD (HR = 1.01; 95% CI: 1.00–1.02), whereas no significant associations were observed for glaucoma, DED, or RVO. RCS analyses demonstrated mild J-shaped associations between SB and AMD and glaucoma, with the lowest risk observed at approximately 3.7 h/day; however, these associations became nonsignificant after full adjustment. Stratified interaction analyses identified significant effect modification for AMD by body mass index (BMI), glycated haemoglobin (HbA1c), and physical activity (PA) level, and for DED by sex and BMI. Sensitivity analyses, including additional adjustment for physical activity and sleep duration and analyses of domain-specific sedentary behaviors, did not alter the overall conclusions. Conclusion After adjustment for potential confounders, no statistically significant independent associations were observed between SB and incident AMD, glaucoma, DED, or RVO. However, these findings should not be interpreted as evidence against a causal relationship. Further studies incorporating objective exposure assessment, repeated measurements, and causal inference approaches are warranted to clarify the potential role of sedentary behavior in ocular disease development.