Associations between sleep patterns and diabetic retinopathy: insights from cross-sectional and longitudinal evidence

To investigates the associations between various sleep patterns and DR prevalence and incidence.

We conducted a multi-stage analysis combining cross-sectional and longitudinal data from UK Biobank (UKB) and Shanghai Cohort of Diabetic Eye Disease (SCODE), alongside two-sample Mendelian randomisation (MR) using genome-wide association study (GWAS) summary statistics. Sleep traits included daytime napping, sleep duration, chronotype, snoring, insomnia, dozing and sleep apnoea. Associations with DR prevalence and incidence were evaluated using logistic and Cox regression models, adjusting for key covariates. Daytime sleepiness was further assessed using the Epworth Sleepiness Scale (ESS). MR analysis employed inverse variance weighting and sensitivity tests to assess causality, heterogeneity, and pleiotropy.

UKB cross-sectional data showed that frequent daytime napping was associated with higher DR prevalence (Usually vs. Never/rarely, adjusted odds ratio [OR] = 1.847, P < 0.001), while SCODE data linked more frequent napping (P = 0.013), longer daytime sleep duration (P = 0.023), and higher Epworth Sleepiness Scale scores (P = 0.046) to higher DR prevalence. Longitudinal analysis found that frequent daytime napping (Usually vs. Never/rarely, hazard ratio [HR] = 2.317, P < 0.001) was associated with DR incidence. Mendelian randomisation analysis suggested a possible link between daytime napping and DR (IVW OR = 1.46, P = 0.04), though this did not withstand multiple testing correction.

Excessive daytime sleepiness may be associated with an increased risk of DR.