Abstract
PURPOSE: To investigate the associations of omics-based biological aging with diabetic retinopathy (DR) and life expectancy among individuals with DR.
METHODS: We included 68,672 UK Biobank participants with prediabetes or diabetes in the metabolomic cohort and 8716 in the proteomic cohort, from which biological age was estimated using a metabolomic aging score and proteomic age (ProtAgeGap), respectively. Associations with prevalent and incident DR were assessed using logistic regression and Cox proportional hazards models, respectively. Life expectancy was estimated from survival curves.
RESULTS: Cross-sectionally, both the higher metabolomic aging score and ProtAgeGap were associated with a higher risk of prevalent DR. Prospectively, during a median follow-up of 13.40 years, 2491 participants (3.7%) developed DR among those free of DR at baseline in the metabolomic cohort and 358 (4.14%) in the proteomic cohort. Higher biological aging was associated with increased DR risk, with hazard ratios of 1.73 (95% confidence interval [CI], 1.59-1.88) for the metabolomic aging score and 1.12 (95% CI, 1.07-1.17) for ProtAgeGap. These associations were independent of and additive to glycated hemoglobin levels and diabetes duration. At age 45 years, individuals with DR had reduced life expectancy compared with those without DR. Among those with DR, the highest quartile of metabolomic aging was associated with a reduced life expectancy compared with the lowest quartile.
CONCLUSIONS: Advanced metabolomic and proteomic aging were associated with a higher risk of DR, particularly in individuals with poor glycemic control or longer diabetes duration, while advanced metabolomic aging was associated with reduced life expectancy among those with DR.