Clinical & Experimental Ophthalmology , Feb. 2025
Qingsheng Peng, Tyler Hyungtaek Rim, Zhi Da Soh, Miao Li Chee, Yih-Chung Tham, Zhuoting Zhu, Simon Nusinovici, Charumathi Sabanayagam, Ah. Young Leem, Chan Joo Lee, Byoung Kwon Lee, Sungha Park, Sung Soo Kim, Hyeon Chang Kim, Marco Chak Yan Yu, Tien Yin Wong, Ching-Yu Cheng
Summary
Background
Previously, based on retinal photographs, we developed a deep-learning algorithm to predict biological age (termed, RetiAGE) that was associated with future risks of morbidity and mortality. This study specifically aimed to evaluate the performance of RetiAGE in predicting future risks of chronic obstructive pulmonary disease (COPD).
Methods
RetiAGE scores were generated from retinal images in the UK Biobank and stratified into tertiles. We used Cox proportional hazards models to evaluate the longitudinal association between RetiAGE and incident COPD, adjusting for calendar age, gender, smoking, asthma history, and socio-economic status. In addition, we performed a cross-sectional analysis using generalised linear models to examine the association between RetiAGE and baseline respiratory function, specifically the forced expiratory volume in 1 s to forced vital capacity ratio (FEV1/FVC) and peak expiratory flow (PEF), adjusting for the same confounders.
Results
Among 45 438 UK Biobank participants without a history of COPD at baseline, 448 (0.9%) developed COPD over a mean follow-up period of 9.8 ± 0.7 years. Participants in the moderate-risk and high-risk tertiles of RetiAGE had significantly lower baseline respiratory function (all p < 0.05) and a higher risk of incident COPD (HR = 1.60; 95% CI, 1.18–2.19) compared to the low-risk tertile, after adjusting for confounders. Adding RetiAGE to the multivariable risk model improved predictive performance, as demonstrated by significant enhancements in C-statistics (p < 0.001) and likelihood ratio tests (p = 0.002).
Conclusion
Our deep-learning-based retinal aging biomarker, RetiAGE, can potentially stratify the risk of developing COPD.