ABSTRACT Background: Diabetic retinopathy (DR) is a leading cause of preventable visual impairment. Comparative evidence from rural and urban secondary care settings in Nigeria is limited. We evaluated prevalence, pattern, awareness, screening uptake, and clinical correlates of DR among adults with diabetes attending rural and urban secondary health facilities in Lagos. Methods: A comparative hospital-based cross-sectional study recruited 288 adults with diagnosed diabetes (144 rural and 144 urban). Data collection included structured questionnaires, clinical records review, and comprehensive ocular examination with clinical DR grading. Associations were examined using Chi-square tests and multivariable logistic regression; the adjusted model included age, sex, duration of diabetes, blood-pressure status, and area of residence. Statistical significance was set at P < 0.05. Results: Mean age was 60.4 ± 10.8 years; 64.2% were female. Overall awareness of DR was 16.3% (47/288), markedly lower in rural (0.7%, 1/144) than urban participants (31.9%, 46/144). Overall DR prevalence was 29.5% (85/288), 28.5% in rural and 30.6% in urban participants. Non-proliferative DR accounted for 90.6% (77/85) of cases; diabetic macular edema occurred in 10.6% (9/85). Prior retinal examination uptake was low (rural 35.4%; urban 38.2%). HbA1c results were available for only 17 participants (5.9%). In multivariable analysis, longer diabetes duration (adjusted odds ratio AOR per year 1.10; 95% confidence interval CI 1.04–1.16; P = 0.000623) and poor glycemic status (random blood glucose ≥200 mg/dL; AOR 2.93; 95% CI: 1.64–5.21; P = 0.000267) were independently associated with DR. Conclusion: DR is common among clinic-attending adults with diabetes in Lagos, while awareness and screening uptake are low, particularly in rural settings. Longer disease duration and clinic-based glycemic measures were associated with DR. Given the cross-sectional, hospital-based design and limited HbA1c data, these findings indicate priorities for service strengthening and should be confirmed in prospective, population-based studies.
Alabi et al. (2026) studied this question.