Global prevalence of diabetic retinopathy and projection of burden through 2045

Authors: Teo ZL, Tham YC, Yu M, Chee ML, Rim TH, Cheung N, et al.

Geographical coverage: Western Pacific, South East Asia, North America, Caribbean, the Middle East, North Africa, Europe, South and Central America and Africa

Sector: Burden of disease

Sub-sector: Epidemiology, prevalence

Equity focus: Not reported

Study population: Patients with diabetic retinopathy

Review type: Prevalence review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background: With a rapidly growing ageing global population, increasing lifespans of people living with diabetes mellitus (DM), and lifestyle changes leading to an increased risk for DM, a higher burden of diabetic retinopathy (DR) and demand for eye care and treatment are expected. Thus, up-to-date and accurate estimation of the prevalence of DR is critical to devising health policies and allocating adequate resources to address this global problem.

Objectives: To update the global prevalence of DR and to provide future projections of the number of people with DR, vision-threatening DR (VTDR), and clinically significant macular edema (CSME) through 2045.

Main findings:

In total, 59 population-based studies were included in the final analysis. Fifty-nine studies were from 27 countries and consisted of 9,685 patients with DR. By International Diabetes Federation world regions, 17 study populations were from Western Pacific (WP), 14 were from South East Asia (SEA), 12 were from North America and the Caribbean (NAC), seven were from the Middle East and North Africa (MENA), six were from Europe, two were from South and Central America (SACA), and one was from Africa. The included studies were of varying quality with the quality assessment scores ranging from 0 to 4; four studies scored 0, 20 studies scored 1, 20 studies scored 2, 12 studies scored 3, and three studies scored 4.

Of the included studies, 56 reported complete data on region, habitation type, response rate, year of study, and DR diagnosis method, while a subset of 50 provided ethnicity information, 49 reported gender proportion data and 27 provided mean age data.

Among individuals with diabetes, global prevalence was 22.27% (95% confidence interval [CI], 19.73-25.03) for DR, 6.17% (95% CI, 5.43-6.98) for VTDR, and 4.07% (95% CI, 3.42-4.82) for CSME. However, it was estimated that the global number of adults with DR, VTDR and CSME will increase to 160.5 million, 44.8 million, and 28.6 million, respectively, by 2045.

Prevalence of DR was highest in Africa (35.90%) and NAC (33.30%), and was lowest in SACA (13.37%). Adjusted meta-regression models showed that among people with diabetes, Hispanics (odds ratio [OR], 2.92; 95% CI, 1.22-6.98) and Middle Easterners (OR, 2.44; 95% CI, 1.51-3.94) were more likely to have DR compared with Asians. Sensitivity analysis excluding these three studies (that scored 4 on the quality assessment scale) showed that findings remained broadly similar.

Overall, the findings suggested that approximately one in five people with diabetes worldwide have DR, and the number of people losing vision as a result of DR may continue to rise.

The authors highlighted the need for high quality population-based studies of DR, especially in Africa and SACA, so the findings would be used in planning health care strategies to prevent diabetes-related vision loss.

Methodology:

Authors conducted a search in PubMed, Medline, Web of Science and Scopus to identify population-based studies on DR, VTDR and CSME prevalence, published in the English language only and up to March 2020.

The review authors included population-based studies that clearly defined random or clustered sampling procedures. In addition, the studies were included if they: (1) had 60% or more participation rate of the eligible population; (2) provided DR, VTDR or CSME prevalence, or a combination thereof, among the DM group(s); (3) clearly defined DM with at least one of the following used definition for DM diagnosis: fasting blood glucose ≥7 mmol/l, random blood glucose of more than 11.1 mmol/l, oral glucose tolerance test results of 11.1 mmol/l or more, glycated haemoglobin findings of 6.5% (48 mmol/mol) or more, self-reporting of physician-diagnosed DM, existing DM treatment and medical records; and (4) defined DR by the presence of retinal haemorrhages, microaneurysms, cotton-wool spots, pan retinal photocoagulation laser scars, or a combination found on colour fundus photographs, dilated slit-lamp examination by an ophthalmologist, or a combination thereof.

The authors reviewed the reference lists of relevant articles, such as previous country or region-based systematic reviews and meta-analyses providing DR prevalence to identify additional relevant publications. In addition, authors obtained original unpublished DR, VTDR and CSME data from Asian population studies via the Asian Eye Epidemiology Consortium.

The quality of the included studies was assessed using a tool developed by Hoy et al. 2012. The tool, modified from the Grades of Recommendation, Assessment, Development and Evaluation approach, assessed both internal and external validity. The findings were synthesised quantitatively using random-effects meta-analysis, and meta-regression. The authors also conducted sensitivity analyses to evaluate the effect of study quality on overall results. Heterogeneity was assessed using I² statistics. Publication bias among included studies was assessed using funnel plots and Begg and Mazumdar tests (for DR, VTDR and CSME).

Applicability/external validity: The review reported using a specific risk-of-bias tool to assess internal and external validity. However, authors did not provide further details on external validity.

Geographic focus:

Authors included studies from a wide range of regions: 17 study populations were from Western Pacific (WP), 14 were from South East Asia (SEA), 12 were from North America and the Caribbean (NAC), seven were from the Middle East and North Africa (MENA), six were from Europe, two were from South and Central America (SACA), and one was from Africa.

Summary of quality assessment:

The research process was thorough, involving an extensive search, reference list checks and outreach for unpublished data. The results were integrated through meta-analysis and meta-regression, and a sensitivity analysis was performed to determine the impact of study quality on the overall outcomes. Variability among studies was quantified using I² statistics, and potential publication bias was examined with funnel plots and Begg and Mazumdar tests. While the review’s limitations were recognised, it was limited to English-language articles, and it is not specified whether two reviewers independently conducted the screening and data extraction processes. For these reasons, a low confidence was attributed to the conclusions about the effects of this review.