Prevalence and Risk Factors of Diabetic Macular Edema in Type 1 Diabetes: A Systematic Review and Meta-Analysis.

Topic: This systematic review and meta-analysis investigates the prevalence, natural history, clinical risk factors, and distinctive characteristics of diabetic macular edema (DME) in type 1 diabetes mellitus (T1D), highlighting comparisons with type 2 diabetes mellitus (T2D).

Clinical relevance: DME is the most common vision-threatening complication of diabetic retinopathy, significantly impacting quality of life. While extensively studied in T2D, data in T1D remains limited, restricting targeted prevention and management. Given the increasing longevity of individuals with T1D, understanding its specific risk profile is critical to refine screening and treatment guidelines.

Methods: This systematic review and meta-analysis was registered (PROSPERO: CRD420251011398) and performed according to PRISMA guidelines. A comprehensive literature search was performed in MedLine (via PubMed), Scopus, and Web of Science databases (July 15, 2024). Eligible studies included original observational studies explicitly reporting prevalence, incidence, natural history, or risk factors of DME in T1D patients. Studies not differentiating T1D from T2D were excluded. The risk of bias was assessed using NIH criteria. Meta-analysis and meta-regression were used for data synthesis.

Results: Of the 63 studies included, 27 were eligible for meta-analysis, encompassing 40 distinct study populations and a total of 20,074 patients. The pooled prevalence of DME in T1D was 11.1% (95% CI: 7.9-14.3%, I²=99.6%). Meta-regression identified disease duration and HbA1c as significant predictors: at the sample means (19.9 years of diabetes; HbA1c 8.6%), the adjusted prevalence was 14.8%. Each additional year of disease increased prevalence by 1.2 percentage points, and each 1% increase in HbA1c raised prevalence by 4.7 percentage points, on average. Additional factors such as hypertension, dyslipidemia, and nephropathy were also associated with higher DME prevalence. The certainty of the evidence was rated as very low using the GRADE approach.

Conclusions: DME prevalence in T1D is substantial, with a strong dependency on disease duration and glycemic control. Nonetheless, these findings should be interpreted with caution given the very low certainty of the evidence. These findings highlight the need for personalized screening intervals and early intervention strategies. Future research should focus on refining diagnostic criteria, integrating emerging biomarkers, and evaluating the impact of novel diabetes management technologies.