Introduction:
Lipid metabolism is pivotal in diabetic retinopathy (DR) development. Nevertheless, the relationship between lipid-lowering drugs and the risk of DR remains controversial. This study utilized Mendelian randomization (MR) to investigate the potential effects of pharmacological targets for lowering lipid levels on DR and to clarify the causal link between blood lipid characteristics and DR.

Material and methods:
The data comprised genetic variations related to lipid traits and genetic variations associated with lipid-lowering drug targets obtained from the Global Lipid Consortium. Total DR, non-proliferative DR (NPDR), and proliferative DR (PDR) were sourced from the Finnish R9 database. Lipid-lowering drug targets were tested using inverse variance-weighted MR (IVW-MR) and statistics-based MR (SMR). Colocalization and mediation analysis were conducted to validate the results and explore potential mediating factors.

Results:
Results: A reduced risk of total DR and NPDR was linked to genetically improved 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR) (OR = 0.62; 95% CI: 0.46–0.83; p = 1.30× 10^-2; OR = 0.49; 95% CI :0.34–0.70; p = 9.70× 10^-4). Strong colocalization (PP.H4 = 0.85) was observed between whole blood tissue HMGCR expression and a significant MR relationship with total DR (OR = 0.66; ‌95% CI: 0.52–0.85; p = 7.31× 10^-4). Furthermore, Body mass index (BMI)and glycated hemoglobin (HbA1c) are critical factors that mediate the impact of HMGCR and APOB on DR risk.

Conclusions:
This Mendelian randomization study suggests that abnormalities in triglyceride (TG) levels serve as a pathogenic element in DR. Of the nine lipid-lowering drug targets assessed, HMGCR and APOB have emerged as potential promising targets for managing NPDR.