Causal Association Between Commonly Used Medicines and Diabetes-Related Eye Diseases: Univariable and Multivariable Mendelian Randomization Study.
Diabetic patients require long-term polypharmacy, yet the causal effects of these medications on ocular complications-specifically diabetic retinopathy (DR), diabetic maculopathy (DMac), and glaucoma-remain unclear due to confounding in observational studies.
We performed a Mendelian randomization (MR) study using GWAS summary statistics for 23 common medication classes and five diabetic eye diseases. Multivariable MR (MVMR) was employed to adjust for key comorbidities (e.g., hypertension, type 2 diabetes), while enrichment analyses explored biological mechanisms.
In the univariable MR (UVMR) analysis, drugs used in diabetes, agents acting on the renin-angiotensin system, and several other medication classes showed significant associations with increased risks of diabetic eye diseases. Crucially, MVMR confirmed robust causal links after adjusting for key comorbidities. Specifically, drugs used in diabetes remained associated with DMac (OR = 1.44, p = 1.31 × 10-11), DR (OR = 1.23, p = 1.26 × 10-8), neovascular glaucoma (OR = 1.19, p = 0.003), and senile cataract (OR = 1.10, p = 9.25 × 10-12) independent of type 2 diabetes liability. Similarly, thyroid preparations retained significance for DMac (OR = 1.26, p = 4.35 × 10-10), DR (OR = 1.17, p = 9.53 × 10-10), and senile cataract (OR = 1.04, p = 0.004) after adjusting for hypothyroidism. Additionally, adrenergic inhalants were independently linked to senile cataract (OR = 1.07, p = 0.008) after adjusting for asthma. Pathway analysis highlighted hormone transport and MAPK signaling as potential mechanisms.
Our findings provide genetic evidence supporting potential comorbidity-independent causal links between specific systemic medications-particularly diabetes and thyroid drugs-and ocular complications, suggesting the importance of ophthalmological monitoring in these patients.
We performed a Mendelian randomization (MR) study using GWAS summary statistics for 23 common medication classes and five diabetic eye diseases. Multivariable MR (MVMR) was employed to adjust for key comorbidities (e.g., hypertension, type 2 diabetes), while enrichment analyses explored biological mechanisms.
In the univariable MR (UVMR) analysis, drugs used in diabetes, agents acting on the renin-angiotensin system, and several other medication classes showed significant associations with increased risks of diabetic eye diseases. Crucially, MVMR confirmed robust causal links after adjusting for key comorbidities. Specifically, drugs used in diabetes remained associated with DMac (OR = 1.44, p = 1.31 × 10-11), DR (OR = 1.23, p = 1.26 × 10-8), neovascular glaucoma (OR = 1.19, p = 0.003), and senile cataract (OR = 1.10, p = 9.25 × 10-12) independent of type 2 diabetes liability. Similarly, thyroid preparations retained significance for DMac (OR = 1.26, p = 4.35 × 10-10), DR (OR = 1.17, p = 9.53 × 10-10), and senile cataract (OR = 1.04, p = 0.004) after adjusting for hypothyroidism. Additionally, adrenergic inhalants were independently linked to senile cataract (OR = 1.07, p = 0.008) after adjusting for asthma. Pathway analysis highlighted hormone transport and MAPK signaling as potential mechanisms.
Our findings provide genetic evidence supporting potential comorbidity-independent causal links between specific systemic medications-particularly diabetes and thyroid drugs-and ocular complications, suggesting the importance of ophthalmological monitoring in these patients.