[The problems that should be noticed in choosing animal models with diabetic retinopathy]

Animal models of diabetes mellitus (DM) are vital for research on pathogenesis and pharmaceutical therapy of diabetic retinopathy (DR). At present, diabetic animal models include chemicals or dietary-induced models, transgenic or gene knockout mice, and spontaneous mice, etc. Among these models, rat model induced by streptozotocin is simple and imminent with high repeatability, which made it the most popular one. However, no available diabetic model could reproduce all vascular and neural pathological changes observed in human non-proliferative and proliferative DR. It is also known that not all DM models could lead to DR. In addition, features and severity of retinopathy in the model depend on animal kinds, animal lifespan, time course of the disease and induction methods. Therefore, researchers should consider characteristics and limitations of different models while choosing suitable DM model based on research objectives and resources. It is particularly emphasized that the results from animal models do not always fit human conditions.