Development and validation of a risk prediction model for early diabetic peripheral neuropathy based on a systematic...

Early identification and intervention of diabetic peripheral neuropathy is beneficial to improve clinical outcome.

To establish a risk prediction model for diabetic peripheral neuropathy (DPN) in patients with type 2 diabetes mellitus (T2DM).

The derivation cohort was from a meta-analysis. Risk factors and the corresponding risk ratio (RR) were extracted. Only risk factors with statistical significance were included in the model and were scored by their weightings. An external cohort were used to validate this model. The outcome was the occurrence of DPN.

A total of 95,604 patients with T2DM from 18 cohorts were included. Age, smoking, body mass index, duration of diabetes, hemoglobin A1c, low HDL-c, high triglyceride, hypertension, diabetic retinopathy, diabetic kidney disease, and cardiovascular disease were enrolled in the final model. The highest score was 52.0. The median follow-up of validation cohort was 4.29 years. The optimal cut-off point was 17.0, with a sensitivity of 0.846 and a specificity of 0.668, respectively. According to the total scores, patients from the validation cohort were divided into low-, moderate-, high- and very high-risk groups. The risk of developing DPN was significantly increased in moderate- (RR 3.3, 95% CI 1.5-7.2, P = 0.020), high- (RR 15.5, 95% CI 7.6-31.6, P < 0.001), and very high-risk groups (RR 45.0, 95% CI 20.5-98.8, P < 0.001) compared with the low-risk group.

A risk prediction model for DPN including 11 common clinical indicators were established. It is a simple and reliable tool for early prevention and intervention of DPN in patients with T2DM.