Abstract
Diabetic polyneuropathy (DPN) shares overlapping clinical and electrodiagnostic features with chronic inflammatory demyelinating polyneuropathy (CIDP), which complicates the differential diagnosis of CIDP in diabetic patients. 32 patients with diabetes mellitus and CIDP, 68 patients with CIDP without diabetes, 83 patients with DPN, and 28 diabetic patients without polyneuropathy were examined using clinical scores (Overall Neuropathy Limitation Scale (ONLS), Neuropathy Symptom Score, Neuropathy Deficit Score), nerve conduction studies, and nerve ultrasound (Ultrasound Pattern Sum Score (UPSS)). The ONLS was significantly higher in the CIDP patients with diabetes than in DPN (median [interquartile range]: 4.0 [3.0] vs. 0 [1.0], p < 0.001) as well as the UPSS (4.0 [6.0] vs. 0 [2.9], p < 0.001). Multiple binary logistic regression revealed UPSS and ONLS as statistically significant predictors to differentiate between CIDP with diabetes and DPN. Receiver operating characteristic curve analysis showed the ONLS with an area under the curve (AUC) of 0.918 (95% CI: 0.868-0.0.967, p < 0.001). The UPSS total score had an AUC of 0.826 (95% CI: 0.743-0.909, p < 0.001). An UPSS ≥ 2.5 had a sensitivity of 77.4% and a specificity of 68.7% to detect CIDP. An ONLS ≥ 1.5 had a sensitivity of 87.1% and a specificity of 81.9% to detect CIDP. ROC curve analysis of a composite score of ONLS and UPSS revealed an AUC of 0.959 (95% CI: 0.928-0.991, p < 0.001). CIDP is an important differential diagnosis in people with diabetes mellitus. This study reports that the UPSS is well suited to differentiate between DPN and CIDP.