Prediction of diabetic peripheral neuropathy in type 2 diabetes using high-frequency ultrasound and shear-wave elastography of the median and tibial nerves: a nomogram study.

BACKGROUND

Diabetic peripheral neuropathy (DPN) is a common and disabling complication of type 2 diabetes mellitus (T2DM).Noninvasive imaging such as high-frequency ultrasound (HFUS) and shear wave elastography (SWE) can capture structural and mechanical changes in peripheral nerves. This study sought to identify predictors for DPN and develop a clinically useful nomogram based on HFUS and SWE parameters of the median nerve and tibial nerve plus routine clinical and laboratory indicators.

METHODS

A total of 118 adults with T2DM classified as DPN or non-DPN (NDPN) were retrospectively enrolled and randomized into the training dataset and testing dataset in a 7:3 ratio. HFUS and SWE were utilized to measure the cross-sectional area (CSA) of the median nerve and tibial nerve, the shear wave velocity (SWV) of the same nerve. These factors, plus clinical variables, were compared between DPN and NDPN. Univariate analysis and multivariate logistic regression analysis were conducted on the training dataset to identify independent predictors for DPN, construct a binary logistic regression model, and develop a corresponding nomogram. Its performance was evaluated using the receiver operating characteristic (ROC) analysis, decision curve analysis (DCA), and clinical impact curve (CIC).

RESULTS

Significant differences were detected in age, course of T2DM, postprandial 2-h C-peptide (2-h C-P), CSA of the median nerve and tibial nerve, and SWV of the tibial nerve between the DPN and NDPN (P < 0.05). Multivariate logistic regression analysis identified the course of T2DM, 2-h C-P, TN-CSA1, TN-CSA3, TN-SWV2, and TN-SWV3 as independent predictors for DPN. Using them, the nomogram demonstrated an AUC of 0.867 in the training dataset and 0.824 in the testing dataset. DCA plotted a net benefit within the risk threshold of 0.15-0.85. CIC validated the nomogram's high predictive accuracy at the risk threshold of 0.4 and above.

CONCLUSION

Course of T2DM, 2-h C-P, TN-CSA1, TN-CSA3, TN-SWV2, and TN-SWV3 are independent predictors for DPN. The DPN nomogram based on HFUS and SWE demonstrated a favorable diagnostic performance and may assist in clinical decision-making.