Context: Due to the heterogenous clinical symptoms and deficits, the diagnosis of diabetic polyneuropathy (DPN) is still difficult in clinical routines, leading to increased morbidity and mortality.

Objective: We studied the correlation of phase angle (PhA) of bioelectrical impedance analysis (BIA) with clinical, laboratory, and physical markers of DPN to evaluate PhA as a possible diagnostic method for DPN.

Materials and methods: In this cross-sectional observational study as part of the Heidelberg Study on Diabetes and Complications, we examined 104 healthy individuals and 205 patients with type 2 diabetes mellitus (T2D), among which 63 had DPN. The PhA was calculated from multifrequency BIA. Nerve conduction studies, quantitative sensory testing (QST) and diffusion-weighted magnetic resonance neurography to determine fractional anisotropy (FA) reflecting peripheral nerve integrity were performed.

Results: T2D patients with DPN had lower PhA values (5.71 ± 0.10) compared to T2D patients without DPN (6.07 ± 0.08, P = .007, + 6.1%) and healthy controls (6.18 ± 0.08, P < .001, + 7.9%). Confounder-adjusted analyses showed correlations of the PhA with conduction velocities and amplitudes of the peroneal (β=.28; β=.31, P < .001) and tibial nerves (β=.28; β=.32, P < .001), Z-scores of QST (thermal detection β=.30, P < .05) and the FA (β=.60, P < .001). Receiver-operating characteristic analysis showed similar performance of PhA in comparison to the mentioned diagnostic methods.

Conclusion: The study shows that PhA is, in comparison to other test systems used, at least an equally good and much easier to handle investigator-independent marker for detection of DPN.