Diagnostic contribution of multi-frequency vibrometry to detection of peripheral neuropathy in type 1 diabetes mellitus compared with nerve conduction studies

Abstract

Aim The aim was to assess the use of multi-frequency vibrometry (MFV) in detecting diabetic peripheral neuropathy (DPN) in type 1 diabetes in comparison to nerve conduction studies (NCS) and neurothesiometer (NT). Our objectives were to examine how VPTs correlated with NCS parameters, evaluate the efficacy of MFV in distinguishing DPN as well as to investigate whether MFV procedure could be based on fewer frequencies. Methods Adults with type 1 diabetes with previous MFV examinations were recruited at Skåne University Hospital in Malmö, Sweden, between 2018 and 2020. Participants were examined regarding nerve function in the lower limbs through MFV, NT and NCS. Results A total of 66 participants (28 women and 38 men) with a median age of 50 (39 to 64) years were included in the study. Through NCS assessment, 33 participants (50%) were diagnosed with DPN. We found negative correlations between VPTs and all NCS parameters, where the strongest correlation was found between sural nerve amplitude and the 125 Hz frequency of MFV. A combination of four frequencies, two low (4 and 8 Hz) and two high (125 and 250 Hz), showed the highest classification efficacy (AUC 0.83, 95% CI 0.73–0.93). Conclusion We conclude that a strong correlation exists between the sural nerve amplitude and the VPTs at 125 Hz and that VPT testing with MFV can be focused on only four frequencies instead of seven, thus shortening test time, to distinguish DPN in the lower limb.