Comparing the Diagnostic Characteristics of Neuropathy Questionnaires in Adults With Obesity

Aim To compare the diagnostic characteristics of three peripheral neuropathy (PN) patient‐completed questionnaires. Methods We recruited participants with obesity from a previously completed randomized controlled trial. Patient‐completed PN questionnaires included the Michigan Neuropathy Screening Instrument questionnaire (MNSIq), the Diabetic Neuropathy Symptom (DNS) score, and the Erasmus Polyneuropathy Symptom Score (E‐PSS). PN was defined using the Toronto consensus definition for confirmed neuropathy. Secondary PN definitions included probable neuropathy and definitions based on skin biopsy, nerve conduction studies, and clinical scales. Diagnostic accuracy was determined using the area under the receiver operating characteristic curve (AUC), and significant differences in AUC were assessed with DeLong's test. Potential diagnostic thresholds and corresponding predictive values and likelihood ratios were calculated. Results We enrolled 59 participants (mean [SD] age: 53 [7] years, 76% female, 81% White), and 20.3% had neuropathy. E‐PSS had the largest AUC (95% confidence interval) for confirmed neuropathy (0.84, 0.71–0.97), followed by MNSIq (0.76, 0.63–0.90) and the DNS (0.72, 0.54–0.90), but these differences were not significant ( p > 0.05). No significant differences were seen for secondary PN definitions (all p > 0.05). All three questionnaires were highly correlated with one another ( r range: 0.70–0.72, all p < 0.01), and diagnostic thresholds were estimated. Interpretation All three patient‐completed questionnaires displayed good diagnostic accuracy for PN and were highly correlated. Our findings suggest that questionnaires can be used to determine PN in large epidemiological studies and might be useful as screening instruments in clinical care.