Electrodiagnostic Classification of Sensory Neuropathies: Severity Clusters and Clinical Correlates

INTRODUCTION/AIMS: Grading sensory neuropathy severity on nerve conduction studies (NCS) is operator dependent, which limits comparisons across patients, laboratories, and studies. We aimed to develop an objective severity classification to standardize NCS reporting and support clinical decision-making. METHODS: We retrospectively analyzed NCS performed at IRCCS San Raffaele Hospital (2018-2024), extracted with a validated parser. Adult studies with sensory involvement were selected using sural sensory nerve action potential (SNAP) amplitude thresholds after excluding motor abnormalities. Hierarchical clustering combined normalized sural and "best" upper-limb (UL) SNAP amplitudes with binary indicators for absent responses. Classes were ordered by an NCS severity index and validated against clinical scales. RESULTS: Among 15,744 consecutive examinations, 308 met inclusion criteria. Three well-separated NCS severity classes emerged: mild (reduced but present sural and present UL SNAPs), moderate (absent sural, present UL SNAPs), and severe (absent sural and UL SNAPs). Clinical severity increased monotonically across classes. Patients in the severe class had younger age at symptom onset than those in mild-moderate classes and were more disabled (modified Rankin scale ≥ 2: 50.0% in severe vs 30.4% in moderate vs 16.2% in mild). The severe class was enriched in immune/paraneoplastic (12.5%) and genetic (40.6%) causes. DISCUSSION: By stratifying patients based on SNAP amplitudes, we identified a differential etiologic enrichment rather than a continuum of disease progression. Severe electrophysiology with motor sparing should prompt RFC1 testing and immune/paraneoplastic work-up, whereas mild to moderate abnormalities should lead to assessment for diabetes and toxic or nutritional exposures.