Develop an AI model that estimates biological age from routine nerve conduction studies and determine whether positive age gaps identify subclinical peripheral neuropathic change that current cutoffs miss.
Chronological age can obscure early peripheral nerve dysfunction. Although nerve conduction parameters vary with aging, there is no electrophysiology based biological age tool in clinical use to surface covert abnormalities.
Supervised regressors (gradient boosted and random forests) were trained on handcrafted features from motor, sensory, and F wave recordings (27,730; 9,863; 6,637 signals). Within normal-labeled studies, the upper tail of the age-gap distribution defined suspected subclinical cases per modality. Clinical validity was assessed by prespecified tests: dose–response across clean normal, subclinical, and clinically abnormal groups; age-matched comparisons; cross-modal consistency; and nerve-specific analyses.
Large positive age gaps in young adults identified a reproducible subclinical cohort with intermediate physiology. Pooled Dose–response checks showed monotonic worsening of motor metrics from normal to subclinical to clinically abnormal groups. Motor amplitude showed a large effect (p<0.001). Pooled motor conduction velocity declined across the same gradient (normal 42.7, subclinical 38.8, clinically abnormal 37.7 m/s; p<0.001). At the nerve level, subclinical cases showed significant abnormalities in 10 of 13 motor nerves, 5 of 5 sensory nerves, and 4 of 4 F wave nerves examined, typically lower amplitudes and velocities with modest latency increases (p<0.001). Sensory amplitudes were reduced about 48.8% versus controls and median nerve F wave latency was higher about 28.8% (both p<0.001). Cross modal ages were tightly aligned (±1.8 years), supporting a coherent phenotype.
An AI derived biological age from standard nerve conduction signals flags young patients with subclinical peripheral neuropathic change that exceeds contemporary normal ranges yet falls short of overt abnormality. The age gap threshold functions as a practical signal for early detection and risk stratification in neuromuscular care.