Neurophysiological index as a biomarker for ALS progression: validity of mixed effects models

Our objective was to evaluate the neurophysiological index (NI) as a biomarker for amyotrophic lateral sclerosis (ALS) and to assess the validity of linear mixed effects models for describing longitudinal changes. Functional assessment and nerve conduction studies were undertaken in 58 ALS patients. Neurophysiological data were collected on four occasions over 12 weeks (baseline, weeks 4, 8 and 12). The NI was calculated for the abductor digiti minimi and ulnar nerve at the wrist. NI declined at a rate of 0.04 per week (S.E. 0.006, p < 0.0001). Patients with bulbar-onset disease had 0.88 greater NI than patients with upper limb-onset disease over the follow-up period (S.E. 0.39, p = 0.03). There were no differences in the rates of decline among patients with different disease phenotypes. Rates of change in NI and functional impairment were weakly correlated (Spearman's p = 0.29, p = 0.03). Linear mixed effects models were appropriate for detailing the longitudinal changes in NI. The present findings support incorporation of NI as an outcome measure for ALS clinical trials conducted over short time periods.