A Bottleneck Coupling Principle in Early Amyotrophic Lateral Sclerosis Electrophysiology From motor-unit instability and output loss to open-data multiscale triangulation

Early amyotrophic lateral sclerosis (ALS) is difficult to distinguish from neurological mimics becausefasciculation, active denervation, and reduced compound muscle action potential (CMAP) are individually nonspecific. We asked whether these abnormalities are more informative when represented as a coupledphysiological state. In a clinical discovery cohort, 29 patients with ALS within 6 months of symptom onset werecompared with 27 neurological mimics using conventional electrophysiological variables. We defined a motorunit Instability axis from fasciculation potentials, fibrillation and positive sharp waves, and an Output-Loss axisfrom CMAP. Among difficult CEILS-positive mimics, the minimum of these axes, B=min(Instability, Output Loss),was the best of 12 pre-specified coupling operators (AUC 0.812; P=0.00535; FDR q=0.036) and outperformedeither axis alone in leave-one-out analysis. A medically constrained nested model family achieved an aggregateout-of-fold AUC of 0.935 (bootstrap 95% CI 0.858-0.990), sensitivity 82.8% and specificity 96.3%, whilerepeatedly selecting the same instability-output structure. We then triangulated the model using open data.Longitudinal human CMAP data showed mean decline from baseline of 10.05% at 3 months and 36.87% at 12months across four muscles. An independent SPiQE/AVID dataset contained 165 ALS recordings with medianfasciculation frequency 40.5/min and strong bilateral concordance (Spearman rho=0.723, P=2.46x10^-14).Finally, open SOD1G37R neuromuscular-junction data showed a demand-dependent intervention effect onsoleus output: little separation at 5-20 Hz but increasing force preservation at high stimulation frequencies, withthe largest mean difference at 120 Hz (+29.84 mN). These results support a candidate framework in which earlyALS is represented not by a single abnormality or total electrical burden, but by concurrent motor-unit instabilityand output loss, with the weaker process acting as a bottleneck. The study is exploratory and requirespreregistered patient-level external validation of the locked formula.