New Statistical Model Captures How Stretching Injures Nerve Fibers

When the brain or spinal cord is subjected to sudden forces, as in a car crash, a fall on the sports field, or a blast wave, the damage often falls not on the neurons themselves but on the delicate wiring that connects them. White matter, the pale tissue composed of bundled nerve fibers called axons, is exquisitely vulnerable to being stretched. When these axons elongate even slightly beyond their tolerance, the injury cascades into lasting neurological dysfunction, from cognitive impairment to motor deficits. Yet despite decades of research, scientists have struggled to build a single mathematical model that can predict exactly how mechanical deformation disrupts the electrical signals that axons carry. A new study published in PLOS Computational Biology by Zexuan Chen, Liqun Tang, Bao Yang, Yiping Liu, Zhenyu Jiang, Licheng Zhou and Zejia Liu now offers a fresh approach, one that treats axonal injury not as a uniform event but as a statistical process unfolding across thousands of individual fibers.

The electrical signaling of neurons has been understood since the pioneering work of Alan Hodgkin and Andrew Huxley in the 1950s through a set of equations describing how sodium and potassium ions flow across the cell membrane through gated channels. When an axon fires, voltage-gated sodium channels snap open, allowing positively charged sodium ions to flood inward and drive the membrane potential upward, while potassium channels follow with a delayed opening that restores the resting state. This choreography of channel gating, captured mathematically by rate constants that depend on voltage, is what allows a nerve impulse to propagate reliably along a fiber at speeds of up to a hundred meters per second. Any injury model that hopes to describe damaged nerve function must therefore grapple with how mechanical stretching perturbs this delicate ionic machinery.

Several research groups have previously proposed axonal electrophysiological damage models built on the Hodgkin-Huxley framework, in which the deformation of the membrane is assumed to alter the behavior of ion channels. These models have provided valuable insights, but as the new study points out, none of them can fully reproduce the results of stretching experiments performed on white matter nerve bundles across different rates and amplitudes of deformation. The mismatch matters because real injuries are not neat, controlled events. A traumatic impact stretches a nerve fascicle rapidly and unevenly, and the electrophysiological consequences depend on both how fast and how far the tissue is deformed. A model that works for one loading condition but fails for another offers limited guidance to clinicians trying to understand or treat diffuse axonal injury.

The team behind the new model identified two crucial ingredients that earlier frameworks had neglected. The first is the coupled damage of sodium and potassium channels. Stretching an axon does not simply disable one type of channel in isolation; it affects both channel families simultaneously, and critically, it alters the rate constants that govern how quickly their gates open and close in response to voltage. By incorporating the combined effects of sodium and potassium channel damage on these gating kinetics, the model captures a subtler picture of how mechanical deformation reshapes the electrical behavior of an injured fiber, rather than merely scaling down its excitability.

The second ingredient is perhaps the more conceptually striking one: the recognition that no two axons in a nerve bundle are alike. Within a single fascicle, individual fibers differ in their mechanical stiffness, their geometry, and their susceptibility to injury. When the bundle is stretched, some axons accumulate damage quickly while others resist longer. Earlier models typically treated the fascicle as a homogeneous population, averaging away this variability. The new framework instead represents the differences among axons continuously using a statistical distribution, so that the response of the whole nerve emerges from the spectrum of individual behaviors rather than from a single representative fiber. This statistical treatment allows the model to reproduce the gradual, heterogeneous loss of function observed experimentally as stretch severity increases.

One of the most telling tests of the model concerns a phenomenon known as the sodium left-shift. After mechanical injury, the voltage dependence of sodium channel activation shifts toward more negative membrane potentials, meaning the channels open at lower voltages than they would in a healthy axon. This left-shift is a hallmark electrophysiological signature of damaged nerve tissue, and reproducing it has been a benchmark for any credible injury model. The new formulation, the authors report, describes the sodium left-shift phenomenon reasonably while also accurately capturing the broader neuro-electrophysiological responses of nerve fascicles under a variety of stretching conditions, spanning different rates and amplitudes of deformation.

The significance of this dual achievement lies in its universality. Previous models often required case-by-case tuning to match a particular experiment, effectively fitting the damage parameters to each loading scenario. A model that can span the experimental range with a single consistent formulation is far more valuable, because it suggests the underlying mathematics is capturing genuine mechanisms rather than curve-fitting artifacts. For researchers studying traumatic brain injury and spinal cord injury, such a model provides a computational bridge between the biomechanics of an impact and the functional consequences for neural signaling, allowing simulations of injury scenarios that would be impossible or unethical to reproduce in the laboratory.

The clinical implications could be substantial. Diffuse axonal injury is one of the most devastating consequences of traumatic brain injury, contributing to prolonged coma and long-term disability, yet it is notoriously difficult to assess with imaging alone because the damage occurs at a scale below the resolution of conventional scans. A mechanics-based electrophysiological model offers a way to translate measurable mechanical inputs, such as the strains and strain rates estimated from computational reconstructions of an accident, into predictions of functional impairment. The authors emphasize that their model provides a mechanical mechanism-based reference for future clinical treatment strategies for nerve injury, potentially informing decisions about protective equipment design, surgical intervention, and rehabilitation timing.

The work also highlights a broader trend in computational biology: the move toward models that embrace heterogeneity rather than smoothing it away. Biological tissue is inherently variable, from the molecular noise of channel gating to the anatomical diversity of fiber diameters and myelination within a nerve. Models that encode this variability statistically, as the new axonal damage model does, tend to be more robust when confronted with data they were not explicitly tuned to fit. In this sense, the study contributes not only a tool for neurotrauma research but also a methodological example for modeling other coupled mechano-physiological systems, where mechanical forces and biological function are tightly intertwined.

Much remains to be done before such models reach the clinic. The framework will need validation against additional experimental datasets, extension to three-dimensional tissue-level simulations, and integration with imaging techniques that can estimate patient-specific strain fields. Nevertheless, by coupling sodium and potassium channel damage to gating kinetics and by treating axon-to-axon variability as a continuous statistical distribution, Chen and colleagues have produced what may become a reference point for the field. Their model demonstrates that the electrical silence of a stretched nerve is not a simple on-off failure but a graded, statistically structured degradation of the ionic machinery, and that capturing this structure is the key to predicting, and perhaps one day preventing, the hidden damage of traumatic nerve injury.

Subject of Research: A statistical damage model for coupled mechano-electrophysiological axonal injury

Article Title: A statistical damage model for coupled mechano-electrophysiological axonal injury

Article References: Chen, Z., Tang, L., Yang, B., Liu, Y., Jiang, Z., Zhou, L., & Liu, Z. (2026). A statistical damage model for coupled mechano-electrophysiological axonal injury. PLOS Computational Biology, 22(10), e1014837. https://doi.org/10.1371/journal.pcbi.1014837

Image Credits: AI Generated

DOI: 10.1371/journal.pcbi.1014837

Keywords: statistical, damage, model, coupled, mechano-electrophysiological, axonal, injury, scientific research