Brain Tissue Remembers: Prior Loading Softens Its Response to Repeated Shear

Every year, millions of people sustain traumatic brain injuries, and for a substantial subset of them the trauma does not stop at a single event. Athletes in contact sports, military personnel exposed to blasts, and victims of repeated falls or vehicle collisions can experience multiple head impacts within days, weeks, or years of one another. Yet the computational models that engineers and clinicians use to predict how the brain deforms during such events are largely built on a simplifying assumption: that brain tissue responds to each load as if it were the first. A new experimental study published in Medical & Biological Engineering & Computing challenges that assumption directly, showing that the mechanical behavior of brain tissue depends not just on how much it is loaded, but on the precise sequence and type of loading it has already experienced.

The research, conducted by Hadi Nosrati and Mehdi Shafieian of Amirkabir University of Technology in Tehran together with Kurosh Darvish of Temple University in Philadelphia, set out to answer a deceptively simple question: does the loading history of brain tissue change its subsequent mechanical response? To find out, the team performed multistage simple shear experiments on fourteen bovine brain specimens. Bovine tissue is a widely used surrogate in brain biomechanics because it is readily available, structurally similar in many respects to human brain matter, and can be obtained as waste material from slaughterhouses without any animals being euthanized specifically for research purposes.

The experimental design was deliberately more sophisticated than the single-stage loading protocols that dominate the literature. Each specimen was subjected to successive loading stages combining ramp-and-hold tests, in which the tissue is sheared to a target strain and held there while stress relaxes, and cyclic loading, in which the tissue is sheared back and forth repeatedly. By varying the order of these protocols, the strain amplitudes, and the loading modes, the researchers could probe whether the tissue’s response depended on the path it had taken through its own mechanical history. All tests were conducted within a reported sub-damaging loading range, meaning the deformations were gentle enough to avoid gross structural failure while still revealing how the material evolves under repeated loading.

To quantify the mechanical behavior, the team fitted their data to a viscoelastic–hyperelastic constitutive framework, extracting a shear modulus and a set of Prony-series parameters that describe how the tissue’s stress relaxes over time. In such models, the hyperelastic component captures the instantaneous, rubber-like stiffness of the tissue, while the viscoelastic terms, here denoted g1 through g3 and g-infinity, describe the time-dependent dissipation and relaxation that arise from the brain’s complex microstructure of neurons, glial cells, axons, and interstitial fluid. These parameters are precisely the numbers that feed into the finite element head models used in crash safety research, sports equipment design, and injury prediction, which makes any systematic change in them potentially consequential for how we simulate real-world head impacts.

The headline finding is that prior cyclic loading generally reduced the peak stress and the shear modulus of the tissue in subsequent loading stages. From the initial to the final loading stage, the shear modulus decreased from 2.94 plus or minus 0.94 kilopascals to 2.49 plus or minus 0.63 kilopascals, a statistically significant softening. In other words, after experiencing a history of cyclic deformation, the brain tissue became measurably less stiff, even though the individual loading episodes remained within the sub-damaging range. This is a subtle but important result: it suggests that the tissue accumulates a form of mechanical memory even at loads that would not normally be classified as injurious.

The viscoelastic parameters told a more nuanced story. The parameter g2 increased from 0.14 plus or minus 0.19 to 0.19 plus or minus 0.17, while g3 decreased from 0.21 plus or minus 0.08 to 0.16 plus or minus 0.06, both shifts reaching statistical significance. Meanwhile, g1 and the long-term equilibrium modulus g-infinity showed no significant changes. This pattern implies that the softening is not a uniform degradation of all time-dependent mechanisms. Instead, specific relaxation timescales are being redistributed, indicating that repeated loading alters the balance between the fast and slow dissipative processes within the tissue without fundamentally changing its long-term equilibrium stiffness. For modelers, this means that a single set of viscoelastic parameters measured on pristine tissue may misrepresent the relaxation behavior of tissue that has already been cycled.

Perhaps the most intriguing result concerns what did not correlate. The researchers tested whether the cumulative strain experienced by each specimen predicted its final shear modulus, and found no significant relationship, with a Spearman correlation coefficient of minus 0.31 and a p-value above 0.05. What did matter was the sequence of the loading protocols: different orderings of ramp-and-hold and cyclic stages produced different final moduli. This is the hallmark of path dependence, a property familiar from complex materials science but rarely demonstrated so explicitly in brain tissue. The tissue does not simply tally up the total deformation it has endured; it responds to the specific choreography of how that deformation was delivered.

The implications for repetitive traumatic brain injury research are considerable. Epidemiological studies have long shown that individuals with a prior brain injury face elevated risks of re-injury, dementia, and death, and animal models have demonstrated that traumatic axonal injury can be exacerbated by repeated closed head impacts. Computational models attempting to capture these phenomena have been hampered by a lack of experimental data on how prior loading alters tissue mechanics. The new findings suggest that constitutive models for repeated loading scenarios should incorporate loading-history dependence, rather than treating each impact as an independent event acting on virgin material. A brain that has already been rattled, even sub-clinically, may deform differently under the next blow than the models currently assume.

The study also speaks to a persistent methodological problem in brain biomechanics. The field has long struggled with inconsistencies between ex vivo mechanical measurements, partly because different laboratories use different preconditioning protocols, the preliminary loading cycles applied to a specimen before formal data collection. The authors’ own prior work has catalogued these inconsistencies in detail. The present results suggest that preconditioning is not a mere technical nuisance to be standardized away; it is itself a manipulation of the tissue’s mechanical state that can shift the measured parameters. Two laboratories testing ostensibly identical tissue with different preconditioning histories may legitimately obtain different answers, because the tissue itself has been changed by its history.

Certain caveats frame the scope of these findings. The experiments used bovine tissue under ex vivo conditions, and the reported loading range was sub-damaging, so the results speak most directly to modeling of repeated mild loading rather than severe injury. The authors note that the findings may inform constitutive models for repeated loading scenarios, including those relevant to repetitive traumatic brain injury, which is a careful and appropriately measured claim. Translating the results to living human brains, with their differences in scale, vasculature, and age-related variation, will require further work. Nevertheless, by demonstrating that brain tissue mechanics are path- and sequence-dependent within a controlled experimental framework, the study provides both a caution and an opportunity: a caution against relying on history-blind material models, and an opportunity to build the next generation of head injury simulations on a more faithful description of how soft neural tissue actually behaves when it is loaded again and again.

Subject of Research: Path-dependent viscoelastic and hyperelastic mechanical response of brain tissue under repeated multistage shear loading

Article Title: Path-dependent viscoelastic–hyperelastic response of brain tissue under multistage shear loading

Article References: Nosrati, H., Shafieian, M., & Darvish, K. (2026). Path-dependent viscoelastic–hyperelastic response of brain tissue under multistage shear loading. Medical & Biological Engineering & Computing. https://doi.org/10.1007/s11517-026-03692-z

Image Credits: AI Generated

DOI: 10.1007/s11517-026-03692-z

Keywords: brain tissue mechanics, traumatic brain injury, viscoelasticity, hyperelasticity, shear loading, loading path dependence, repetitive head impact, constitutive modeling, biomechanics, Prony series, ex vivo testing, soft tissue rheology