Blood Tau Rises in ALS but Tracks Motor Neuron Damage, Not Thinking Skills

A protein that has become one of the most celebrated blood tests in Alzheimer’s disease research is now forcing neuroscientists to rethink what it means when it appears in amyotrophic lateral sclerosis, a very different and far more lethal disorder of the nervous system. Phosphorylated tau 181, or p-tau181, earned regulatory approval as a blood-based diagnostic marker for Alzheimer’s disease because it reliably reflects the tangled tau pathology that accumulates inside the brains of people with that condition. Yet in ALS, a disease that destroys the motor neurons controlling muscles, several research groups have reported that blood levels of the very same protein are markedly elevated, sometimes reaching concentrations comparable to those seen in Alzheimer’s patients. A new multicenter study from the German Center for Neurodegenerative Diseases, published in Annals of Clinical and Translational Neurology, set out to determine whether this blood signal in ALS is a fingerprint of Alzheimer’s co-pathology in the brain, a mirror of cognitive decline, or something else entirely. The answer, drawn from more than two hundred patients followed over a full year, is perhaps the most unexpected possibility: the protein appears to originate largely outside the brain, and it rises relentlessly as the disease runs its course.

The research team analyzed 202 people living with ALS, including some with comorbid frontotemporal dementia, alongside 94 healthy controls drawn from two multicenter cohorts. Every participant underwent standardized clinical assessment using the revised ALS Functional Rating Scale, diagnosis was confirmed according to the revised El Escorial Criteria, and disease severity was staged using the King’s staging system. Cognitive function was assessed with the German version of the Edinburgh Cognitive and Behavioural ALS Screen, a battery of fifteen subtests spanning five domains: verbal fluency, executive function including social cognition, language, memory, and visuospatial ability. Blood plasma concentrations of p-tau181 were measured using the ultrasensitive Quanterix SIMOA platform, with every sample run in technical duplicates and internal control aliquots monitoring consistency between runs. Neurofilament light chain, the established workhorse biomarker of axonal injury in ALS, was quantified in parallel, and a subset of patients also underwent cerebrospinal fluid analysis.

The cross-sectional results were striking in their asymmetry. At baseline, people with ALS showed significantly higher plasma concentrations of both p-tau181 and neurofilament light chain than healthy controls. Mean plasma p-tau181 in patients stood at 3.9 picograms per milliliter, more than double the control average of 1.7, while mean NfL soared to 73.2 picograms per milliliter against a control value of just 14.4. Yet when the researchers turned to cerebrospinal fluid, the picture changed dramatically: no significant difference in CSF p-tau181 emerged between patients and controls. This dissociation matters enormously. If the elevated blood signal were simply spilling over from tau pathology inside the brain, the fluid bathing that brain should carry a matching elevation. Its absence argues that the plasma p-tau181 in ALS is generated predominantly by peripheral processes rather than by cerebral disease.

The cognitive findings reinforced that conclusion. Cognitive impairment affects up to half of all ALS patients, typically eroding verbal fluency, executive function, language, and social cognition, with memory deficits sometimes appearing as well. If plasma p-tau181 were a marker of Alzheimer’s co-pathology, it should track those memory and visuospatial deficits. It did not. Biomarker levels at baseline did not differ significantly among cognitively unimpaired patients, cognitively impaired patients, and those with full frontotemporal dementia. Partial correlations between cognitive test scores and biomarker levels, controlled for age and corrected for multiple comparisons, reached statistical significance for no domain and no biomarker. The correlation between plasma p-tau181 and verbal fluency, for instance, was essentially zero. Even numerically, the strongest tendencies ran through CSF p-tau181 rather than its plasma counterpart, hinting that central tau relates to cognition in a way plasma tau simply does not.

Genetic analysis deepened the puzzle. Carriers of the APOE ε4 allele, the best-known genetic risk factor for late-onset Alzheimer’s disease, showed the expected elevations in cerebrospinal fluid Alzheimer’s markers, including altered amyloid beta ratios and higher p-tau181 in CSF, confirming that the assays were capturing genuine central pathology. But their plasma p-tau181 levels were indistinguishable from those of non-carriers. The same held true when patients were grouped by ATN status, the standard framework combining amyloid, tau, and neurodegeneration markers: patients with high-risk profiles showed clear CSF differences but no plasma tau differences whatsoever. ALS-associated gene mutations such as FUS and SOD1, meanwhile, had no measurable effect on any biomarker. Whatever is driving the blood tau signal in ALS, it is neither Alzheimer’s genetics nor ALS genetics.

Longitudinal monitoring over approximately one year revealed a dynamic that distinguishes the two blood proteins even more sharply. Plasma p-tau181 increased significantly over time in patients while remaining flat in controls, and a significant group-by-time interaction confirmed that disease status, not aging, was the engine of change. Neurofilament light chain, by contrast, remained statistically stable across the same interval, despite its much higher baseline concentrations. This divergence suggests the two markers capture fundamentally different biological processes, one rising steadily with ongoing disease activity and the other plateauing once a certain threshold of axonal damage has been reached.

Where the proteins parted ways most cleanly was in their relationship to clinical phenotype and prognosis. Neurofilament light chain behaved exactly as its reputation predicts: higher levels were associated with more advanced disease severity, faster progression rates, and shorter disease duration, and it clearly separated slow from fast progressors and discriminated across King’s disease stages. Plasma p-tau181, in contrast, showed no relationship with severity, progression, staging, or disease duration, and it failed to distinguish slow from fast progressors entirely. What it did do, and did clearly, was rise in patients with predominantly lower motor neuron involvement, the muscle-wasting form of the disease, reaching significantly higher concentrations than in patients whose upper motor neurons dominated the clinical picture. Combined with a positive correlation between plasma p-tau181 and NfL, but no correlation between plasma and CSF p-tau181, the data sketch a protein that accompanies peripheral neurodegenerative processes without measuring how sick a patient is or how fast they are declining.

The authors are careful to situate these results against a complicated literature. Previous studies have reported elevated serum p-tau181 in ALS with concentrations comparable to Alzheimer’s disease, and one recent investigation using the novel NULISA assay even reported robust associations between plasma p-tau181, clinical severity scores, and survival, with higher concentrations predicting poorer outcomes. That study, however, was a single-center pilot with a limited sample size and a different assay technology, factors that may explain the discrepancy. The present findings, converging with several independent cohorts across two large multicenter samples, suggest that p-tau181 in ALS is best understood as a marker with diagnostic relevance rather than prognostic power. One limitation the researchers acknowledge is that survival data were not available for analysis, leaving open whether the protein might yet relate to longevity in ways the current design could not detect.

The practical implications reach into the design of clinical trials, where biomarkers that respond to therapy are desperately needed. Because plasma p-tau181 rises over the course of the disease independently of progression rate, it could serve as a sensitive readout of biological response to experimental treatments, complementing rather than replacing neurofilament light chain, which remains the gold standard for tracking disease severity. The authors also point toward the future of the field: newer tau markers such as p-tau217 have largely superseded p-tau181 as indicators of Alzheimer’s pathology and may offer sharper discrimination between Alzheimer’s and non-Alzheimer’s biology. For now, the message of this study is a sober lesson in biomarker humility. A protein approved to diagnose a dementia of the brain can rise to dementia-like levels in a motor neuron disease without reflecting a single tangle in the brain or a single failed word on a cognitive test. In ALS, blood tau is a messenger from the periphery, and learning to read it on its own terms may prove as important as the excitement that first brought it to the clinic.

Subject of Research: Plasma p-tau181 as a peripheral biomarker in amyotrophic lateral sclerosis and its dissociation from cognitive impairment and disease severity

Article Title: Impact of Plasma p‐tau181 on Cognition, Motor Phenotypes, and Disease Course in ALS

Article References: Kasper, E., Lehto, A., Nordmann, N., Peters, O., Hellmann, J., Priller, J., Spruth, E. J., Petzold, G. C., Vogt, I., Weydt, P., Bernsen, S., Dinter, E., Falkenburger, B., Günther, R., Düzel, E., Glanz, W., Synofzik, M., Beichert, L., Spottke, A., … Hermann, A. (2026). Impact of Plasma p‐tau181 on Cognition, Motor Phenotypes, and Disease Course in ALS. Annals of Clinical and Translational Neurology, 13(8), 1711-1718. https://doi.org/10.1002/acn3.70423

Image Credits: AI Generated

DOI: 10.1002/acn3.70423

Keywords: ALS, p-tau181, biomarkers, neurofilament light chain, motor neuron disease, cognitive impairment, frontotemporal dementia, Alzheimer's disease, plasma biomarkers, lower motor neurons, disease progression, clinical trials