Subtle cognitive deficits emerge before ALS symptoms appear, study finds

In the years before amyotrophic lateral sclerosis announces itself with weakened muscles and wasting limbs, the disease may already be leaving fingerprints on the mind. A longitudinal analysis of people carrying genetic variants for ALS has found that roughly one in seven unaffected carriers showed mild cognitive impairment at one or more study visits, with deficits concentrated in language and executive functioning — the same cognitive territories that define the frontotemporal dementia that can accompany ALS. The findings, drawn from the Pre-symptomatic Familial ALS (Pre-fALS) study, characterize cognition during the presymptomatic stage of the disease, a phase long hypothesized but difficult to measure. Using newly refined research criteria, the research team identified mild cognitive impairment in 20 percent of presymptomatic C9ORF72 carriers and 13 percent of SOD1 carriers, compared with just 4 percent of non-carriers. The results suggest that the frontotemporal axis of ALS begins stirring well before motor neurons fail — and provide a framework for identifying who might benefit from future prevention trials.

ALS is best known as a degenerative disease of the motor neurons, but cognitive and behavioural features are common in clinically manifest illness. Up to 10 to 15 percent of people with ALS also meet criteria for frontotemporal dementia, most often the behavioural variant subtype, and deficits in language, letter fluency and executive functioning are typical across the patient population. Apathy is the most common behavioural manifestation, and the constellation of non-motor symptoms has been encompassed by the term ALS frontotemporal spectrum disorder. These symptoms matter far beyond the clinic: they are associated with reduced quality of life, increased caregiver burden and compromised decision-making capacity regarding medical care. The discovery of the genetic causes of ALS and frontotemporal dementia transformed the ability to study these deficits before illness begins, because unaffected carriers of pathogenic variants can now be identified, followed and assessed with neuropsychological testing while they are still entirely asymptomatic.

The Pre-fALS study, initiated in 2007, is a natural history and biomarker study that recruits individuals at elevated genetic risk for ALS or frontotemporal dementia from across North America, including carriers of pathogenic variants in genes such as SOD1, C9ORF72, FUS, TARDBP and VCP. Participants undergo genetic counselling and testing before enrolment and may choose whether or not to learn their genetic status. Those in the non-disclosure group who turn out not to carry the familial pathogenic variant serve as blinded controls. All participants are asymptomatic at enrolment and are evaluated every 12 to 36 months with multimodal assessments including neuromuscular examination, electromyography and biomarker procedures. As of the May 2024 data cut-off, roughly 470 people had been screened and about 235 enrolled in the longitudinal cohort. Cognition has been assessed since 2011 with a comprehensive neuropsychological battery and since 2015 with the Edinburgh Cognitive and Behavioural ALS Screen, an instrument designed specifically for ALS.

A central challenge was deciding what should count as mild cognitive impairment before disease is manifest. The team compared published frameworks from the ALS and frontotemporal dementia communities and found unexpectedly divergent rates of impairment, which prompted them to refine the criteria proposed following the first International Pre-Symptomatic ALS Workshop in 2020. The resulting MCI-ft Criteria designate mild cognitive impairment — envisioned as a prodromal stage along the frontotemporal axis of the Miami Framework — based on two technical requirements. First, subthreshold cognitive performance, defined as scores at least two standard deviations below appropriate normative data at a single evaluation, or meaningful decline of at least 1.5 standard deviations from baseline performance. Second, this subthreshold performance or decline must be evident on two or more non-overlapping tests of executive functioning, including letter fluency, social cognition, or language. The criteria also retain an uncertain research designation for results whose interpretation is unclear, prioritizing specificity.

Applying these criteria to 216 Pre-fALS participants with neuropsychological data at 807 study visits, the researchers found that 31 participants — 14 percent — met criteria for mild cognitive impairment at one or more visits, accounting for 4 percent of all visits. The prevalence differed sharply by genotype: 15 of 74 C9ORF72 carriers, or 20 percent, showed impairment, as did 12 of 94 SOD1 carriers, or 13 percent, and 3 of 20 carriers of variants in other genes, or 15 percent. By contrast, only one of 28 non-carriers, or 4 percent, was ever classified as impaired. Notably, cognitive functioning in presymptomatic SOD1 carriers had not previously been described, and the finding adds to recent research challenging the view that affected SOD1 carriers are cognitively unimpaired — although the team cautions that research on this question relied on small samples and excluded those unable to complete testing because of motor symptoms.

The cognitive profile of the impairment was strikingly consistent. Of the 35 visits classified as showing mild cognitive impairment, 89 percent involved executive functioning, with letter fluency contributing to 40 percent of designated visits, 49 percent involved language, and 6 percent involved social cognition. Most visits — 69 percent — showed multi-domain impairment, and isolated subthreshold performance alone accounted for 46 percent of visits, evidence of decline alone for 23 percent, and both for 31 percent. Importantly, 63 percent of the impaired visits were classified based on comprehensive neuropsychological testing while performance on the ALS-specific screening instrument remained unimpaired. The researchers attribute this to two factors: the screening tool does not capture functions such as semantic fluency or cognitive flexibility, which were impaired on animal fluency and Stroop testing at several visits, and its summary scores can mask poor performance on a specific task — such as letter fluency — when other contributing tests fall within the expected range.

Perhaps the most intriguing finding was the temporal pattern. Fluctuations in cognitive status between assessments were common, with participants moving in and out of the impaired classification over repeated visits. The researchers could not identify specific reasons for these fluctuations, which did not appear to reflect varying performance on any single test. They note that fluctuating performance is not surprising in cognitive testing, where factors such as mood and fatigue can influence results — a concern heightened in people who know their genetic status and may self-monitor more closely. Alternatively, they suggest, frequent fluctuations may indicate early disease instability, a sign that the underlying neurodegenerative process is beginning to assert itself. Whether these fluctuations are associated with an increased risk of future phenoconversion to clinically manifest ALS or frontotemporal dementia is a question the team says future research must address.

The study also carries a caution about existing diagnostic tools. When the team compared their criteria with the Revised Strong Criteria, developed for clinically manifest ALS, the latter identified significantly higher rates of impairment — 17 percent of visits and 35 percent of individuals — because they permit classification based on the ALS screening instrument alone. Although 77 percent of visits identified by the new criteria were also captured by the Revised Strong Criteria, the older framework classified an additional 113 visits as impaired, most of which the new criteria designated as uncertain rather than impaired. The researchers argue that applying criteria developed for symptomatic patients to presymptomatic populations may overestimate impairment, and they recommend consensus review by a neuropsychologist and neurologist wherever feasible. Such review proved essential for interpreting confounding factors and for tests with ceiling effects, such as the Boston Naming Test, where an increase of just one or two errors can represent a large standardized decline.

The implications reach toward the emerging era of ALS prevention trials. Reliably designating unaffected carriers as impaired — or not — permits clear communication of cognitive status to research participants, empowers the study of the biological mechanisms underlying early frontotemporal dysfunction, and facilitates evaluation of the prognostic utility of impairment in predicting future phenoconversion. The team suggests that combining cognitive status with biomarkers such as neurofilament light chain, which has shown promise as a susceptibility marker for both ALS and frontotemporal dementia, might strengthen prediction of who will convert to clinically manifest disease. Not everyone with mild cognitive impairment will progress to frontotemporal dementia, and more longitudinal data are needed to understand the predictive value of the designation. But as early intervention trials aim to delay or even prevent phenoconversion, identifying carriers in the earliest stages of the disease process may prove decisive — and the mind, it turns out, may be where that process first becomes visible.

Subject of Research: Mild cognitive impairment in presymptomatic amyotrophic lateral sclerosis

Article Title: Mild cognitive impairment in presymptomatic amyotrophic lateral sclerosis

Article References: McHutchison, C., Wuu, J., Cosentino, S., Saldana, R., Vadella, E., Benacerraf, A., McBane, K., Gonzalez, A. S., Stark, J. H., Li, Y., Abrahams, S., & Benatar, M. (2026). Mild cognitive impairment in presymptomatic amyotrophic lateral sclerosis. Journal of Neurology, 273(11), Article 685. https://doi.org/10.1007/s00415-026-14206-0

Image Credits: AI Generated

DOI: 10.1007/s00415-026-14206-0

Keywords: ALS, mild cognitive impairment, presymptomatic disease, C9ORF72, SOD1, frontotemporal dementia, neuropsychology, Pre-fALS study, genetic carriers, phenoconversion, executive function, biomarkers