February 2026 Research Update

This month added early but real-world signals that non-drug approaches can measurably change movement in some people with ataxia, although responses were not uniform and the evidence is still developing. A sham-controlled study in SCA3 found that cerebellar rTMS (a form of non-invasive brain stimulation) changed resting brain activity patterns, and those changes were linked to longer-term improvement in ataxia. A feasibility study of motor imagery paired with real-time MRI feedback found that some participants improved a finger-tapping task, suggesting this kind of training may help a subset of people rather than everyone. [1] [2]

In SCA3, active cerebellar rTMS increased beta-band brain rhythms and synchronization between certain brain regions, while the sham group did not show these EEG changes. The study also reported that these EEG features predicted longer-term improvement in ataxia, raising the possibility that simple brain-wave measurements could help track or anticipate response in future work. The report did not provide safety details in the summary provided. [1]

With motor imagery plus real-time fMRI neurofeedback, improvement was mixed: 9 of 16 people had better tapping accuracy after training, while the group as a whole did not show a clear statistical change. People who improved tended to show stronger imagery-related activation in motor-related brain regions during training, which may help explain why some people benefit more than others. One participant could not complete MRI due to dizziness, and the study noted key limits such as small size and no control group. [2]

A single-person gait report suggested that mixed reality (a headset overlaying digital content onto the real world) can make walking more cautious when used without physical support, with shorter steps and slower speed than normal walking. In that same person, pairing mixed reality with a stabilizing smart walker improved step length and walking speed compared with the walker alone, and the person reported less fear of falling. This highlights that assistive stability may strongly shape whether a new technology feels helpful or unsettling, but the finding is from one individual and may not apply broadly. [3]

For patients and caregivers, this set of studies mainly adds understanding about what is plausible and what might be measurable, rather than establishing a new standard approach. The strongest evidence here is the sham-controlled rTMS study in SCA3, which links a specific brain-wave pattern to later improvement, but details like how many people were studied and how large the clinical change was were not provided in the summary. The neurofeedback and mixed-reality reports are earlier-stage and show that benefits may be limited to some individuals or depend on the setup (for example, adding stable walking support). [1] [2] [3]

What became clearer this month is that researchers are not only testing whether symptoms change, but also trying to measure how the brain and movement change at the same time, which could support more personalized research in the future. What remains uncertain is how well these approaches work across different SCA types, how durable the effects are, and what the safest and most practical ways are to deliver them outside research settings. [1] [2] [3]

Across several human studies, this month strengthened the picture that SCA affects more than balance and gait, and that better measurement tools may detect meaningful change earlier or more sensitively than standard clinic scales. In SCA3, cognition measured by a brief screening test looked normal before clear ataxia but worsened after onset and tracked with disease severity. In SCA10 and SCA19/21, brain imaging measures showed changes that related to brain network activity or progressed over time, suggesting potential tracking tools. Separate work also showed that patient-reported vision difficulties are common in genetic ataxias and that a specific eye movement sign was closely tied to worse vision-related quality of life. [4] [5] [6] [7] [8]

In the READISCA cohort following SCA1 and SCA3 participants for about 3 years, several MRI measures changed more sensitively than the standard ataxia rating scale (SARA). Pons volume stood out as the most sensitive measure in both pre-ataxic and ataxic stages, and some MRI changes tracked with changes in function and daily activities. The study also provided sample-size estimates suggesting MRI-based outcomes could make early-stage trials more feasible, though the exact design assumptions matter. [5]

In SCA3, a large multicenter cohort found that cognitive differences (MoCA) were clearer after ataxia begins, and lower scores were linked to both worse disease severity and markers of brain change and nerve cell injury in blood (lower pallidal volume and higher plasma NfL). The study also reported that education level interacted with ataxia severity, meaning the same level of movement impairment did not relate to cognition equally in everyone. This helps explain why cognitive symptoms can feel quite different across people with the same genetic diagnosis. [4]

In SCA10, brain scans showed loss of gray matter in the cerebellum and motor-related brain areas, alongside increased connectivity in cerebellar and sensorimotor networks. Higher cerebellar network activity related to lower MoCA scores, while network activity did not clearly track the ataxia severity scale in this snapshot-in-time study. This pattern is consistent with the idea that brain networks may rewire as cells are lost, but the functional meaning can be complex. [6]

In a 7-year follow-up of a mother and daughter with SCA19/21, cerebellar mGluR1 PET (a scan that measures a specific brain receptor involved in signaling) decreased as disease progressed, and the authors suggested it may detect subtle changes better than MRI. Because this was only 2 related individuals, it is best viewed as an early signal about a possible tracking tool rather than a confirmed biomarker. [7]

A separate study focused on daily life impact rather than imaging and found that adults with genetic ataxias reported worse vision-related quality of life than healthy controls, with the biggest gap in visual processing speed. Among ataxia eye movement findings, gaze-evoked nystagmus was the only sign that independently predicted worse vision-related quality of life. This helps connect a specific, observable symptom to how vision problems may show up in everyday activities. [8]

For patients, this theme mainly matters because it supports clearer ways to describe and track what is happening, including thinking changes, vision difficulties, and brain changes that may not be fully captured by a single walking or balance score. Some of these measures (like MRI changes over time and blood NfL) are already widely used in research, while others (like mGluR1 PET) are still specialized and early. None of these studies, by themselves, establish what any one person should expect, but they do make it more concrete which symptoms and measurements tend to move together as disease progresses. [5] [4] [6] [7] [8]

What remains uncertain is how well these measurements perform across all SCA types and across different stages, especially when studies are small or cross-sectional. What became clearer this month is that researchers are building a more complete picture of progression that includes the brainstem and cerebellum on MRI, short cognitive tests, and real-world impacts like visual functioning. This work is most immediately relevant to research studies and future clinical trials, where choosing sensitive measures can change what researchers are able to detect. [5] [4] [6] [7] [8]

This month, multiple studies pointed in the same direction: genetics and blood-based measurements may help explain why people with SCA can look different from each other and may offer additional ways to classify or track disease. In SCA3, two independent blood-marker studies reported changes in metabolism-related signals compared with controls, and one linked a stress-responsive protein (IGFBP1) to both ataxia severity and a nerve-injury marker (NfL). In SCA3 and SCA2, large genetic studies suggested that factors beyond the main repeat expansion can modify when symptoms begin, including a modifier repeat in TCF4 for SCA3 and mitochondrial DNA haplogroups for SCA2. Alongside this, smaller genetic reports expanded the list of rare genes and variant types that can cause ataxia. [9] [10] [11] [12] [13]

In SCA3, a study of 32 people reported higher IGFBP1, IGF2, and free IGF1, along with lower insulin and a higher glucose-to-insulin ratio compared with controls. IGFBP1 levels tracked with both ataxia severity (SARA) and plasma NfL, and were lower in people with higher body mass index, suggesting a link between metabolic state and neurodegeneration markers in this group. The study also reported that people in the lowest insulin group had higher IGFBP1 and higher NfL than others. [9]

Another SCA3 study measured 24 routine metabolic markers in 101 people and reported that a 5-marker panel could distinguish SCA3 from controls, with similar performance in an independent validation cohort. Several markers also showed relationships with age at onset, repeat length, symptom severity scores, and disease duration. This supports the idea that blood chemistry patterns may carry information about SCA3, although the summary did not include how these markers behave over time within the same person. [10]

A large Chinese SCA3 study (1,439 people) tested whether an intronic repeat in TCF4 changes age at onset. TCF4 repeat length did not shift age at onset by itself, but it did strengthen the relationship between the ATXN3 repeat and age at onset, especially in early-onset cases. Cell experiments in the same report supported a two-way interaction, where expanded ataxin-3 worsened TCF4-related RNA problems and TCF4 expansion increased mutant ataxin-3 aggregation. [11]

In SCA2, a cross-sectional study of 217 people in India reported that mitochondrial DNA haplogroups were linked to age at onset even after accounting for ATXN2 repeat length, with haplogroup T tied to later onset and haplogroup M tied to earlier onset. The study did not find evidence that haplogroup changed the effect of CAG length itself, and the authors noted the findings need confirmation in other groups. This supports the broader concept that background genetics may partly explain why onset timing varies between families and populations. [12]

Several additional genetic studies expanded what is known about rare causes or unusual presentations linked to ataxia pathways. A Pakistani family study reported a new recessive GRM1 frameshift variant causing SCA13, adding to the known set of GRM1 changes tied to ataxia. A functional study in 8 patients with DNMT1-related disorders (including ataxia) reported that different DNMT1 variants can lead to different degrees and patterns of DNA methylation changes in blood, rather than one uniform signature. [13] [14]

One small case series highlighted that the same ATXN2 repeat length can behave very differently depending on whether it is inherited from one parent or both. In an Acadian family, three affected individuals had a homozygous ATXN2 expansion (39/39), while several heterozygous relatives (22/39) remained without symptoms into adulthood. This raises the possibility of reduced penetrance or other genetic modifiers in that family, but it is uncertain how often this pattern occurs elsewhere. [15]

For patients, these studies add to the idea that SCA is not explained by a single gene change alone, and that additional genetic background and blood signals may help explain differences in symptom timing and severity. The blood-marker findings in SCA3 are still exploratory and do not establish a single test that can diagnose or predict outcomes on its own, but they make it clearer which metabolic pathways are being examined and which markers line up with severity measures in research. The genetic modifier studies (TCF4 in SCA3 and mtDNA haplogroups in SCA2) may matter most for future research on prognosis and study design, rather than for day-to-day decisions now. [9] [10] [11] [12]

What remains uncertain is how stable these blood markers are over time and whether they change in step with symptoms in the same individual, since the reported studies were cross-sectional. It is also uncertain how widely genetic modifiers identified in one population apply to others. Even with these gaps, this month made the bigger picture clearer by linking specific modifier candidates and blood patterns to measurable features like age at onset, ataxia severity, and NfL in defined SCA groups. [9] [10] [11] [12] [15]

This month, case reports and small case series mainly reinforced two practical messages: ataxia-related genes can show up in unexpected ways, and standard testing may miss certain mutation types. Several reports centered on FGF14-related SCA27B, where the key change is an intronic repeat expansion that may not be captured by typical panel or exome testing. Other cases expanded the known range of rare genetic ataxias and showed that unusual genetic combinations can complicate how symptoms appear and when they start. [16] [17] [18]

Across 3 people described with SCA27B, the shared lesson was that the diagnosis may be delayed if testing does not include the right type of analysis for the FGF14 intronic GAA repeat expansion. Two older adults had negative first-line testing until whole-genome sequencing or targeted repeat testing identified SCA27B. A separate single-person report described prominent downbeat nystagmus and oscillopsia and noted a robust symptom improvement after 4-aminopyridine, with only transient improvement reported on lithium. [16] [17]

Other cases expanded rare genetic causes and presentations. A child with early neuroregression and ataxia had compound heterozygous VWA3B variants, broadening what has been reported for VWA3B-related ataxia. A Saudi patient report described SCA35 due to a TGM6 change, with FDG-PET showing reduced metabolism in one cerebellar hemisphere alongside cerebellar atrophy on MRI. [19] [20]

Two additional single-patient reports illustrated that ataxia can be misattributed or modified by other factors. One adult with no family history was initially labeled as having alcohol-related cerebellar damage but was later found to have SCA6, emphasizing that family history and common explanations do not always settle the diagnosis. One child with very early-onset SCA3 had both an ATXN3 expansion and an HTT expansion, and the authors proposed the large ATXN3 expansion likely drove early onset while the HTT expansion might influence the overall picture, though this is unproven from a single case. [18] [21]

Finally, one report connected an ataxia-related gene to a different movement disorder presentation. A child with tremor and dystonia but no ataxia had a new ITPR1 variant, and lab testing suggested it increased receptor activity, expanding the range of symptoms linked to ITPR1. This matters because it suggests some genes tied to ataxia can also appear as other movement problems, depending on the specific variant. [22]

For patients and caregivers, these reports mostly matter as context for why diagnoses can take time and why different types of genetic changes require different tests. The SCA27B cases particularly highlight that a repeat expansion in FGF14 can be missed by some common genetic approaches, which may help explain prior negative results in similar situations. The treatment response noted in the SCA27B nystagmus case is limited to one person and should be viewed as an individual clinical observation rather than a predictable outcome. [16] [17]

What remains uncertain is how often these unusual presentations occur and how well the single-case observations apply to others. Even so, this month added concrete examples of diagnostic pitfalls (such as mislabeling symptoms as alcohol-related or missing intronic repeat expansions) and expanded the catalog of genetic variants associated with ataxia-like syndromes. [18] [20] [19] [22] [21] [16]

Several laboratory and computational studies this month focused on what drives nerve cell stress and protein clumping in SCA, and which molecular pathways might be targeted in the future. In SCA1, patient-derived microglia (brain immune cells) and a mouse model suggested mutant ATXN1 can directly push microglia toward a more inflammatory, overactive state, and reducing mutant ATXN1 in microglia improved multiple disease features in mice. In SCA3 models, researchers reported that protein quality-control pathways are disrupted, and boosting a key protein (CHIP) partially restored stress responses and improved disease-related features in model systems. Other work explored aggregation modifiers in generic human cell models and tested small molecules or computational candidates that may reduce aggregation, but these findings are still far from patient use. [23] [24] [25] [26] [27]

In SCA1, human iPSC-derived microglia carrying mutant ATXN1 showed changes in shape and gene activity, along with increased phagocytosis and higher production of pro-inflammatory signaling molecules. In a conditional mouse model, reducing mutant ATXN1 specifically in microglia/macrophages improved microglial features, reduced supportive-cell scarring in the cerebellum, improved Purkinje neuron problems, and improved cognitive, mood, and motor deficits. This supports the idea that non-neuronal cells may actively contribute to symptoms and progression, at least in model systems. [23]

In SCA3 models, CHIP (a protein involved in handling misfolded proteins) was reported to become trapped in neuronal nuclear inclusions over time, with an age-related drop in soluble CHIP and reduced activity of the stress-response regulator HSF1. When CHIP was overexpressed, the study reported partial restoration of HSF1 function and improvement in several disease-related phenotypes in cells and mice. This places the CHIP-HSF1-DNAJB1 stress-response axis as a potential mechanism contributing to SCA3 pathology in preclinical work. [24]

Broader cell-based screening work developed reporter systems for polyQ and polyG aggregation and found that DNAJC7 acted as a strong suppressor of polyQ aggregation, with more context-dependent effects on polyG. In a separate cell study, gastrodin reduced insoluble ataxin-3 aggregation and altered signaling pathways, and blocking ERK signaling removed gastrodin's anti-aggregation effect in that model. These results identify possible modifier proteins and pathways, but they were generated in lab cell systems and do not predict how a treatment would work in people. [25] [26]

Two additional studies explored targets and pathways using computational or patient-derived neuron tools. A computational screening study predicted that the natural product crocin could bind tightly to AKT1 and TP53 in models, proposing it as a lead candidate that still requires experimental validation. In patient-derived SCA1 neurons, another study reported reduced GRM1 and GRM5 RNA and weaker responses of mGluR1/5 signaling, while also demonstrating a platform that could support higher-throughput testing of signaling-modifying compounds. [27] [28]

Finally, a human genetics plus mouse model study identified SNUPN variants in two families and showed that the altered protein had impaired nuclear transport, while a knock-in mouse developed cerebellar ataxia and widespread RNA splicing changes in Purkinje cells. This connects a specific genetic cause to a clearer cellular chain of events, from transport problems to splicing disruption to Purkinje cell abnormalities, in a model system. [29]

An evolutionary analysis of ATXN2 across species also argued that certain parts of the gene (such as exon 1B) might be preferable targets for exon-1-specific knockdown approaches, aiming to avoid disrupting conserved core regions. Because this was bioinformatic work, its main contribution is a hypothesis about how to design future gene-targeting strategies, not evidence that a specific therapy works. [30]

For patients, this theme is mainly early research that helps explain possible drivers of disease, such as immune-cell activation in SCA1 and impaired cellular stress responses in SCA3. These studies do not establish new treatments, but they provide clearer mechanistic targets that future therapies might aim to modify, and they show that research is expanding beyond neurons alone. The strongest reported effects here were in cell and mouse models, so it is still unknown which targets will translate into meaningful symptom changes in people. [23] [24] [29]

What became clearer this month is that several different research groups are converging on protein-handling and cellular stress pathways (such as chaperones and stress-response regulators) as relevant to polyQ diseases, while others are building tools to test signaling in patient-derived cells. What remains uncertain is which candidate molecules or targets can be safely and effectively reached in the human brain and whether changing these pathways will slow or change symptoms. [25] [26] [27] [28] [30]