May 2026 Research Update

Across multiple adult ataxia cohorts, a repeat expansion in the FGF14 gene was repeatedly found as a relatively common explanation for otherwise unexplained, later-in-life cerebellar ataxia, and it was not reliably captured by all standard genetic approaches. Results also suggested that repeat size thresholds matter for how confidently the expansion points to this diagnosis, but repeat length did not consistently track with symptom severity or age at onset. [1] [2] [3] [4]

In a Spanish cohort of 53 adults with idiopathic late-onset cerebellar ataxia, 21 people had an expanded FGF14 repeat over 200, and expansions over 250 repeats were described as highly predictive for this cause. Brain MRI in this group sometimes showed cerebellar atrophy and a signal change in a pathway called the superior cerebellar peduncle. [1]

In a large Polish screening study of 701 adults with undiagnosed adult-onset cerebellar ataxia, 31 people had expansions of 250 repeats or more, while none were seen in 66 healthy controls. The study described both fully penetrant and incompletely penetrant expanded alleles, and it did not find a clear link between repeat length and age at onset. [3]

A German cohort of 107 genetically unresolved adults with ataxia found pathogenic expansions in 13% of the full group and also reported intermediate expansions, with yield depending strongly on how typical the clinical picture was. The authors highlighted that specialized methods (long-range PCR and long-read sequencing) can help detect these expansions and also assess mosaicism (different repeat sizes in different cells). [2]

In Brazil, FGF14 repeat expansions explained 2.8% of unrelated spinocerebellar ataxia families assessed at one center, with onset around the mid-50s and slow progression on average. In this group, repeat length did not correlate with disease severity or progression, while age at onset correlated with progression rate. [4]

FGF14 expansions were uncommon in a vestibular-focused cohort: among 116 people with definite bilateral vestibulopathy, three carried expansions of 250 repeats or more, and all three had downbeat nystagmus, with two showing cerebellar features consistent with an ataxia syndrome. One person had both an FGF14 expansion and a different genetic cause of vestibular-ataxia symptoms (RFC1-related disease), illustrating that more than one genetic condition can sometimes coexist. [5]

For adults with later-onset ataxia (and sometimes downbeat nystagmus or episodic symptoms), these studies reinforce that an FGF14 repeat expansion is a relatively frequent explanation in several populations, while it appears to be a rare cause when the main diagnosis is bilateral vestibular loss. The work mainly affects genetic diagnosis rather than day-to-day symptom management right now, and it also shows that the lab method used can affect whether this type of repeat change is detected. Uncertainties remain about how intermediate repeat sizes relate to symptoms and why repeat length does not consistently predict how severe disease will be.

Across several rare reports, genetic answers came from looking beyond standard spinocerebellar ataxia panels, including deep non-coding DNA changes, mitochondrial-related genes, and conditions that can mimic typical inherited ataxias for years. The overall message was that the same broad symptom of balance and coordination problems can arise from very different genetic mechanisms and may come with clues outside the cerebellum, such as vision changes, involuntary movements, epilepsy, or adrenal hormone abnormalities. [6] [7] [8] [9]

One person with SCAR20 (an autosomal recessive ataxia) was found to have a deep intronic SNX14 variant that created an abnormal "pseudo-exon" and lowered the SNX14 message, consistent with loss of gene function. The report showed how RNA sequencing of patient cells can reveal abnormal splicing that may be missed by initial DNA testing and later confirmed by reanalyzing whole-genome data. [6]

A separate single-patient report described adult-onset progressive ataxia later attributed to X-linked adrenoleukodystrophy due to a pathogenic ABCD1 variant, after very-long-chain fatty acids were found to be elevated and brain MRI changed over time to show white matter lesions. The case also noted adrenal testing abnormalities (high ACTH with normal cortisol), emphasizing that endocrine and imaging features can become clearer only with long follow-up in some people. [7]

A case-series described two children with de novo heterozygous AFG3L2 variants who had developmental delay along with movement disorders and optic involvement, and it also noted that inherited AFG3L2 variants in adults can cause spinocerebellar ataxia 28 with variable expressivity. The authors reported that levodopa had variable benefit in the described spectrum. [8]

One family report of spinocerebellar ataxia type 43 linked to an MME exon 21 variant described prominent involuntary movements (including mouth and tongue dystonia and choreiform movements) years before gait ataxia, with MRI showing no cerebellar atrophy and normal nerve conduction studies. The variant was seen in two affected sisters and a minimally symptomatic son, supporting autosomal dominant inheritance in this family while also showing that symptoms can start outside the classic "ataxia first" pattern. [9]

These reports matter most as background for diagnosis: they illustrate why some people with ataxia do not get an answer from initial genetic tests and why additional testing approaches (such as RNA-based studies, repeat analysis, or broader sequencing) may sometimes be needed to identify a cause. They also highlight that certain "extra" features (developmental delay, vision problems, involuntary movements, seizures, or adrenal-related lab changes) can point toward specific rare genetic explanations. Because the evidence comes from single cases or small series, it is uncertain how often these exact patterns apply to other patients.

Across these two studies in spinocerebellar ataxia type 3 (SCA3), brain imaging pointed to a clear pattern: the pons (part of the brainstem), the cerebellum, and the wiring between the cerebellum and the rest of the brain show the most consistent damage. Some measures on scans moved in the same direction as worsening symptoms or longer time with disease, supporting their use as progression markers, although how best to use them over time is still being worked out. [10] [11]

In a large multisite MRI study including people before clear symptoms (pre-ataxic), tissue loss increased in a steady pattern, with the greatest changes in the pons, cerebellar white matter, and the cerebellar peduncles (major connection pathways). The amount of tissue loss in these regions related to both symptom severity and disease duration, and cerebellar gray matter loss clustered in regions linked to movement control, attention, and emotion-related processing. The study also reported reduced coordinated size-relationships between the cerebellum and the cerebrum, and only limited involvement of the outer surface of the cerebrum except in the most severe group. [10]

In a smaller study using diffusion-based imaging of white matter connections, people with SCA3 showed widespread damage, especially along the cerebellum-thalamus-cortex pathway that helps connect coordination centers to movement areas. The overall network was less efficient at information transfer (lower global and local efficiency and longer paths), and a measure in a specific midline cerebellar region (vermis VI) was lower in people with longer genetic repeat length and worse movement impairment. This supports the idea that connection-based measures from brain scans could reflect both genetic load and day-to-day motor function. [11]

For patients and caregivers, these papers mainly add background about where SCA3 affects the brain most strongly and which scan measures tend to change as the condition becomes more severe. This matters now primarily for research studies and clinical trials that need objective ways to track progression, including in people who are not yet clearly symptomatic. What remains uncertain is how well any single scan measure will predict an individual person's future changes, and which measures will be most practical and consistent across different hospitals and scanners. [10] [11]

Across two cross-sectional human studies, blood-based measures differed from people without spinocerebellar ataxia, but the patterns were not the same across types. One study suggested changes in blood amyloid-beta balance in SCA12, while another suggested that cholesterol measures (especially HDL cholesterol) may track disease stage and severity in SCA3. Both studies support the idea that simple blood tests might reflect aspects of disease biology, while also showing that not every blood marker lines up with day-to-day symptoms. [12] [13]

In SCA12, researchers measured amyloid-beta 40, amyloid-beta 42, and tau proteins in plasma and compared results with healthy controls. People with SCA12 had lower amyloid-beta 40 and a higher amyloid-beta 42 to amyloid-beta 40 ratio, while total tau and phosphorylated tau were similar to controls. These blood differences did not correlate with cognition scores or ataxia severity scores in this study, meaning the test changes did not clearly mirror how a person was doing clinically. [12]

In SCA3, researchers compared cholesterol patterns across pre-symptomatic gene carriers, symptomatic patients, and controls, and they adjusted analyses for body mass index (BMI). After adjustment, symptomatic patients had higher LDL cholesterol and lower HDL cholesterol, and BMI partly hid (masked) these differences. HDL cholesterol was higher in pre-symptomatic carriers but lower in symptomatic patients, and lower HDL cholesterol predicted higher ataxia severity on the Scale for the Assessment and Rating of Ataxia. [13]

The SCA3 study also tested how well HDL cholesterol separated pre-symptomatic from symptomatic stages. HDL cholesterol contributed to stage discrimination with an area under the curve (AUC) of 0.837, suggesting it may help distinguish stages at a group level in research settings. This does not mean HDL cholesterol alone can define an individuals stage, but it points to a measurable blood signal that changes with disease progression in this dataset. [13]

For patients and caregivers, these studies are mainly background research about whether routine blood measures could eventually help track spinocerebellar ataxia in studies. The SCA12 work suggests an amyloid-beta signal in blood, but it did not track with symptom severity in that dataset, so its practical meaning is still uncertain. The SCA3 work suggests HDL cholesterol may reflect stage and severity at a group level, but because the research was cross-sectional (one time point), it cannot show how well these markers change within the same person over time or how they would perform in clinical care.

Across this months papers, most treatment research was preclinical: lab-grown nerve cells and mouse models were used to test ways of reducing toxic proteins, improving nerve cell structure, and changing inflammation or cell cleanup processes. These studies showed measurable changes in brain-related markers and, in mice, changes in movement tests, but they do not establish that these approaches help people with spinocerebellar ataxia. [14] [15] [16]

In lab-grown cerebellar Purkinje cells (key cells for coordination), two aromatic-turmerone analogs (A2 and A4) increased a protein-recycling pathway called chaperone-mediated autophagy and improved dendrite development. In several Purkinje-cell models expressing spinocerebellar ataxia proteins, the compounds reduced dendrite shrinkage and restored autophagy-related activity, and the effects were blocked by inhibitors of NRF2 or p38 signaling. [14]

A gene-silencing approach aimed at long CAG repeats (a cause of several polyglutamine diseases) used an engineered microRNA delivered by a viral vector (AAV5) to lower mutant huntingtin in cell and mouse models of Huntington disease, and it also reduced some polyglutamine proteins in patient-derived cells that included SCA1 and SCA3. In mice, it lowered mutant huntingtin in brain regions for up to 28 weeks with multiple safety checks reported as unchanged, but the authors noted they did not see clear improvement in mouse symptoms and that human safety and delivery questions remain. [15]

In a controlled study in SCA3 transgenic mice, intermittent theta-burst stimulation to the cerebellum improved coordination and gait measures compared with sham and also outperformed continuous theta-burst stimulation. The same mice had fewer ataxin-3 and ubiquitin-positive inclusions, lower signals linked to pro-inflammatory cytokines, and higher levels of autophagy markers. [16]

For patients and caregivers, these studies mainly serve as background research: they show multiple ways scientists are trying to influence processes thought to matter in some spinocerebellar ataxias (toxic proteins, inflammation, and the cells normal protein disposal systems). Because these results come from cells or mice, it is still uncertain whether they will translate into symptom improvement, which ataxia types might benefit, and what the real-world safety and practicality would be in people.

Two case reports focused on care approaches tailored to a persons particular situation rather than a one-size-fits-all treatment for spinocerebellar ataxia. Both reports described improvement after targeting a specific contributing issue (blood flow to the back of the brain in one case, and early visual function support in another), but each involved only one patient so the results may not apply broadly. [17] [18]

In a 22-year-old male with overlapping symptoms of amyotrophic lateral sclerosis (ALS) and spinocerebellar ataxia, clinicians identified an ischemic cryptogenic vascular dissection and treated it with endovascular stent repair plus dual antiplatelet therapy. Follow-up imaging showed improved blood flow in the posterior circulation and reported improvement in cerebellar atrophy measures alongside marked clinical improvement, with a favorable disability score at 21 months. [17]

In a 4-month-old infant with Spinocerebellar Ataxia type 29 (SCA29) and cerebral visual impairment, weekly visual stimulation therapy using high-contrast, brightly colored stimuli was followed by improvement in visual behavior and acuity. After 20 sessions, the report described a change from severe impairment to normal classification on a structured visual assessment scale. [18]

These case reports highlight that, alongside research into disease-targeted therapies, some improvements reported in real-world care came from identifying and treating a specific additional problem (such as impaired blood flow) or supporting a specific function (such as early vision) in a defined ataxia type. It remains uncertain how often similar situations occur, which patients might see similar changes, and how durable the improvements are beyond these individual reports.

Across these preclinical studies, several different inherited ataxias converged on a few shared drivers of nerve cell harm: trouble handling proteins, changes in the cell's waste-disposal systems, inflammation from supportive immune cells in the brain, and breakdown in the support systems that keep nerve signals fast and coordinated. [19] [20] [21] [22] [23]

In one mouse model, adding extra human STAU1 in the central nervous system led to abnormal gene activity, disrupted autophagy (a cell-cleanup process), activation of support cells, and changes in neuron markers, even though the mice did not show clear behavior changes up to 55 weeks. Reducing STAU1 with RNA interference improved these molecular problems, while combining STAU1 with a separate disease-related protein change (TDP-43 Q331K) made them worse, suggesting STAU1 can modify how degeneration unfolds. [19]

In a spinocerebellar ataxia type 34 (SCA34) mouse model, activation of microglia (immune cells in the brain) showed up before neuron loss and before motor problems started, and later changes were seen at synapses that connect key cerebellar circuits. The authors interpreted this timing as consistent with the idea that microglia-driven synaptic disruption may come before neuron death in this model. [20]

Support cells that make myelin also appeared able to drive disease features in an oligodendrocyte-specific SCA1 mouse model: mutant ataxin-1 limited to oligodendrocytes was enough to cause demyelination, damage to Purkinje cell axons, and impaired motor coordination. The study also identified distinct oligodendrocyte subtypes and highlighted transcriptional regulators (TCF7L2 and HTT) linked to oligodendrocyte dysfunction and loss of neuroprotective support. [21]

Two studies focused on how mutant ataxin-3 (SCA3) can amplify harmful cell stress in non-neuronal and neuronal contexts. In human microglia, polyglutamine-expanded ataxin-3 increased breakdown of CFTR through a specific tagging process (K63-linked ubiquitination) and pushed microglia toward a more inflammatory state, with the disease form (Ataxin-3Q80) having a stronger effect than normal protein. [22]

In cell and mouse model work on SCA3, Rad23b acted as a regulator that could worsen mutant ataxin-3 buildup: increasing Rad23b raised mutant ataxin-3 aggregates and neuron death, while lowering Rad23b reduced both. Although Rad23b helped deliver mutant ataxin-3 to the proteasome (a major protein disposal machine), it also reduced proteasome activity, which fit with the observed increase in aggregates; Rad23b also overlapped with mutant ataxin-3 aggregates in cerebellar neurons in SCA3 mice. [23]

Other lab studies pointed to additional cellular pathways that may feed into nerve cell vulnerability in ataxia. In engineered neuronal cell lines with ANO10 mutations linked to autosomal recessive spinocerebellar ataxia type 10 (SCAR10), protein profiling suggested disruptions in synaptic function, cell cycle control, extracellular matrix remodeling, and immune balance, with ties to calcium signaling. [24]

Finally, work in Purkinje neurons showed that the protein MAP6 helps place key ion channels (Kv3.3 and Cav2.1) in the right parts of the neuron, and lowering MAP6 reduced these channels in the soma and dendrites and reduced burst firing. Because these channels are linked to some spinocerebellar ataxias, the study supports the idea that incorrect channel targeting and altered firing may contribute to how cerebellar circuits become dysfunctional. [25]

For patients and caregivers, this set of studies mainly adds background about why nerve cells and cerebellar circuits may become stressed in different inherited ataxias: inflammation in microglia, reduced myelin support from oligodendrocytes, and multiple ways proteins can be misprocessed and accumulate. Because these are lab and animal studies, they do not yet show which of these mechanisms is most important in people, nor whether changing any single pathway would change day-to-day symptoms, but they help map out the biological steps that research groups are testing in models.

Across this animal study, a clear genetic cause was tied to a spinocerebellar ataxia-like condition in horses, and the researchers also showed how that DNA change affects gene activity and protein levels. The work points to a specific gene and a specific type of genetic change (in non-coding DNA) that disrupts how the gene is read, rather than changing the protein sequence directly. What remains uncertain from this report is how, or whether, this exact gene change relates to human spinocerebellar ataxia, since the study was done in horses. [26]

In 7 affected Quarter Horse foals compared with 36 unaffected horses (including carriers), the authors identified an 82 kb DNA region linked to equine juvenile spinocerebellar ataxia and narrowed this to an intronic (non-coding) variant in FDXR. Affected foals were homozygous for this variant, meaning they had two copies. [26]

The variant was tied to abnormal RNA processing: a cryptic exon (an extra RNA segment that is not usually included) appeared in spinal cord tissue, and overall FDXR mRNA was decreased. The study also reported lower FDXR protein levels in spinal cord and liver from affected horses, linking the DNA change to reduced gene output in tissues relevant to movement control and general metabolism. [26]

For people living with spinocerebellar ataxia, this study mainly serves as background research showing that ataxia-like diseases in animals can sometimes be traced to a single inherited gene change, including changes outside of protein-coding regions that still disrupt how a gene works. It does not show that FDXR causes human spinocerebellar ataxia, and it does not establish any implications for human testing or care. The most direct value right now is scientific: it illustrates how researchers can connect a DNA change to altered RNA and protein levels in nervous system tissue, while leaving open whether the same pathway matters in people.