Most practitioners working in functional or integrative medicine are familiar with anti-tissue transglutaminase 2 (TG2) as the coeliac disease marker. Fewer are routinely testing for its neurological counterpart, transglutaminase 6 (TG6), and the clinical consequences of that gap are substantial. A patient can have significant, progressive gluten-driven neurological damage with a completely normal TG2, a normal duodenal biopsy, and no gastrointestinal symptoms at all. If TG6 is not in the differential, the diagnosis is simply missed.
This piece is an attempt to bring the TG6 literature into clinical focus, because the science has been building quietly and steadily for over a decade, and the most recent evidence is compelling enough to change how we think about serological screening in neurological practice.
Understanding TG6: Where It Lives and What It Does
Transglutaminase 6 belongs to a family of calcium-dependent protein cross-linking enzymes. The family members relevant to gluten-related autoimmunity are TG2 (the intestinal autoantigen in coeliac disease), TG3 (the skin autoantigen in dermatitis herpetiformis), and TG6, whose primary home is the central nervous system.
TG6 is encoded by the TGM6 gene and is expressed abundantly in the brain, particularly in the cerebellar cortex where it is found in Purkinje cells, and in the thalamus, which functions as a key relay point for motor control. Its expression in mouse models correlates with neurogenesis during development, and it is upregulated in cortical and cerebellar neurons during differentiation. Mutations in the TGM6 gene have been independently linked to familial spinocerebellar ataxia, which is significant because it establishes that TG6 function is genuinely critical to cerebellar integrity, regardless of any immune-mediated mechanism.
From a biochemical standpoint, TG6 is regulated by calcium and guanine nucleotides, and molecular modelling suggests it shares binding site architecture with both TG2 and TG3. There is also some degree of antibody cross-reactivity between these three isoforms, which has implications for interpreting serological panels where TG2 and TG6 are both run.
The Gluten Connection: How Anti-TG6 Antibodies Arise
The pathogenic hypothesis runs broadly as follows. In genetically susceptible individuals consuming gluten, the same immune activation that drives TG2 autoimmunity in the gut can generate antibody responses directed against TG6 in the nervous system. TG6 acts as an autoantigen in gluten-related neurological disease, and IgA anti-TG6 deposits have been found in the blood vessels of cerebellar tissue in patients with gluten ataxia, which tells us that these antibodies are reaching and binding to CNS targets.
The question of where these antibodies are produced and how they gain access to brain parenchyma has been a persistent gap in the mechanistic literature. A 2025 study published in The Cerebellum provided the first detailed CSF immunological assessment in gluten ataxia patients, offering evidence that IgA anti-TG6 antibodies produced in the gut gain access to the CNS through a dysfunctional blood-brain barrier and contribute directly to cerebellar degeneration. The presence of plasma cells in the CSF of gluten ataxia patients and upregulated chemokines including IP-10 provide further support for an active intrathecal humoral response.
Animal model data adds another layer. Intraventricular injection of patient-derived TG-specific immunoglobulins in mice produces ataxia-like deficits, which moves the argument from association to plausible causation. These antibodies appear capable of causing neurological dysfunction directly, not merely serving as a bystander marker of an underlying process.
Gluten Ataxia: The Primary Neurological Manifestation
Gluten ataxia is defined as sporadic ataxia associated with serological evidence of gluten sensitivity, underpinned by autoimmunity to TG6, and characterised pathologically by loss of Purkinje cells throughout the cerebellar cortex.
A critically important clinical point is that gastrointestinal symptoms are absent in the majority of cases, present in only around 10% of patients. Enteropathy, if looked for with duodenal biopsy, is found in approximately half of confirmed gluten ataxia patients. In the remaining half, the neurological presentation is the only manifestation, and anti-TG2 and anti-EMA can be entirely negative. In this scenario, AGA and anti-TG6 are the only available serological markers.
The landmark diagnostic study by Hadjivassiliou and colleagues, published in Neurology in 2013, established the prevalence and specificity of TG6 antibodies across patient groups:
- 73% of confirmed gluten ataxia patients were TG6-positive
- 32% of patients with idiopathic sporadic ataxia were TG6-positive
- 32% of coeliac disease patients were TG6-positive (without overt neurological disease)
- 5% of neurological controls were TG6-positive
- 4% of healthy controls were TG6-positive
These figures carry two important messages. The first is that TG6 has meaningful specificity; background positivity in healthy controls is low. The second is that a substantial proportion of patients labelled as having idiopathic sporadic ataxia likely have an identifiable, treatable aetiology that is simply not being screened for.
Anti-TG6 antibodies were shown to be gluten-dependent in that study: after one year of strict gluten-free diet, titres were significantly reduced or undetectable in the majority of patients. That responsiveness to dietary intervention is both diagnostically informative and clinically actionable.
TG6 in Coeliac Disease Without Overt Neurological Symptoms
The neurological dimension of coeliac disease is almost certainly underappreciated. In a prospective cohort of newly diagnosed coeliac disease patients presenting to a gastroenterology department, 40% tested positive for anti-TG6, and this seropositive subgroup showed measurable atrophy of the cerebellum and thalamus on brain MRI, even in the absence of overt neurological complaints.
A seven-year follow-up of the same cohort found that the antibody-positive group had significantly greater rates of cerebellar grey matter atrophy compared to antibody-negative patients over time. Some patients in this group developed new-onset incoordination during the follow-up period despite being on a gluten-free diet, and all of those patients had residual positive serology for one or more gluten-related antibodies at follow-up. That finding matters clinically: dietary adherence sufficient to resolve gastrointestinal serology may be inadequate to eliminate TG6 antibodies, and TG6 monitoring provides a separate and neurologically specific endpoint to target.
The 2025 Sheffield Cohort: Cumulative Antibody Exposure and Brain Outcomes
The most recent major contribution to the field comes from Croall and colleagues, published in the Annals of Neurology in August 2025, using longitudinal data from a specialist neurological centre with routine TG6 testing. This study introduced the concept of cumulative IgA TG6 “exposure” as an analogue of smoking pack-years, representing the total antibody burden over time rather than a single cross-sectional titre.
The findings were clinically significant on several counts. Cumulative IgA TG6 exposure correlated with regional brain atrophy after correcting for age. Patients with persistent TG6 positivity had worse physical functioning scores, higher rates of anxiety and depression, and lower quality of life on validated measures. Greater self-reported adherence to a gluten-free diet was the strongest predictor of achieving a negative TG6 result.
The study’s conclusion is worth stating plainly: TG6 testing can identify patients at risk of accelerated brain atrophy, and achieving negative serology through strict dietary adherence should be the explicit clinical goal. The authors argue that IgA TG6 should be used as both a diagnostic and a monitoring tool in patients with relevant neurological presentations.
TG6 Antibodies in Other Neurological Conditions
Beyond ataxia, anti-TG6 antibodies have been identified across a range of other neurological presentations, though the strength of evidence varies considerably.
Amyotrophic Lateral Sclerosis. A study in JAMA Neurology found that mean IgA TG6 levels were significantly elevated in ALS patients compared to controls, with the authors concluding that a subset of ALS presentations may be associated with gluten-driven autoimmunity. The clinical picture in seropositive ALS patients was otherwise typical, making serological screening the only way to identify this subgroup. The question of whether gluten-free diet alters disease course in TG6-positive ALS has not yet been formally studied.
Cerebral Palsy. Anti-TG6 antibodies were found in 13% of cerebral palsy patients overall and 35% of the tetraplegic subgroup, compared to 6% of controls. Interestingly, there was no correlation between TG6 positivity and seropositivity to gliadin or other food proteins, which suggests the autoimmunity in this context may be driven by neurological injury and subsequent antigen exposure rather than by active gluten ingestion. An early brain insult may predispose to TG6 autoimmunity through a mechanism analogous to other post-injury autoimmune responses.
Multiple Sclerosis. A prospective cohort study found TG6 autoimmunity associated with brain atrophy in newly diagnosed gluten sensitivity patients, and TGM6 has been investigated as a potential biomarker of disease activity and astrocytic proliferation in MS. However, some studies in MS populations have reported negative TG6 titres across the whole cohort, and the relationship between TG6 and MS appears less robust than in ataxia. This remains an area requiring further well-designed study.
Huntington’s Disease. TG6 protein physically interacts with mutant huntingtin (mHTT) by co-immunoprecipitation, and mHTT aggregates colocalise within TG6-positive cells in rodent models. TG6 expression is particularly abundant in regions showing the highest mHTT aggregate burden. This raises the possibility that TG6 enzymatic activity contributes to aggregate formation in HD pathogenesis, though this is mechanistic animal model data and does not yet translate into clinical recommendations.
Paediatric presentations. Gluten ataxia in children is rare and poorly characterised in the literature. A 2025 BMJ case report described anti-TG6-positive gluten ataxia in a toddler, confirming that the presentation is not confined to adults, and noting that cerebellar MRI changes may be absent in early paediatric cases even where TG6 serology is positive. The paediatric literature remains sparse.
Antibody Development and Gluten Exposure Duration
A paediatric coeliac disease study found that anti-TG6 antibodies are more prevalent in children with untreated disease, and antibody levels correlated with duration of gluten exposure rather than with TG2 titre. There was no significant correlation between anti-TG6 and anti-TG2 concentrations, which reinforces the view that these two antibodies arise through distinct immunological pathways and are measuring different aspects of gluten-driven autoimmunity. Running TG2 alone will not capture TG6 status, and the two should be considered complementary rather than interchangeable.
When to Consider Anti-TG6 Testing: Clinical Indications
The following presentations warrant consideration of anti-TG6 serology as part of the diagnostic workup:
Neurological presentations with a suspected gluten connection:
- Cerebellar ataxia of unknown or unclear aetiology, particularly sporadic adult-onset ataxia where genetic testing is negative or inconclusive
- Gait disturbance with cerebellar features, with or without prior coeliac diagnosis
- Progressive balance problems with normal or borderline TG2 serology
- MRI findings of cerebellar atrophy, thalamic volume loss, or white matter abnormalities without a clear structural or vascular cause
- Abnormal cerebellar MR spectroscopy
Established gluten-related diagnoses with neurological involvement:
- Known coeliac disease with any neurological symptom including unexplained headache, cognitive change, peripheral neuropathy, or coordination difficulties
- Non-coeliac gluten sensitivity presenting with extraintestinal or neurological features
- Dermatitis herpetiformis with neurological symptoms (given the overlapping autoimmune spectrum)
Serologically complex or incomplete coeliac presentations:
- Seronegative or equivocal coeliac disease with neurological features
- Positive AGA with negative TG2 and EMA, particularly where neurological symptoms are present
- Patients on a gluten-free diet with persistent neurological symptoms and normalised TG2, where TG6 may remain elevated
Monitoring:
- Known TG6-positive patients on a gluten-free diet, where achieving negative serology is the therapeutic target
- Coeliac disease patients with subclinical neurological findings at baseline, to track whether antibody burden is accumulating over time
Other contexts where evidence is emerging:
- Unexplained peripheral neuropathy with evidence of gluten sensitivity elsewhere in the clinical picture
- Cognitive decline with functional medicine markers suggestive of gluten reactivity
- CNS hyperexcitability or myoclonus in the context of CD or gluten sensitivity
- Psychiatric presentations (psychosis, schizophrenia spectrum) with concomitant evidence of gluten-driven immune activation, where TG6 has been flagged in exploratory research
Treatment and Monitoring
The gluten-free diet is currently the only treatment with evidence of benefit. In gluten ataxia, studies show that patients on a strict GFD have improvements detectable on repeat cerebellar MR spectroscopy at one year. Given that Purkinje cell loss is largely irreversible, early identification and early dietary intervention are the clinical priority. The longer the delay, the more likely some neurological deficit will be permanent even after gluten is removed.
Strict GFD adherence matters specifically for TG6 normalisation. Adherence sufficient to resolve gastrointestinal symptoms or normalise TG2 may be inadequate to eliminate TG6 antibodies. Monitoring with serial TG6 serology, rather than relying on symptom improvement alone, gives a more accurate picture of whether neurological autoimmunity has been genuinely suppressed.
Limitations and Open Questions
Intellectual honesty requires acknowledging where the evidence has gaps.
The mechanistic case for TG6 antibody pathogenicity is supported by animal model data and CSF findings, but formal proof of causality in human disease is still lacking. The correlation between TG6 antibody levels and the degree of cerebellar damage is not straightforward, and the 2025 Sheffield study noted a somewhat unexpected absence of correlation between antibody titre at any single timepoint and symptom severity, which is why cumulative exposure over time appears to be the more informative variable.
TG6 testing is still not widely available in standard clinical laboratories, particularly in North America. The Sheffield Centre has driven the majority of the clinical research, and access outside specialist neurology settings remains limited. In functional medicine practice, panels such as Cyrex WheatBurden (previously Array3X) include TG6 IgA and IgG as part of a comprehensive gluten and wheat immune reactivity screen, which is currently one of the more accessible routes in clinical practice.
The relationship between TG6 seropositivity in the absence of any neurological symptoms (as seen in the 40% of newly diagnosed coeliac disease patients) and the long-term risk of developing overt neurological disease has not been formally quantified. Whether asymptomatic TG6-positive coeliac patients warrant more aggressive monitoring or faster escalation in GFD support is a clinical question the current literature cannot yet fully answer.
Finally, the anti-TG6 literature, while growing, is still heavily concentrated in a small number of research groups. Independent replication across larger and more geographically diverse cohorts would strengthen confidence in the prevalence figures and clinical thresholds currently in use.
Conclusion
TG6 represents the neurological arm of the transglutaminase autoimmune axis in gluten-related disease. The evidence is now strong enough to support its use as both a diagnostic and monitoring biomarker in patients presenting with cerebellar ataxia, progressive balance disorders, or neurological features in the context of known or suspected gluten sensitivity.
The clinical picture that TG6 serology reveals is one where the gut and the brain are connected by a shared autoimmune vulnerability, and where standard coeliac serology can be entirely normal even as neurological damage accumulates. For practitioners who see complex neurological presentations alongside chronic inflammatory or autoimmune conditions, adding TG6 to the differential is a low-risk, high-yield move that the literature now supports with reasonable confidence.
Hadjivassiliou et al., Neurology 2013;
Croall et al., Annals of Neurology 2025;
Thomas et al., Amino Acids 2013;
Hadjivassiliou et al., Brain 2003;
Croall et al., PMC 2021 (7-year follow-up cohort); Rouvroye et al.,
The Cerebellum 2025 (CSF markers study).
