Glutaric Acidaemia Type 1

Glutaryl-CoA dehydrogenase deficiency (glutaric aciduria type I), related to GCDH.

Literature and regulatory information updated: 6 October 2026.

0. Initial Reference Table

1. Introduction

Glutaric acidaemia type 1 is a “cerebral” organic aciduria: unlike methylmalonic, propionic or isovaleric acidaemia, its course is rarely dominated by severe acidosis, hyperammonaemia or hypoglycaemia, and the decisive injury occurs in a restricted anatomical territory, the striatum, during a limited period of brain development. An affected child may appear healthy until gastroenteritis or a respiratory infection precipitates acute encephalopathy, losing head control, sucking and sitting abilities within hours. The resulting putaminal and caudate necrosis is irreversible and determines a lifelong dystonic movement disorder [6, 13]. This asymmetry between an apparently normal previous state and permanent sequelae explains why GA1 has become one of the clearest examples of benefit from tandem mass spectrometry newborn screening and structured family education [8, 14].

The disease was described in 1975 by Goodman and colleagues in two siblings who, from three and seven months of age, developed a neurodegenerative disease with opisthotonus, dystonia and spasticity; one had chronic compensated metabolic acidosis, with serum bicarbonate between 7.5 and 15.7 mEq/l [13, 15]. In the same year, Goodman and Kohlhoff demonstrated deficient glutaryl-CoA dehydrogenase activity [16]. The group led by Stephen Goodman, who died in 2020, maintained a central role in disease investigation, and the 2023 recommendations credit him with its discovery in 1974 and successive contributions to its characterisation, diagnosis and treatment [1, 3].

Over subsequent decades, three observations changed clinical understanding. The first was recognition that macrocephaly, present at birth or within the first weeks, is often the initial sign and precedes neurological injury, as described by Iafolla and Kahler in 1989 and Hoffmann and colleagues in 1991 [13]. The second was identification of communities with very high disease frequency, such as the Old Order Amish of Lancaster County, Pennsylvania, and Oji-Cree populations in Manitoba and northwestern Ontario, whose founder variants allowed study of series of children with the same genotype and prolonged follow-up, such as the 77 patients described by Strauss and colleagues in 2003 [17–19]. The third was demonstration that many patients excrete small or intermittent quantities of glutaric acid, making conventional biochemical diagnosis insensitive and leading to isotope dilution techniques, 3-OH-GA measurement and C5DC assays [6, 13].

The international cross-sectional study published in 2006, involving 279 patients from 35 centres, established that 95% of first crises occur by two years, that the latest recurrent crisis occurred at 70 months and that diagnosis while asymptomatic followed by L-carnitine and lysine-restricted diet best predicted a good outcome [7]. The first evidence-based guideline was published in 2007 and revised in 2011, 2017 and 2023 [3, 20–22]. German prospective studies of children identified through newborn screening showed that full adherence to recommendations reduces movement disorder to approximately 5–7%, whereas delayed emergency treatment is associated with striatal injury in all reported cases [8, 23].

The current landscape combines this success with new questions. Fewer acute crises have made insidious onset, progressive white matter abnormalities in high excretors, cognitive impairment, subclinical peripheral neuropathy and chronic kidney disease relatively more visible; the latter appears unmodified by recommended therapy [3, 8, 24, 25]. In parallel, research seeks substrate reduction and gene replacement therapies, one of which has entered a phase 1 trial [26, 27]. This article organises information from reference textbooks, GeneReviews, the third revision of international recommendations and recent literature, consistently distinguishing natural history, treated disease and experimental data.

2. Molecular Genetics and Pathophysiology

2.1. Gene, protein and metabolic pathway

The GCDH gene is located at 19p13.13 according to current nomenclature; original mapping by in situ hybridisation in 1994 placed it at 19p13.2, a designation still appearing in the Atlas and Orphanet entry [1–3, 13]. It has 11 exons spanning approximately 7 kb [13]. It encodes glutaryl-CoA dehydrogenase, a mitochondrial matrix protein of the acyl-CoA dehydrogenase family functioning as a homotetramer of 45 kDa subunits, binding flavin adenine dinucleotide (FAD) and transferring electrons to electron transfer flavoprotein (ETF) and ETF:ubiquinone oxidoreductase [4, 13]. The enzyme catalyses glutaryl-CoA dehydrogenation and subsequent glutaconyl-CoA decarboxylation, generating crotonyl-CoA and CO₂; labelled-substrate studies show that glutaconyl-CoA remains enzyme-bound, as the only released products are crotonyl-CoA and CO₂ [6, 13].

The reaction is the shared step in L-lysine, L-hydroxylysine and L-tryptophan degradation. Lysine, quantitatively the most relevant precursor, is converted to saccharopine by aminoadipate-semialdehyde synthase (AASS), then to 2-aminoadipic and 2-oxoadipic acids, which the 2-oxoadipate dehydrogenase complex (with DHTKD1 subunit) converts to glutaryl-CoA [6, 13]. Enzyme deficiency causes upstream glutaryl-CoA accumulation, hydrolysed to GA, converted to 3-OH-GA by a mechanism not yet clarified, and esterified with carnitine to form C5DC. Urinary glutarylcarnitine excretion consumes free carnitine and causes secondary carnitine deficiency, with an increased acylcarnitine/free carnitine ratio in plasma and urine [1, 6].

2.2. Injury mechanisms

The explanation for selective striatal vulnerability during the first years remains incomplete. Experimental data suggest several complementary mechanisms. Glutaryl-CoA inhibits the 2-oxoglutarate dehydrogenase complex, analogously to succinyl-CoA, compromising the Krebs cycle; GA disturbs the dicarboxylic acid shuttle metabolically coupling astrocytes and neurons; and 3-OH-GA promotes excitotoxicity, inducing neuronal degeneration through NMDA receptor activation in striatal slice cultures and seizures with striatal injury after intrastriatal injection in rats [6, 13]. Structural similarity between glutaric and glutamic acids and inhibition of cerebral glutamic acid decarboxylase by GA and 3-OH-GA supported the initial excitotoxic hypothesis; these metabolites also indirectly modulate glutamatergic and GABAergic neurotransmission [13].

The blood–brain barrier has low permeability to dicarboxylic acids. Metabolites produced within the central nervous system become trapped in the brain compartment, where GA and 3-OH-GA concentrations have exceeded plasma levels by 10 to 1000 times [6, 13]. Disturbed autoregulation and regional perfusion pressure gradients have also been proposed, and increased lysine residue glutarylation has been demonstrated in glial cells, impairing mitochondrial proteins such as glutamate dehydrogenase and carbonic anhydrase 5b, with clinical significance still undetermined [6]. A 2025 mouse study showed that an odd-chain dicarboxylic acid (undecanedioic acid) is shortened in peroxisomes to glutaryl-CoA and reproduces the GA1 biochemical profile, suggesting that ω-oxidation stimulated by catabolism may increase mitochondrial substrate load; this is preclinical proof of concept [28]. GeneReviews emphasises that it is unknown whether toxins cross the blood–brain barrier or are synthesised in the brain, and why a childhood striatal vulnerability window exists [1].

Neuropathology shows neuronal loss and astrocytic proliferation in the caudate and putamen, sometimes globus pallidus, more extensive in older patients, with spongiform change predominantly in white matter; despite apparent cortical atrophy on imaging, cortical neuronal loss has not been found, and microvesicular lipid may be present in the liver [13]. The kidney strongly expresses GCDH; GA and 3-OH-GA interfere with proximal tubular cell organic anion transporters, and animal models have demonstrated acute renal toxicity during induced crises and chronic toxicity [3].

2.3. Variant spectrum and genotype–phenotype correlation

More than 200 pathogenic or likely pathogenic variants have been described, 290 according to HGMD consulted by the recommendations group in December 2021, mostly missense and generally private [1, 3]. Many appear to affect protein stability, folding and tetramerisation [1, 29]. The disease mechanism is loss of function [1].

Genotype correlates with biochemical phenotype but not striatal injury risk. High excretors have absent or very low residual activity (0–2% according to recommendations; 0–3% according to GeneReviews), and newborn screening detects them with 100% sensitivity. Low excretors have up to 30% residual activity and at least one hypomorphic missense allele; screening sensitivity is 84% [1, 3, 8]. In the series of 215 enzymatically confirmed patients in Copenhagen since 1975, predominantly of European ancestry, more than half had undetectable activity, 34% had up to 5% and four exceptional cases reached 30%; no correlation of genotype or biochemical phenotype with clinical course was found among the 79 patients with clinical data [30]. In the Spanish 43-patient cohort, variants p.Ala293Thr and p.Arg402Trp had allele frequencies of 30% and 28% in high excretors, and p.Val400Met and p.Arg227Pro, present only in low excretors, accounted for 53% of that group’s alleles [29]. Despite equal striatal injury risk in both groups, high excretors have higher intracerebral GA and 3-OH-GA on magnetic resonance spectroscopy, more extrastriatal abnormalities, greater head circumference, higher subdural haematoma risk and poorer cognitive performance [3].

The following table collates clinically or population-relevant variants, normalised to MANE Select transcript NM_000159.4 (numbering in NM_000159.3, used in GeneReviews, is identical for these variants, verified in VariantValidator) [1, 5].

The Atlas attributes “c.1240C>T” to p.Arg402Trp and designates p.Ala293Thr as “Ala293Trp” resulting from “c913G>A”; these designations do not match normalised nomenclature, and VariantValidator verification confirms c.1204C>T for p.Arg402Trp and c.877G>A for p.Ala293Thr, with c.1240G>A being the Lumbee p.Glu414Lys variant [5, 13]. The Atlas also reports, in a Saudi family, variant p.Leu179Arg, then undescribed in another family, in a 47-year-old man with mild hand dystonia and impaired gait, whose two diagnosed nephews had no neurological abnormalities [13]. Intrafamilial variability with the same genotype is the rule: course depends mainly on whether acute striatal injury occurs, and the Atlas describes one four-year-old sibling unable to sit and another eight-year-old performing well at school [1, 13].

GeneReviews notes that an infant with positive screening, normal blood GA and 3-OH-GA and only one identified pathogenic variant may warrant close clinical surveillance, but that, given high molecular analysis sensitivity, the likelihood of disease risking striatal necrosis is low [1].

3. Clinical Spectrum and Manifestations by System

3.1. Untreated disease phases

GeneReviews describes a spectrum from severe dystonia due to basal ganglia injury to asymptomatic individuals, with late neurological manifestations in between, and distinguishes infantile onset (acute injury between three months and three years, rarely up to six) from late onset, defined by manifestations appearing after six years [1]. Untreated, 80–90% of children sustain irreversible striatal injury during a vulnerable brain development period, mainly between 3 and 36 months, with individual reports up to 72 months [3].

Neonatally and in the first months, most infants are asymptomatic or show non-specific signs: muscle hypotonia with delayed head control, high palate, feeding difficulties, irritability, jitteriness and vomiting. These signs are reversible and alone have little prognostic value [6, 13]. Macrocephaly is present at or shortly after birth in 75%, may begin prenatally and prompt caesarean delivery, peaks between 3 and 6 months and, at six months, may place head circumference above the 98th percentile or 2 to 5 standard deviations above the mean; crossing percentiles between 3 and 9 months is the most useful clue [1, 6, 13]. In an 11-infant series, the two without macrocephaly never experienced an encephalopathic crisis [13].

Acute encephalopathic crisis occurs in untreated patients at a median 9 to 10 months [6]. In the international 279-patient study, 95% of first crises occurred by two years, at a median nine months, and the latest recurrent crisis occurred at 70 months [7, 13]. In the Amish series of 77 patients followed for 14 years, 78% were diagnosed after striatal necrosis, developing between two and 18 months, and no child sustained basal ganglia injury after their second birthday [13, 19]. The trigger is generally febrile infectious illness, mainly gastroenteritis, but also fasting, surgery or a febrile vaccine reaction [3, 6]. Strauss and colleagues described putaminal injury as heralded by abrupt behavioural arrest, with degeneration resembling stroke in tempo, radiological appearance and irreversibility, symmetrical and selective [19].

During a crisis, motor skills are acutely lost: head control, sucking and swallowing, sitting, standing and grasping objects. The child often remains alert, with profound neck and trunk hypotonia, limb rigidity and dystonic or athetoid hand and foot movements, particularly when agitated [6, 13]. Seizures and paroxysmal electroencephalographic changes may occur, and raised cerebrospinal fluid protein reinforces a mistaken suspicion of encephalitis; in one Atlas patient, cerebrospinal fluid protein was 500 mg/dl [13]. Usually the metabolic decompensation of “classical” organic acidurias is absent: bicarbonate may be chronically low or low during acute illness, but in most patients is normal, and ketosis, hypoglycaemia, hyperammonaemia or raised transaminases occur only rarely [6, 13]. Recovery is slow and incomplete, and successive episodes may occur, each followed by further deterioration [13].

3.2. Insidious and acute-on-insidious onset

Some patients develop striatal injury without an apparent crisis. This form, initially estimated at 10–20% (or 15%) of symptomatic patients, accounts for up to 50% of symptomatic cases in newborn screening cohorts, mainly associated with deviations from recommended diet [1, 3, 13]. Dystonia develops over weeks or months, sometimes preceded by hand tremor or mild oral dyskinesia, and tends to be less severe [6]. In a study of 21 symptomatic patients, median dystonia onset was 10 months in both acute and insidious forms; severity was greater after crisis (8 severe and 5 moderate) than with insidious onset (4 mild, 3 moderate, 1 severe); lesions were confined to dorsolateral putamen in 8 of 10 insidious-onset patients, and an asymptomatic latency of 3.5 months to 6.5 years separated lesion detection from clinical manifestation [34]. Two insidious-onset patients subsequently experienced an acute crisis, with extensive lesions superimposed on previous dorsolateral lesions, termed acute-on-insidious [3, 34]. In the prospective German cohort, insidious-onset movement disorder manifested later (median 630 days) than acute onset (median 270 days) [3].

3.3. Movement disorder and neurological course

Striatal injury produces a complex movement disorder, predominantly dystonic, with orofacial dyskinesia and dysarthria superimposed on axial hypotonia; chorea, grimacing, opisthotonus, clenched fists and tongue protrusion may occur [6, 13]. With age, mobile dystonia tends to become fixed and may be associated with akinetic-rigid parkinsonism or spasticity [3, 13]. Severe dystonia may progress to status dystonicus, causes pain and reduces quality of life and life expectancy [3]. Many children have sudden dystonic spasms mistaken for seizures, although chronic epilepsy is rare [13]. Cognitive function may initially be preserved; the Atlas describes siblings with pronounced dystonia, normal intelligence and normal magnetic resonance imaging (MRI) [13].

Chewing and swallowing difficulties, vomiting, aspiration and increased energy expenditure from hypertonia lead to poor weight gain and malnutrition; kyphoscoliosis and chest wall dystonia cause restrictive lung disease [6]. Profuse sweating, unexplained hyperpyrexic crises, irritability, anorexia and insomnia are frequent in patients with neurological sequelae; hepatomegaly, fasting hypoglycaemia in older children and adults, one death with a Reye-like syndrome and a case of recurrent rhabdomyolysis, fatal during the third episode, with creatine kinase between 78,000 and 189,000 IU, have been described [13]. In a retrospective study, 30% of 114 symptomatic patients underwent surgery for orthopaedic complications such as scoliosis or hip dislocation [3].

Seizures occur in 7%; they may accompany crisis or be the initial manifestation, and rare infantile spasms have been described [1]. Epilepsy prevalence was not increased in early-treated patients in the two largest newborn screening cohorts [3].

3.4. Subdural haematoma and retinal haemorrhages

Enlarged frontotemporal subarachnoid spaces stretch bridging veins, which may rupture spontaneously or after minor trauma, particularly near the first birthday when walking begins; haematoma may be accompanied by retinal haemorrhages, and parents have been wrongly accused of abusive head trauma [1, 6, 13]. The initially reported 20–30% subdural haematoma frequency was not confirmed by systematic studies: in a Dutch 25-patient registry, lifetime incidence was 4%; in a systematic review, 19 of 20 children with GA1 and subdural haematoma had other disease-specific brain abnormalities [35, 36]. However, systematic review of 168 MRI scans from 69 patients found haematomas in eight high excretors aged 5.8 to 24.4 months (two after minor accidental trauma, six incidental findings); in 36 patients without trauma studied between 3 and 30 months, incidence was 16.7% (16% in screen-detected cases), higher after acute onset (33.3%) than insidious onset (14.3%) or in asymptomatic patients (5.9%), and always with wide frontoparietal spaces and frontotemporal hypoplasia [37]. Recommendations add that haematoma peaks at 10–14 months, has not been observed after 36 months and is mainly associated with “relative” macrocephaly (disproportion between cranial cavity and brain) [3]. It may occur in screen-detected, appropriately treated children without macrocephaly and may resolve spontaneously [1].

3.5. Late onset and adult disease

Some patients are diagnosed in adolescence or adulthood through family investigation, non-specific neurological manifestations or a false-positive newborn screen in their child. Reported symptoms include chronic headache, vertigo, transient ataxia, reduced fine motor skills, syncope after exercise, tremor, epilepsy, polyneuropathy, incontinence, dementia and psychiatric symptoms, with MRI showing periventricular white matter abnormalities, frontotemporal hypoplasia and subependymal nodules without striatal injury [1, 3, 13]. In a series of eight late-onset patients, all were high excretors; two were diagnosed after 60 years with dementia, tremor and epilepsy, three were asymptomatic mothers identified through their children’s screening; all had frontotemporal hypoplasia and white matter abnormalities, and six had subependymal lesions [38]. Authors concluded that these patients are not categorically different from those treated early and differences probably reflect greater cumulative neurotoxicity, leading recommendations to question whether they constitute a distinct variant [3, 38]. In 2026, a case was published of a 43-year-old woman with levodopa-responsive parkinsonism and documented presynaptic dopaminergic dysfunction, with compound heterozygous GCDH variants, one (c.1178G>A, p.Gly393Glu) very rare [39].

3.6. Extraneurological manifestations and long-term complications

Chronic kidney disease was the first recognised extraneurological manifestation. In the European phenotypic spectrum study of organic acidurias and urea cycle disorders, chronic renal failure occurred in adolescents and adults with GA1 at similar frequency with or without movement disorder [40]. In the German newborn screening cohort, renal function tended to decline with age, independently of neurological phenotype, biochemical subtype and treatment quality: three had stage 2 chronic kidney disease and ten intermittent stage 2–3a, and no patient known to the recommendations group required dialysis [3, 8]. Isolated acute kidney injury cases and a fatal atypical haemolytic uraemic syndrome have been described [3, 14].

Three cases of central nervous system malignancy (medulloblastoma, glioblastoma) have been published in patients aged 6 to 23 years without guideline-compliant treatment or with late onset; causation remains uncertain, and GeneReviews considers increased susceptibility unproven speculation [1, 3, 41]. A study of 21 patients using sciatic nerve MRI neurography showed frequent abnormalities, particularly in high excretors, with reduced motor amplitude, slowed conduction and prolonged tibial somatosensory evoked potentials exclusively in that subgroup; neuropathic symptoms were rare (4 of 15 high excretors) [25]. Optic chiasm thickening in six of ten patients in a French series and mild hearing loss in a Taiwanese series of 13 with methodological limitations have also been described [3].

Cognitively, 72 German screen-detected children had a median IQ of 87, below the general population; high excretors had a median 84 and low excretors 98, without influence of treatment quality, sex or motor phenotype [24].

3.7. Neuroimaging

MRI reflects region- and age-dependent pathology. Before any symptoms, temporal hypoplasia (in 95% of patients), enlarged cerebrospinal fluid spaces anterior to the temporal lobes and Sylvian fissures, open opercula, immature gyral pattern, delayed myelination, subependymal pseudocysts and isolated globus pallidus T2 hyperintensity may exist; these abnormalities may develop in the third trimester and regress completely with treatment begun neonatally [6, 13]. Acute striatal injury starts in the dorsolateral putamen and progresses dorsoventrally, with the putamen more affected than the caudate and the caudate more than the globus pallidus; ventricular dilatation and post-crisis basal ganglia lesions indicate permanent damage and correlate with motor deficit [6, 13]. Periventricular white matter and U-fibre abnormalities increase with age and are more frequent in high excretors [6, 13]. In 18 Dutch patients aged 11 months to 33 years, mostly diagnosed before universal screening, open opercula were found in 15, enlarged cerebrospinal fluid spaces or ventriculomegaly in 9, abnormal basal ganglia signal in 8, white matter abnormalities in 5 and subdural haematoma in 1 [1]. The French PNDS distinguishes three stages of striatal injury during a crisis — an acute phase with cytotoxic oedema, a subacute phase with vasogenic oedema and a chronic phase with atrophy —, notes that the injury is usually symmetrical but may be asymmetrical and that, outside crises, there are progressive forms limited to the dorsolateral putamen, with white matter abnormalities independent of decompensations (delayed myelination, signal change, corpus callosum atrophy, subependymal pseudocysts) that may worsen with age [11].

4. Clinical Recognition and Warning Signs

Clinical recognition of GA1 remains necessary despite newborn screening: some patients were born before its introduction, some children come from countries without screening for this disease and low excretors may be missed. Recommendations therefore require investigation of suggestive signs even after a normal screen [3]. Isolated macrocephaly is frequent but has a low positive predictive value because approximately 3% of the general population lies above the 97th percentile [3]. The French PNDS adopts a broader threshold: GA1 should be considered for any macrocrania above +2 standard deviations, present from birth or even prenatally, or for rapid head circumference growth in the first two years, even if isolated, and particularly when macrocephaly is associated with dystonic movements; suspicion is also strong in a child under 6 years who, after a febrile episode or another catabolic situation, develops dystonia, chorea or myoclonus, axial hypotonia, dysarthria and, less often, epilepsy, with relatively preserved intellectual abilities [11].

In a child with known GA1, signs requiring immediate metabolic centre contact are temperature above 38.5 °C, vomiting or diarrhoea or other intercurrent illness symptoms, and any new neurological signs (reduced consciousness, hypotonia, irritability, rigidity, dystonia, chorea) [1, 3]. During the vulnerable period, crisis may develop within hours without gradual or substantial metabolic disturbance [13]. Rapid head circumference growth may indicate subdural collections or haemorrhages [1].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Diagnostic criterion

Diagnosis is established in an individual with suggestive findings and biallelic pathogenic or likely pathogenic GCDH variants or, when molecular findings are uncertain, significantly reduced glutaryl-CoA dehydrogenase activity in cultured fibroblasts or leukocytes [1, 3]. Two variants of uncertain significance, or a pathogenic variant with a variant of uncertain significance, neither establish nor exclude diagnosis [1]. Macrocephaly, acute encephalopathy, bilateral basal ganglia injury, dystonia, subdural haematoma, retinal haemorrhages and elevated GA, 3-OH-GA or C5DC are suggestive but not confirmatory because none is specific [1, 3].

5.2. Newborn screening

Population screening measures dried blood spot C5DC by tandem mass spectrometry (MS/MS); some laboratories use ratios with other acylcarnitines as secondary parameters. Each laboratory defines its cut-off, and no controlled studies establish pathological values [3]. Introduction of multiple reaction monitoring increased sensitivity and reduced false positives in German newborn screening society reports [3]. GA1 is included in national panels in 24 European Union countries and Switzerland [3].

Sensitivity is not uniform. In the prospective German study including 98.3% of children with GA1 identified through screening nationally between 1999 and 2016, cumulative sensitivity was 95.6%, but 84% for low excretors; recommendations cite 75–84% for this group and 100% for high excretors [3, 8]. GeneReviews gives 96% as overall sensitivity [1]. Patients missed by screening had neurological outcomes as poor as in the pre-screening era [8]. The Oji-Cree variant c.91+5G>T is associated with such low excretion that patients are not reliably diagnosed by gas chromatography–mass spectrometry or MS/MS, so these communities had DNA-based newborn screening [13, 18]. The Atlas describes an infant with elevated C5DC in the first sample and normal on repeat, diagnosed only at 11 months after developing dystonia during an infection; the state programme adjusted its cut-off and began recommending full assessment of every positive result rather than repeat sampling [13]. Over time, C5DC may decrease through depletion of carnitine stores, and low excretors have had normal C5DC even during supplementation [13, 46]. In a 20-year review at a Texas centre, all 20 referred GA1 patients had been identified by screening, including two low excretors [42]. In France, at the time of the PNDS (April 2021), GA1 was not yet screened: the Haute Autorité de Santé had recommended in February 2020 the extension of newborn screening to seven metabolic diseases, including GA1, with C5DC measurement in dried blood collected on the third day of life, and the PNDS notes that, even with screening, low excretors may give false negatives [11].

C5DC also increases in multiple acyl-CoA dehydrogenase deficiency (glutaric aciduria type 2), renal failure, maternal GA1 and pseudoglutarylcarnitinaemia in medium-chain acyl-CoA dehydrogenase deficiency [3]. The 2022 ACMG action sheet lists “none” for elevated C5-DC differential diagnosis; this wording differs from European recommendations listing the above causes, and the ACMG index retains a separate sheet for glutaric aciduria type 2, screened through C4 and C5 [3, 47, 48]. Liquid chromatography-MS/MS methods and acylcarnitine ratios have been developed to improve low-excretor detection; carnitine loading may improve sensitivity but has not been systematically studied [3]. An untargeted metabolomic dried blood spot study in 29 patients identified candidate biomarkers requiring independent validation [49].

5.3. Confirmation of a positive screen and targeted investigation

The 2023 recommendations (recommendation #2, level A) require positive screening results and suggestive clinical, biochemical or neuroradiological findings to be confirmed by quantitative urinary and/or blood GA and 3-OH-GA and, if abnormal, GCDH molecular analysis and/or leukocyte or fibroblast enzyme assay [3]. A positive result should be repeated in the same blood spot, preferably at the same laboratory [3]. GeneReviews prefers quantitative GA and 3-OH-GA studies, where available, over conventional organic acid analysis, requiring simultaneous treatment and confirmatory testing if either is abnormal [1].

Raised 3-OH-GA, generally with raised GA, is highly suggestive of GA1. Normal values do not exclude the disease because some low excretors have intermittently normal concentrations, and isolated borderline values cannot distinguish a low excretor from a carrier or a preanalytical problem [3]. Metabolic treatment should therefore begin as soon as 3-OH-GA is elevated, before genetic or enzymatic confirmation, and there is no treatment stratification by biochemical subtype [1, 3]. The French PNDS, on the same principle, requires treatment to be started without waiting for molecular confirmation when clinical, neuroradiological and biochemical findings converge; it defines the two biochemical phenotypes by urinary GA above or below 100 mmol/mol creatinine, with no correlation with clinical severity, and stresses that in low excretors only 3-OH-GA and C5DC are raised, which requires specialised laboratories with sensitive methods and their own reference ranges; lumbar puncture is not part of the diagnostic strategy [11].

Molecular analysis has 98–99% sensitivity according to recommendations; GeneReviews indicates that sequencing detects over 99% of known pathogenic variants and no intragenic deletions or duplications have been reported [1, 3]. Single-gene testing (sequencing and copy-number analysis), a multigene panel including GCDH and differential diagnosis genes, or exome or genome sequencing may be used; rapid sequencing is necessary in critically ill infants. Most known variants are coding and exome-detectable [1]. Recommendations advise starting with genetic analysis because of greater availability and importance for counselling and prenatal diagnosis, allowing initial enzyme assays according to local availability and family preference [3]. Diagnosis may also be established molecularly without prior biochemical testing [3].

If only one (or no) variant is found and other suggestive findings exist, a variant undetectable by routine techniques should be considered and enzyme activity measured: significantly reduced activity confirms, normal or heterozygote-range activity excludes [3]. Enzyme assay limitations include difficulty distinguishing carriers from low excretors with high residual activity, the need for 3–5 ml of blood for the leukocyte assay and the small number of laboratories; according to the recommendations, clinical assays were available only at Professor Salomons’ laboratory in Amsterdam and Professor Wibrand’s in Copenhagen [1, 3]. The French PNDS confirms that this assay is not performed in France and reserves it for particular circumstances, basing confirmation in all cases on the identification of biallelic pathogenic or likely pathogenic variants in GCDH; a biological and genetic diagnostic algorithm is given in annex 3 of the protocol [11]. Molecular reports should indicate gene, versioned transcript (NM_000159.4), HGVS nomenclature, zygosity and phase when studied in parents, ACMG/AMP classification and method limitations; apparent homozygosity may result from an undetected deletion or uniparental isodisomy [1]. Treatment response does not reclassify a variant of uncertain significance.

5.4. Practical algorithm

In the United States, the American College of Medical Genetics and Genomics (ACMG) ACT Sheet for elevated C5-DC, updated July 2022 and classified as “time critical”, lists immediate actions after positive screening: inform the family, ascertain clinical status (feeding difficulties, macrocephaly, hypotonia), consult a paediatric metabolic specialist, assess the newborn, initiate specialist-recommended confirmatory testing and treatment, provide basic GA1 information and report results to the screening programme [47]. For diagnostic assessment, the sheet lists urinary organic acids (GA and 3-OH-GA, typically elevated but sometimes slightly increased or normal), plasma acylcarnitines (elevated C5-DC) and urinary C5-DC, potentially helpful in low GA excretors; molecular analysis “may confirm” diagnosis [47]. Its algorithm proceeds from elevated C5-DC to one step combining plasma acylcarnitines, urinary organic acids and urinary C5-DC; compatible results lead to GA1 diagnosis, followed by molecular analysis linked by a dashed arrow, indicating optional testing; incompatible results are considered false-positive screening needing no further action [47]. This differs from the European sequence in two respects: 2023 recommendations require molecular and/or enzyme analysis as confirmation, particularly because low excretors may have intermittently normal metabolites, and do not allow normal 3-OH-GA to exclude disease with suggestive findings [3]. The ACMG sheet adds clinically that presymptomatic treatment may prevent irreversible neurological symptoms, untreated late-onset disease (after 6 years) may cause headache, vertigo, dementia and ataxia, even treated patients may risk other manifestations such as kidney disease, and an infant with macrocephaly and subdural haematoma may mimic shaken baby syndrome [47].

In targeted investigation, dried blood spot or plasma C5DC is less important because of lower sensitivity in low excretors and secondary carnitine deficiency; urinary C5DC is a poorly available alternative, less sensitive than quantitative urinary 3-OH-GA [3]. Two patients without elevated blood C5DC had appreciable urinary glutarylcarnitine excretion [13]. No systematic data exist on diagnostic cerebrospinal fluid GA and 3-OH-GA, although the Atlas describes cases in which this was the only fluid with elevated values [3, 13]. Lysine loading and prolonged fasting tests are potentially dangerous and obsolete, and in vitro tests do not increase sensitivity [3, 6].

5.5. Biochemical reference values

The Atlas records GA excretions of 850–1700 and 900–1200 mmol/mol creatinine in initial descriptions, versus 1.1–8 mmol/mol in controls, and excretors with 80–200 mmol/mol. Plasma GA ranged between 5 and 18 µmol/l, with normal values (0.6–3 µmol/l) also observed in patients; cerebrospinal fluid values ranged between 20 and 40 µmol/l [13]. Glutaconic acid may exceed 3-OH-GA during acute ketosis, when urine also contains 3-hydroxybutyric, acetoacetic, adipic, suberic and sebacic acids [13]. Most patients have low plasma free carnitine and increased esterified carnitine, particularly in urine, and low muscle carnitine has been reported in an asymptomatic patient [13]. In the clinical case studies, the operational definition used was urinary GA above 100 mmol/mol creatinine for high excretors and below that value for low excretors; an “intermediate” subgroup with 100–1000 mmol/mol before treatment falls into the low-excretor range during treatment [3, 46].

5.6. Special contexts

ClinGen classified GA1 as moderately actionable for adult secondary findings, explicitly excluding newborn and prenatal screening because effectiveness was extrapolated from paediatric studies and adult disease is probably milder [50]. This framework does not apply to children, in whom early diagnosis benefit is demonstrated. Prenatal diagnosis preferably uses molecular analysis when familial variants are known; GeneReviews does not recommend prenatal biochemical diagnosis, although amniotic fluid GA and amniocyte enzyme assays were used historically [1, 13].

6. Differential Diagnosis

The differential has three axes: macrocephaly, encephalopathy with basal ganglia injury and isolated metabolite elevation. Recommendations list benign familial macrocephaly and communicating hydrocephalus, other metabolic macrocephaly disorders (such as Canavan disease), hepatic and uraemic encephalopathies, “metabolic strokes” in methylmalonic and propionic acidurias, urea cycle and mitochondrial disorders, non-metabolic encephalopathies (encephalitis, meningitis, poisoning, Aicardi–Goutières syndrome), biochemical causes of elevated GA or 3-OH-GA, and asphyxia, HIV encephalopathy, cerebral palsy and abuse [3]. Orphanet adds Reye syndrome, familial infantile bilateral striatal necrosis, familial megalencephaly, postencephalitic parkinsonism, battered child syndrome with chronic subdural effusions, sudden infant death and vaccine-attributed brain injury [2]. Saudubray includes GA1, Canavan disease and L-2-hydroxyglutaric aciduria among small-molecule disorders with macrocephaly [51].

Glutaric aciduria type 2 is addressed here only as a differential diagnosis. Distinguishing GA1 from GA3 relies on the GA:3-OH-GA ratio, which is very high in GA3 and not seen in GA1 [1]. Regarding subdural haematoma, a Texas series of 20 screen-detected patients found only one incidental haematoma, in a patient with other typical abnormalities, and the authors concluded that investigating GA1 in isolated subdural haematoma with negative screening has limited value [42]. GeneReviews and the recommendations agree: without other GA1 findings on MRI, pretest probability is low and targeted investigation is unnecessary [1, 3]. The French PNDS follows the same logic: in an infant with subdural haematoma, the first hypothesis to pursue, in order of frequency, is non-accidental head injury, and GA1 is investigated with specific tests only when clinical abnormalities were already present (abnormal head circumference growth, psychomotor delay) or when imaging is suggestive [11]. For biochemical interpretation, the PNDS summarises four situations that mimic GA1: intense ketosis of prolonged fasting, which raises urinary GA and 3-OH-GA; short-chain 3-hydroxyacyl-CoA dehydrogenase deficiency, with raised 3-OH-GA without GA; multiple acyl-CoA dehydrogenase deficiency, with raised GA and C5DC accompanied by 2-hydroxyglutaric acid and without 3-OH-GA; and renal failure, which raises plasma C5DC [11].

7. Treatments and Risk Management Strategies

7.1. Principles and care organisation

Treatment aims to reduce lysine oxidation and promote physiological glutaryl-CoA detoxification [1]. It rests on three components that are effective only together: a low-lysine diet with a specific amino acid mixture, L-carnitine supplementation and intensified emergency treatment during any potentially catabolic episode [3]. Treatment starts as soon as the disease is suspected, for example with raised 3-OH-GA in a screen-positive newborn, without awaiting confirmation [1, 3]. It should be instituted and assessed by an interdisciplinary team at a centre experienced in inherited metabolic diseases (recommendations #1 and #5, level A), including specialists in metabolic disease, paediatric neurology, neuroradiology, nutrition, dietitians, nursing, physiotherapy, occupational therapy, speech and language therapy, psychology and social work, with a continuous metabolic emergency service [3]. In the German study, metabolic centre supervision reduced movement disorder from 57% to 18% (odds ratio 6.17; 95% confidence interval [CI] 1.15–33.11) [23]. In France, the PNDS assigns follow-up to a reference or competence centre for inherited metabolic diseases, with a specialised physician and dietitian, and gives the general practitioner two roles: referring any suspicion urgently, particularly with acute neurological symptoms, and, during follow-up, supporting dietary adherence, ensuring vaccinations, monitoring growth and development and collaborating in the management of motor disability; amino acid mixtures, which are foods for special medical purposes, are reimbursed under long-term condition (ALD) no. 17, and their prescription, like that of low-protein foods, is the responsibility of teams familiar with their composition [11].

After confirmation, recommendations propose brief centre admission to start maintenance treatment and teach parents maintenance and emergency treatment and impending catabolism signs theoretically and practically, providing emergency cards, contacts and written plans, with interpreters when needed. The nearest hospital and paediatrician should be informed [1, 3]. Recommendations emphasise that while postsymptomatic treatment has very limited effectiveness, presymptomatic guideline-compliant treatment keeps over 90% neurologically asymptomatic [3].

7.2. Dietary treatment

Lysine is essential and sufficient quantities must support growth. Low-lysine diet reduces the quantitatively most relevant neurotoxic metabolite precursor, and animal brain GA and 3-OH-GA concentrations vary with lysine intake [3]. Direct lysine calculation is more precise than total protein restriction because lysine content varies from 2–4% of cereal protein to 9% in fish [1, 3]. Until 6 years (recommendation #6, level A), a low-lysine diet with lysine-free, tryptophan-reduced, arginine-enriched amino acid mixture supplemented with minerals and micronutrients is recommended; the formula should be tryptophan-reduced but not tryptophan-free because depletion may cause severe neurological deficits [1, 3]. Since these mixtures’ introduction, tryptophan deficiency has not been described, and suspected deficiency requires high-performance liquid chromatography or MS/MS measurement [3].

Sources: Table 2 of 2023 recommendations, Table 5 of GeneReviews and Table 22.1 of Saudubray, with matching values [1, 3, 6]. Values should be modified if growth and development are abnormal. Recommendations provide no numerical natural protein values because these depend on food source [3]. Lysine-rich foods include poultry, fish, prawns, shellfish, pork, beef, soya, nuts, seeds, eggs, beans and lentils [1].

Breastfeeding should be encouraged. Most commonly the prescribed amino acid mixture is given first and breastfeeding allowed on demand afterwards, as in phenylketonuria; breast milk contains approximately 86 mg lysine per 100 ml, allowing daily intake calculation [1, 3]. Data on preterm lysine, protein and energy requirements are very limited [1]. In dystonic children, energy expenditure may rise to 130–150% (higher in status dystonicus), but immobility may also reduce it, so intake should be adjusted by intensive clinical and dietary surveillance; sweating and tachypnoea cause water and salt losses [3]. Children with feeding difficulties require metabolic dietitian supervision and may need nasogastric feeding, gastrostomy, jejunostomy or fundoplication [1, 3, 6].

After 6 years (recommendation #7, level B), diet becomes protein-controlled, based on safe levels and avoiding excessive lysine-rich foods, with regular dietary follow-up during transition [3]. Liberalisation rests on the end of striatal vulnerability, uncertain extrastriatal abnormality significance and independence of kidney disease from treatment quality; continued restriction is justified by unknown long-term course, renal manifestations, progressive extrastriatal abnormalities and progressive neurological symptoms in late-diagnosed adults [3]. In the largest US cohort, 47% receiving low-protein diet without formula developed movement disorder, versus 7% receiving low-lysine diet with formula [3, 52]. In that cohort, formula and carnitine adherence fell to 12% and 32% at 7 years without serious adverse effects, and authors consider dietary intervention beyond early childhood uncertain [52]. In the German cohort, diet was safe with appropriate anthropometric development into adulthood except with severe dystonia [3].

Practical implementation of the diet according to the French PNDS. In France, the 2009 national consensus uses the lysine “part”, equivalent to 25 mg, as the unit of calculation; calculation in lysine is preferred to calculation in protein because it reduces day-to-day variation, and natural protein should be distributed throughout the day, ideally half of animal origin (milk and dairy products allowed) and half of plant origin [11]. The amino acid mixture, lysine-free and low in tryptophan, is preferably taken in three doses and also provides vitamins and trace elements; the sum of the protein equivalent of the mixture and of natural protein must cover at least the age-appropriate population nutritional references, and in the absence of a mixture, vitamin, mineral and trace element preparations with a dosage calculated for each micronutrient are essential [11]. Low-protein foods, with less than 10% of the protein of a standard food of the same category, supplement energy but are poor in micronutrients, and must not replace the foods “to be controlled”, because of the risk of malnutrition [11].

Table 7.2a — Classification of natural foods in the French PNDS [11]

The PNDS also gives the mean lysine content per gram of protein: approximately 100 mg in fish, 90 mg in meat, 86 mg in breast milk, 80 mg in cow’s milk, 70 mg in potato, 60 mg in egg and soya, 30–75 mg in vegetables and fruit, 25–55 mg in nuts and 30 mg in cereals [11]. The protocol’s prescription examples illustrate its application: in a 2-month-old infant weighing 5.4 kg, 57 g of infant formula and 50 g of amino acid mixture provide approximately 98 mg lysine/kg/day, 1.2 g protein equivalent/kg/day and 97 kcal/kg/day; at 3 years, weighing 14 kg, the combination of weighed natural foods, low-protein foods and 12.5 g protein equivalent from the mixture gives approximately 60 mg lysine/kg/day and 82 kcal/kg/day, and in a child of the same age with dystonia and dysphagia the mixture moves to overnight enteral nutrition; at 8 years, without central nervous system injury and without mixture, the relaxed diet provides approximately 2 g protein/kg/day [11]. Products available in France for GA1 include GA1 Anamix Infant® and GA1 Anamix Junior® (Nutricia), GA gel® (Vitaflo) and XLYS Low TRY Maxamum® (SHS) [11].

The PNDS describes three diet modalities before 6 years. The cruise diet provides the lysine needed for growth. The emergency diet — prescribed at diagnosis during decompensation, during decompensation and in the face of an at-risk catabolic situation — is given in hospital and for a short time (48 hours at most, because of the risk of malnutrition): it removes natural protein, maintains the amino acid mixture, reintroduces natural protein at 50% and then 100% over the following 24–48 hours and increases energy to 120% of usual, with 50–60% carbohydrate and 40–50% lipid, using low-protein foods, dextrin-maltose, oils or lipid emulsions and glucose-lipid powders, orally, by continuous enteral feeding or, if there is digestive intolerance, intravenously [11]. The semi-emergency diet that follows reduces lysine to half of the cruise diet and keeps energy close to 120% of references [11]. After 6 years, the emergency diet in situations of intense catabolism consists of temporary withdrawal of animal protein and an increase in uncontrolled foods rich in carbohydrate and fat, with glucose-lipid enteral nutrition if the child eats poorly without digestive complaints [11]. Enteral nutrition, by nasogastric tube or by gastrostomy if prolonged, is indicated for energy deficit in infants and when neurological injury compromises swallowing, and allows intake to be maintained, the emergency diet and medicines to be given and intercurrent illnesses to be treated at home [11]. In children with severe dystonia or status dystonicus, protein-energy requirements may reach at least 120% of references despite immobility, and food texture must be adapted to dysphagia (small pieces, smooth purée, gelled water) because of the risk of aspiration pneumonia [11].

Arginine competes with lysine at the blood–brain barrier CAT1 transporter. Although the Atlas described arginine fortification as a promising investigational approach, 2023 recommendations (recommendation #8, level 0) do not recommend additional isolated arginine supplementation during maintenance or emergencies beyond food and amino acid mixture content; only supraphysiological doses reduced brain GA in animal models, and potential adverse effects include metabolic acidosis and arterial hypotension [1, 3, 13].

7.3. L-carnitine

Carnitine conjugates glutaryl-CoA into water-soluble, excretable C5DC, compensates secondary depletion and affects oxidative stress parameters in models. Supplementation is associated with lower striatal injury risk in early-diagnosed patients and lower mortality in symptomatic patients; no randomised trials examine its isolated effect [3, 7]. Recommendation #9 (level B) calls for lifelong supplementation, with an initial oral dose of 100 mg/kg/day in three doses (three to four according to GeneReviews), with no maximum dose defined in the guideline (the US label limits the paediatric oral dose to 3 g/day), adjusted to maintain plasma or dried blood spot free carnitine within the reference interval, measured at trough at least 12 hours after the last dose (recommendation #18) [1, 3]. Reported adverse effects are diarrhoea and fishy body odour, which may be socially stigmatising and prompt dose reduction; in isolated patients riboflavin reduced trimethylamine-related odour. Whether prolonged supplementation increases atherosclerotic risk through intestinal trimethylamine N-oxide formation is unknown, and the recommendations consider that the benefit outweighs the potential risk [1, 3]. C5DC increases during supplementation and is unsuitable for monitoring [3]. The French PNDS sets the same initial dose, 100 mg/kg/day of oral L-carnitine in three doses up to 6 years, allows it to be lowered to 30–50 mg/kg/day after 6 years and in adults, proposes in its text an annual carnitine measurement documenting adherence and normal values (its surveillance table, taken from the 2017 recommendations, links carnitine to the amino acid schedule: every 3 months in the first year, every 6 months up to 6 years and annually thereafter), and doubles the dose in at-risk situations: 200 mg/kg/day in 3 to 4 doses, oral or intravenous (or continuous intravenous), before 6 years, and 100 mg/kg/day in 3 to 4 doses, not exceeding 12 g/day, after 6 years and in adults [11]. This maximum of 12 g/day is higher than the oral limits of the US label (3 g/day in children; 1–3 g/day of solution in adults), which serve only for comparison [9].

Sources: DailyMed/openFDA labels and Drugs@FDA; EMA medicines list consulted 7 October 2026 [3, 9, 10, 54, 55].

7.4. Emergency treatment

Emergency treatment is the most effective component for preventing acute striatal injury and the strongest neurological outcome predictor [3, 6]. Recommendation #10 (level A, “very high” clinical relevance) requires immediate, aggressive initiation during any febrile illness, warning symptom or perioperative procedure until 6 years [3]. Principles are preventing or reversing catabolism through high energy intake (with insulin for hyperglycaemia and lipids if necessary); reducing GA and 3-OH-GA production by withholding or reducing natural protein for 24 (to 48) hours; enhancing detoxification and preventing carnitine depletion; and correcting dehydration, electrolytes and acid–base balance [3]. Organisational measures include regular parental education, updated written protocols for parents, local hospital, paediatrician and school, a laminated emergency card, home maltodextrin, amino acid mixture, carnitine and antipyretic supplies, advance notice to the nearest holiday destination centre, and immediate centre contact if temperature exceeds 38.5 °C, vomiting or diarrhoea or new neurological signs develop [1, 3]. GeneReviews adds medical alert bracelets and car seat stickers and a medical record “alert” [1].

Outpatient treatment (until 6 years). Permissible only if clinically well, temperature below 38.5 °C, oral tolerance and no warning signs, with reassessment every 2 hours by trained parents and centre telephone support; GeneReviews limits this attempt to 12 hours [1, 3]. If successful, maintenance is reintroduced over 48 (to 72) hours [3].

Source: Table 4 of the 2023 recommendations (solutions every 2 hours, day and night; the amino acid mixture may be fortified with maltodextrin if tolerated) [3]. Natural protein is reduced to 50% or withheld for at most 24 hours and then increased to the maintenance plan over 48–72 hours; the amino acid mixture continues if tolerated; oral carnitine is doubled (for example 200 mg/kg/day in infants); if temperature exceeds 38.5 °C, ibuprofen or paracetamol is given at 10–15 mg/kg per dose, 3–4 doses daily, maximum 60 mg/kg/day [3]. For paracetamol, this dose matches the Portuguese SmPC (Ben-u-ron®: 10–20 mg/kg per dose at 6–8-hour intervals, maximum 60 mg/kg/day) [56]; for ibuprofen, the recommended dose is the same, and the SmPC (Brufen®: from 3 months and 5 kg, 20–30 mg/kg/day in 3–4 doses at least 6 hours apart, for example 5 mg/kg three times daily from 3 to 6 months) serves for comparison: 10–15 mg/kg per dose up to 60 mg/kg/day exceeds the labelled daily dose and lies outside its conditions [57]. GeneReviews permits occasional antiemetics to improve enteral tolerance [1].

Inpatient treatment (until 6 years). The child should be transferred without delay to the centre or to the nearest hospital, under centre supervision, if recurrent vomiting or diarrhoea, feeding intolerance, high fever or suspicious neurological signs develop [3]. The table compares the sources consulted. The French PNDS advocates, before 6 years, systematic admission for any at-risk situation, with the emergency diet given orally or enterally and a switch to the intravenous route if feeding is impossible or poorly tolerated or if there is no improvement within 12 hours [11].

Differences in intravenous carnitine (usual dose in the 2023 recommendations, doubled dose in the British and Brazilian protocols) and in glucose rate reflect different versions of the recommendations (the British protocol is based on the 2011 revision and the Brazilian on that of 2017) and have not been directly compared; the choice should follow the reference centre protocol [3, 58, 59]. The recommendations and protocols define no maximum dose for glucose, insulin or carnitine in emergencies: glucose and insulin are adjusted to blood glucose, and the US intravenous levocarnitine label gives 300 mg/kg as the highest dose administered [3, 10]. The Atlas describes Morton’s protocol for the Amish, with an initial bolus of 5% dextrose in Ringer’s lactate with bicarbonate, followed by 12.5% dextrose with electrolytes and intravenous carnitine at 300 mg/kg (the highest administered dose reported in the US intravenous label [10]), ondansetron or granisetron for vomiting, carbohydrate provision of at least 1.5 times basal metabolism and gastrostomy placed at diagnosis; this is historical institutional practice, more intensive than current recommendations [13]. In rhabdomyolysis due to severe dystonia with creatine kinase above 5000 U/l, GeneReviews recommends fluids at 3 l/m² body surface area per day for renal protection [1]. No evidence supports arginine during acute illness [1].

After 6 years and in adults. Acute encephalopathic crises have not been described after 6 years, but subclinical injury cannot be excluded; recommendation #11 (level 0) permits emergency treatment during severe illness or surgery, with age-adjusted glucose: 6–8 g/kg/24 h (4–6 mg/kg/min) from 7 to 10 years, 4–7 g/kg/24 h (3–5 mg/kg/min) from 11 to 15 years and 3–5 g/kg/24 h (2–4 mg/kg/min) from 16 years [3]. The British adult protocol recalls that acute decompensation is uncommon and that another cause should be sought; it provides 10% glucose at 2 ml/kg/h, carnitine 50–100 mg/kg/day (approximately 1 g three times daily) orally or intravenously, with no maximum dose defined in the protocol (US label for the adult oral solution: 1–3 g/day), insulin if blood glucose exceeds 10 mmol/l, protein reintroduction within 24 hours and attention to refeeding syndrome [60]. Effectiveness in adolescents and adults rests only on case reports [1, 3]. The French PNDS allows, after 6 years, outpatient treatment after discussion with the centre if fever is well tolerated, the diet well accepted, general condition preserved and there is no vomiting, diarrhoea or acute neurological signs, with frequent reassessment by the family; the emergency diet removes foods of animal origin, carnitine is increased to 100 mg/kg/day (maximum 12 g/day) and, in a child without improvement within 12 hours or with oral intolerance, glucose and lipid infusion with intravenous carnitine at the same dose is started; adults in principle do not need admission, but dystonia and myoclonus may worsen during infections or prolonged fasting [11]. The G2M network emergency certificate, appended to the PNDS, summarises these measures for the local hospital [11].

7.5. Interventions without demonstrated benefit

Riboflavin, the enzyme cofactor, has been used at 100–300 mg/day but does not improve clinical outcome, there is no protocol for testing responsiveness and no predictive genotype, and it may cause nausea and abdominal pain; it is not recommended as standard therapy [1, 3, 13]. The French PNDS takes the same position: the biochemical reduction of GA and 3-OH-GA observed in some patients does not translate into clinical improvement [11]. Phenobarbital, topiramate, carbamazepine, creatine, glutamate antagonists (such as dextromethorphan), N-acetylcysteine and antioxidants used with neuroprotective intent have shown no benefit [1, 3]. Loading tests and prolonged fasting are dangerous [3].

7.6. Movement disorder treatment

Dystonia is difficult to treat and individual drug effectiveness unpredictable; recommendations issue no specific recommendation and indicate dosing should follow general guidance [3, 6]. None of the following medicines has a specifically approved GA1 indication, so use is extrapolated from dystonia of other causes.

Dystonia severity may be assessed using the Barry–Albright scale, which underestimates severity in infants with marked truncal hypotonia, or the Fahn–Marsden scale, not validated for GA1 [3]. The French PNDS adds the Gross Motor Function Classification System (GMFCS), validated and translated, which does not measure dystonia but classifies global motor disability [11]. The PNDS also adds practical data: oral baclofen, together with benzodiazepines, is the most widely used treatment and should be used at usual doses, because at high doses sedation predominates; in a series of over 100 children, benzodiazepines had benefit exceeding adverse effects, at the cost of dependence and high doses (over 1.5 mg/kg/day); intraventricular baclofen by pump, in a report of two patients, abolished dystonic storms and allowed other medicines to be reduced in a 10-year-old child, with a lesser effect in a 23-year-old adult; in France, the botulinum toxins authorised in children are onabotulinumtoxinA and abobotulinumtoxinA, with incobotulinumtoxinA added in adults, and repetition of injections depends on response and on the injector’s experience; gabapentin has no published data in GA1; and four patients treated with deep brain stimulation have been described, with some positive results [11]. Regarding epilepsy, the PNDS notes that seizures may be the first or only manifestation, that infantile spasms with hypsarrhythmia without encephalopathy have been described and that dystonic movements may be mistaken for seizures [11]. Epilepsy is treated according to semiology and electroencephalography, avoiding valproate and vigabatrin [1, 3]. Physiotherapy, occupational therapy, speech and language therapy, augmentative and alternative communication, positioning devices and orthopaedic surgery form part of supportive treatment [1, 3].

7.7. Subdural haematoma, surgery and anaesthesia

The few reports of neurosurgery for arachnoid cysts or subdural haematomas generally had poor neurological outcomes, and surgery in undiagnosed patients increases the risk of encephalopathic crisis; any intervention requires perioperative management according to recommendations #10 and #11, under metabolic centre supervision [3]. A Portuguese case published in 2025 described a screen-detected infant with macrocephaly from birth who at 8 months had bilateral subdural haematomas with mass effect, without signs of intracranial hypertension or history of trauma; after multidisciplinary discussion, conservative treatment with surveillance was chosen, with clinical stability and reduction of the haematomas [64]. Before surgery and procedures, including dental procedures, the metabolic centre should be contacted to plan fasting, fluids and nutrition; in high-risk individuals a prior anaesthetic consultation may be warranted [1, 3]. The French PNDS specifies that fasting should be avoided and that any surgery requires, from the start of preoperative fasting, an infusion of 10% glucose or a glucose-lipid infusion according to the patient’s emergency protocol, in liaison with the reference or competence centre [11].

7.8. Vaccination and trigger prevention

All children should receive national schedule vaccinations (recommendation #13, level B): gastroenteritis, pneumonia and meningitis are leading crisis triggers, and the recommendations group has observed no vaccination complications since screening introduction; postvaccination febrile reactions are treated with the emergency protocol [3]. Excess protein and protein malnutrition, prolonged fasting and insufficient caloric intake under stress should be avoided; although unsupported by data, GeneReviews considers avoiding contact sports or practising them with extreme caution prudent [1].

7.9. Effectiveness evidence: main studies

No randomised GA1 treatment trials exist. Evidence comes from prospective cohorts, comparative historical series and a meta-analysis.

These studies have limitations preventing simple interpretation. Compared groups differ in period, centre and diet type, and historical controls have ascertainment bias acknowledged by authors [65]. The meta-analysis proportion of 25.3% screened patients with movement disorder includes cohorts with non-compliant treatment, whereas 7% refers to guideline-treated patients in specific cohorts; these estimates are not interchangeable [8, 14]. Effect estimates for maintenance and emergency adherence derive from few “informative” studies comparing two treatments within one cohort, with wide confidence intervals [14, 23]. Oji-Cree experience, with a low-excretion variant and different protocol, cannot be directly compared with European cohorts [14, 18]. Finally, absence of crises after 6 years in cohorts does not prove zero subclinical injury risk, as recommendations emphasise [3].

7.10. Research and developments in the last 12 months

In the ClinicalTrials.gov search performed, this is the only GA1-directed interventional trial; no specifically GA1-authorised medicine was identified in the European Medicines Agency list, nor results announcements, authorisation applications or regulatory opinions. Negative searching does not exclude unregistered programmes [26, 55]. The comprehensive June 2026 GeneReviews revision added no new targeted therapies [1].

7.11. French PNDS and other guidelines: divergences

Table 7.11 — Divergences between the French PNDS (2021) and the other guidance cited [1, 3, 9, 11, 58, 59, 63]

8. Reproductive Options and Family Counselling

GA1 is autosomal recessive. Parents of a patient are presumed heterozygous; parental molecular analysis is recommended to confirm heterozygosity and estimate recurrence risk. If only one parent is a carrier, with maternity and paternity confirmed, a de novo proband variant or parental postzygotic mosaicism should be considered; if apparently homozygous, an exon deletion undetected by sequencing and uniparental isodisomy should be excluded [1]. When both parents are heterozygous, each sibling has at conception a 25% chance of being affected, 50% an asymptomatic carrier and 25% neither affected nor carrier. Heterozygotes do not develop disease [1]. Phenotype may vary greatly between untreated siblings with the same genotype, so severity in one child does not predict the next [1, 13]. Offspring of a patient are obligate carriers unless the other parent is a carrier or affected; each sibling of the proband’s parents has a 50% carrier probability [1].

Carrier detection requires prior identification of familial variants; fibroblast or leukocyte enzyme assays are unhelpful because controls and obligate heterozygotes overlap [1, 13]. Carrier testing should be considered for reproductive partners of known carriers and patients, particularly with shared ancestry and especially in founder communities (Manitoba and Ontario Oji-Cree, South African Xhosa, Pennsylvania Amish, North Carolina Lumbee and Irish Travellers) [1]. Heterozygosity for c.553_570del (p.Gly185_Ser190del), with a suggested dominant-negative effect, was associated with positive screening and 10–20% activity without phenotype; whether these individuals require treatment is unknown [1, 3].

Ideally, genetic risk and options are discussed before pregnancy. Once the familial variants are identified, prenatal diagnosis by amniocentesis or chorionic villus sampling and preimplantation genetic testing are possible; prenatal biochemical diagnosis is not recommended [1]. The French PNDS, which requires a dedicated genetic counselling consultation, considers molecular prenatal diagnosis the method of choice, on chorionic villi from 11 weeks of gestation or on amniotic fluid from 15 weeks; in the absence of the familial variants, it accepts measurement of GA, 3-OH-GA and C5DC in amniotic fluid supernatant, in laboratories with sufficiently sensitive methods authorised for biochemical prenatal diagnosis, which diverges from GeneReviews; the same measurement may be requested when ultrasound or fetal MRI shows suggestive, albeit non-specific, signs (defective operculisation of the Sylvian fissures, germinolytic cysts, macrocephaly with widening of the Sylvian fissures) [11]. The Orphanet summary also mentions enzyme analysis of chorionic villi and GA in amniotic fluid, but considers genetic analysis necessary for confirmation [2]. The value of prenatal diagnosis was historically questioned because of asymptomatic homozygotes, but these had frontotemporal atrophy and were not assessed cognitively [13]. An important objective of prenatal diagnosis or newborn testing is to start treatment immediately after birth: without prenatal diagnosis, a newborn with an affected sibling should be treated from birth until confirmation by enzyme assay or molecular study, and older siblings should be tested even if asymptomatic [1]. DNA banking should be considered when the molecular diagnosis has not been confirmed [1].

Pregnancy in women with GA1. No formal recommendations exist. GeneReviews and the 2023 recommendations advise care by a team including an obstetrician, a metabolic physician and a specialist dietitian. The peripartum period involves high catabolic stress, and most specialists consider emergency treatment and close surveillance necessary, but there are no data on their effectiveness or necessity: two women treated peripartum had uncomplicated courses, as did untreated women [1, 3]. In Portugal, Garcia and colleagues described three children born to two mothers with untreated GA1: in one, isolated hypocarnitinaemia on her child’s newborn screen led to the maternal diagnosis; the newborns’ biochemical abnormalities normalised within a week; at 4 months, imaging showed widened Sylvian fissures in both infants and bilateral temporal arachnoid cysts in one, but physical and neurological development was normal at 2 and 5 years [43]. The US carnitine label permits use in pregnancy only if clearly necessary [9]. The French PNDS, acknowledging that there are no European recommendations for pregnancy, recommends managing the peripartum period as an at-risk situation, with an intensive intravenous emergency diet (high glucose-lipid provision), restriction of natural protein for 24–48 hours and carnitine, and does not contraindicate breastfeeding by women with GA1, provided their nutritional intake is adapted [11].

Disclosure of the diagnosis. The French PNDS asks that the diagnosis be disclosed as early as possible, by an experienced physician, in a dedicated consultation, to both parents whenever possible, in a quiet place and avoiding the eves of weekends and holidays, with subsequent contact with the general practitioner; psychological support should be offered from the time of disclosure, together with information on patient associations and social support, follow-up interviews should be scheduled soon to confirm understanding, the disease should later be explained to the child in simple words, and siblings should not be forgotten [11].

Newborn screening counselling. A positive screen with negative child confirmation may reveal maternal GA1; in France, expanded screening in 2023–2025 led to diagnosis of an asymptomatic homozygous high-excretor mother through low newborn free carnitine [3, 44].

9. Special Considerations: Population Diversity and Epidemiology

9.1. General population frequency

Estimates depend on the source. The 2023 recommendations give a global incidence of 1:90,000 to 1:120,000 newborns and over 600 published patients; Saudubray cites a mean frequency of approximately 1:125,000 based on screening programmes; the Atlas gives 1:106,900 from screening in North America, Australia, Germany and Taiwan; GeneReviews presents estimates between 1:30,000 and 1:100,000–110,000 [1, 3, 6, 13]. The Orphanet summary estimates 1 per 100,000 births worldwide, and Orphanet epidemiological data incorporated in GeneGuidance record validated birth prevalences of 2.85/100 000 in Spain, 2.0 in Israel, 0.76 in the United States, 0.56 in the Czech Republic and 0.48 in Japan [2, 4]. The French PNDS uses a global incidence of approximately 1:110,000, higher in some communities in the United States, Canada, Ireland, the United Kingdom and South Africa, and reports approximately 250 pathogenic variants described [11]. These figures are birth prevalence estimates from different programmes and periods, and should not be read as current population prevalences.

9.2. Founder populations

9.3. Europe, United States and other contexts

In Germany, the prospective newborn screening cohort (1999–2016) is the main source for sensitivity, outcome and cognition data [8, 24]. In Spain, the 43-patient cohort defined two distinct genetic and biochemical groups [29]. In Ireland, 21 patients diagnosed over 16 years (11 clinically, 10 through high-risk screening) showed no striatal degeneration over 14 years in the screened group [69], and the 2022 cohort of 34 confirmed the clinical versus presymptomatic diagnosis contrast [66]. In the Netherlands, the national 25-patient registry provided the 4% subdural haematoma incidence [35]. In Austria, Bavaria and Switzerland, GA1 screening began respectively in 2004, 1999 and November 2014 [46]. In the United States, the 168-patient single-centre Pennsylvania cohort, including genetically diverse and Amish patients, is the largest single-centre series [52]. The Turkish 50-patient cohort illustrates outcomes in countries without screening and with high consanguinity [33].

Recommendations acknowledge that the development group’s composition reflects developed-country healthcare systems and applicability may be limited in countries with restricted access to required care [3]. The 2021 meta-analysis examined “migration background” (at least one parent born in a country with a different language) as a non-interventional variable, because language barriers might affect education and adherence; in the 2010 study, this factor had no significant effect [14, 23]. The recommendations-based parent guide is translated into several languages, including Portuguese [3].

9.4. Portugal and Brazil

In Portugal, the national expanded MS/MS newborn screening programme, begun in 2004 at the national laboratory, screened 316,243 newborns over four years with 99.8% coverage and detected eight maternal diseases, three GA1 [70]. Garcia and colleagues’ untreated maternal GA1 study and the conservatively treated subdural haematoma case in a screened infant are Portuguese contributions cited here [43, 64]. Sources consulted provide no Portuguese GA1 incidence estimate.

In Brazil, the Sociedade Brasileira de Triagem Neonatal e Erros Inatos do Metabolismo (SBTEIM) publishes an acute decompensation emergency letter based on 2017 recommendations [59]. Sources consulted include neither Brazilian epidemiology nor information on public newborn screening inclusion of GA1; absence from these sources does not demonstrate absence of published data.

10. Illustrative Clinical Case

10.1. Four cases missed by newborn screening (Austria, Germany and Switzerland)

Spenger and colleagues described in 2021 four children from three families in whom correctly performed newborn screening missed GA1 [46]. The first, born to non-consanguineous Swiss parents, had normal screening, development and head circumference (46 cm, 63rd percentile at 9 months). At that age he was admitted with rotavirus fever, vomiting, diarrhoea and mild dehydration and several brief clonic seizures; he was initially given low-dose carbamazepine for gastroenteritis-associated seizures. The next day an encephalopathic crisis developed with hypotonia, drowsiness and repeated dystonic posturing, and MRI showed bilateral caudate, putaminal and pallidal abnormalities on T2, diffusion and apparent diffusion coefficient mapping (article Figure 1). Urinary GA was 26 and 3-OH-GA 22 mmol/mol creatinine (reference 0–20 and 0–10), and dried blood spot C5DC had been 0.5 µmol/l on day 4 (cut-off < 0.53). Variants p.Gly241Val (then undescribed) and p.Gly390Ala were identified. The child developed dyskinetic cerebral palsy (GMFCS level 5), without tube dependence and with good social contact at 5 years [46].

Patients 2 and 3 were siblings born to non-consanguineous Moroccan parents, compound heterozygous for p.Arg257Gln (incorrectly designated p.Arg257Glu in the original publication and corrected in 2022) and p.Met405Val, the latter associated with low excretion and more frequent in patients of African origin [46, 71]. The boy was seen at 2 years 4 months for non-progressive ataxia and developmental delay; MRI at 2 years 8 months showed mild laterodorsal hyperintensity around the putamen (article Figure 2), not then recognised as an insidious pattern, and organic acids and amino acids were normal. The diagnosis was made only at 10 years, when his sister was assessed; at 14 he attended mainstream school with subtle dystonia. His sister, with normal screening, was referred at 18 months for motor delay and at 4 years for dystonic movements, with the typical insidious pattern on MRI; after diagnosis, an infection with poor intake worsened the movement disorder, and at 8 years she had evident dystonia and learning difficulties. Adherence to diet and carnitine was uncertain in both siblings [46].

The fourth patient, of German parents, had birth head circumference at the 93rd percentile. Screening at 47 hours showed mildly elevated C5DC (0.78 µmol/l; cut-off < 0.45) with two of three ratios elevated, but repetition on day 8 was normal (0.38). Motor delay was noted at 10 months; at 16 months she had hypotonia, unstable posture and mild dystonia. At 21 months urinary 3-OH-GA was elevated (8.66 mmol/mol; reference 0.17–1.17) and plasma C5DC mildly increased. Variants p.Arg257Trp and p.Lys170Asn, the latter novel, were found, with 7% residual leukocyte activity. With lysine-restricted diet (80 mg/kg) and carnitine (100 mg/kg), she progressed at 26 months without cognitive impairment; phenotype was classified as insidious onset [46].

10.2. Lessons from the case

All four were low excretors and all had one severe allele; three had a known mild variant (4–25% residual activity) on the other allele, and in the first the variant was novel. Authors concluded that normal C5DC does not exclude GA1, normalisation after a borderline result also does not exclude it, low excretors may have normal C5DC even with carnitine, and 3-OH-GA was the best marker, at least mildly elevated on almost every occasion [46]. Following these cases, the Swiss laboratory deliberately lowered its C5DC cut-off to 0.35 µmol/l, Austria adjusted its values and Germany began referring all positive results directly to a metabolic centre [46]. The first patient also illustrates crisis presentation as “seizures with gastroenteritis”, with carbamazepine initiation, a drug recommendations consider ineffective for GA1 dystonia [3, 46]. These actions describe the report rather than routine recommendations; current recommendations advocate full investigation of any positive screen and targeted investigation of suggestive signs even with normal screening [3].

11. Prognosis and Long-Term Follow-up

11.1. Prognosis

In natural history, early death affects 40–50% of symptomatic patients by 20 years, through pneumonia, respiratory failure, hyperpyrexic crises or suddenly [6]. In the international study, 49 deceased patients had median age 79 months (5–490 months), estimated 50% survival time was 25 years, aspiration pneumonia was the most common cause, and earlier first crisis and imaging basal ganglia degeneration were associated with poorer outcome [13]. In the meta-analysis, pooled mortality was 0.2% in screened and 6.4% in symptom-diagnosed patients without significant difference; nine screened patients died, seven with severe movement disorder [14]. In the unscreened Turkish cohort, 24% died during follow-up, at median 6.2 years [33].

With early diagnosis and treatment adherence, 70–90% remain largely asymptomatic; GeneReviews estimates approximately 25% of treated patients develop manifestations, with greater risk when carers do not adhere to treatment [1]. Patients missed by screening have outcomes similar to the prescreening era because post-injury treatment does not reverse damage, although some benefit from preventing additional deterioration [3, 6]. Long-term uncertainties remain regarding white matter progression, chronic kidney disease, high-excretor cognitive impairment, peripheral neuropathy and brain tumours [3, 24, 25].

11.2. Surveillance

GeneReviews proposes plasma amino acids at least quarterly in the first year, six-monthly from 1 to 6 years and annually thereafter [1]. The French PNDS sets medical and dietetic consultations at least every 3 months in the first year after diagnosis, every 6 months up to 6 years and annually thereafter, including in adults, with plasma amino acids on the same schedule (table taken from the 2017 recommendations), and makes no provision for systematic neuroimaging in the absence of clinical change [11]. At each consultation it reviews the history (infections, vaccinations, admissions, diet), growth with head circumference, psychomotor and language development, the neurological examination including assessment for peripheral neuropathy, signs of protein deficiency and malnutrition, adherence and renal function; it checks the emergency card, home supplies and product expiry dates, updates the emergency protocol to the current weight and the individualised school reception plan, and assesses quality of life [11]. Depending on needs, gastroenterology (swallowing, enteral nutrition, reflux, gastrostomy or jejunostomy), neurology (development, psychometric testing at key ages, dystonia, epilepsy, central or peripheral injury in adults), otorhinolaryngology (swallowing, orofacial dyskinesias, botulinum toxin, antisecretory patches), physical and rehabilitation medicine, nephrology, psychology and social work are involved [11]. For patients with severe movement disorder, the recommendations suggest Raven’s Progressive Matrices or the Vineland Adaptive Behavior Scales [3]. Urinary GA and 3-OH-GA (recommendation #16) and C5DC should not be used to monitor or adjust treatment because they do not correlate with the clinical course [1, 3]. Plasma lysine correlates poorly with intake, but the amino acid profile helps detect dietary deviations [1, 3].

During the first three years, even minor or mild head trauma warrants admission and surveillance for at least 24 hours (recommendation #20) because of subdural haematoma risk [3]. Diffusion MRI detects striatal lesions earlier and more accurately than computed tomography; minimum requirements include axial T2, FLAIR, T1 and diffusion with ADC maps, and a 3D sequence when seeking subependymal nodules; frontotemporal hypoplasia is also visible by transfontanellar or prenatal ultrasound [3]. During the vulnerability window, no evidence supports serial MRI without clinical indication [3].

11.3. Transition, education and resources

Regular age-adapted education on course, treatment and prognosis, with social and legal counselling and interdisciplinary quality-of-life assessment, is recommendation #14 [3]. Preparation for transition to adult medicine should start from 14 years as a structured interdisciplinary process with joint paediatric and internal medicine visits (recommendation #24); if adult specialists are unavailable, follow-up should continue at the paediatric centre. Adherence may deteriorate in puberty and early adulthood [1, 3]. Adult follow-up should maintain adherence, detect specific complications and address age-related issues such as metabolic syndrome, musculoskeletal disease, fertility and family planning [3].

In Europe, the relevant reference network is MetabERN (European Reference Network for Hereditary Metabolic Disorders), in which GA1 falls within the amino acid and organic acid disorders (AOA) subnetwork [12]. The 2023 recommendations were endorsed by the E-IMD consortium (European Network and Registry for Intoxication Type Metabolic Diseases) and involved the German patient association Glutarazidurie e.V. [3]. The parent guide accompanying the recommendations is available in several languages, including Portuguese [3]. GeneReviews lists the Organic Acidemia Association and its natural history registry, the British Inherited Metabolic Disease Group and the TEMPLE family resource, MedlinePlus and the HRSA state newborn screening page; in the United States, the Rare Organic Acidemias Research Consortium coordinates clinical and basic research [1]. In Brazil, SBTEIM provides the emergency letter [59].

In France, the 2021 PNDS organises the care pathway around the reference and competence centres for inherited metabolic diseases of the G2M network, with long-term condition (ALD) no. 17 requested for all patients, an application to the Maison Départementale des Personnes Handicapées for support for disabled children and adults, an individualised school reception plan (diet and risk of decompensation), school support or specialised education depending on the neurological injury and, in adults, access to sheltered employment; the patient association is AG1-2.3 Soleil [11]. For travel, the patient carries the translated emergency certificate, the updated protocol is sent to the centre near the holiday destination, the centre issues certificates for carrying medicines and dietetic products in the aircraft cabin, and the family takes what is needed for the emergency diet; holiday camps are assessed with the metabolic team [11]. Therapeutic education covers the cruise diet, risk factors and signs of decompensation, the need for admission before 6 years, the conditions for starting emergency treatment at home after 6 years (well-tolerated fever above 38 °C, food accepted, no warning signs), systematic contact with the team for any intercurrent event, preparation of the emergency diet and use of a tube or gastrostomy, with regular assessment of knowledge [11]. Transition should be planned during a period of metabolic stability, with joint consultation models (adult physician, paediatrician, dietitian, psychologist and social worker), and supervision by a specialised centre continues lifelong, because the long-term prognosis remains poorly characterised [11]. When life is at risk during an encephalopathic crisis or in the context of multiple disability with severe dystonia, the PNDS provides for palliative care and end-of-life support, decided jointly by the paediatric, palliative care and, if necessary, ethics teams, together with the family [11].

12. Summary and Final Considerations

Glutaric acidaemia type 1 results from biallelic glutaryl-CoA dehydrogenase deficiency, the shared enzyme of lysine, hydroxylysine and tryptophan degradation. Intracerebral accumulation of GA, 3-OH-GA and glutaryl-CoA, trapped by the blood–brain barrier, renders the striatum vulnerable in early life. Untreated, 80–90% sustain irreversible striatal injury, almost always before 3 years and exceptionally up to 6, triggered by fever, infection, fasting or surgery, or insidiously, with permanent dystonia and high mortality [3, 7].

C5DC newborn screening, confirmation through quantitative 3-OH-GA and GCDH analysis, low-lysine diet with specific amino acid mixture, carnitine and, above all, early aggressive emergency treatment have reduced movement disorder to approximately 7% in guideline-treated patients. Delayed emergency treatment remains associated with acute injury and dietary deviations with insidious injury [8, 14, 23]. Low excretors may escape screening and have almost normal metabolites, but their striatal injury risk equals high excretors’; clinical suspicion requires continued investigation even with normal screening [3, 46].

Current treatment does not resolve all issues. High excretors accumulate more extrastriatal abnormalities and have poorer cognitive performance and greater peripheral nerve involvement, chronic kidney disease progresses independently of treatment quality, and maintaining dietary restriction after 6 years remains debated [8, 24, 25, 52]. Substrate reduction by AASS inhibition and AAV gene therapy have shown effects in animal models, and a phase 1 intracerebroventricular gene therapy trial is ongoing without results [26, 27, 67, 68]. For geneticists, central tasks remain rigorous diagnosis, interpreting test limitations, counselling families and founder communities, and ensuring every child has a written emergency plan and a team available around the clock.

13. Bibliographic References

1. Larson A, Coughlin C. Glutaric Acidemia Type 1. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Published 19 Sep 2019; comprehensive update 4 Jun 2026. PMID:31536184. NBK546575.
2. Orphanet. Glutaryl-CoA dehydrogenase deficiency. ORPHA:25 [Internet]. Summary updated April 2013. Disease entry. Accessed 7 Oct 2026.
3. Boy N, Mühlhausen C, Maier EM, Ballhausen D, Baumgartner MR, Beblo S, et al. Recommendations for diagnosing and managing individuals with glutaric aciduria type 1: Third revision. J Inherit Metab Dis. 2023;46(3):482–519. doi:10.1002/jimd.12566. PMID:36221165.
4. GeneGuidance. GCDH — identifiers, clinical curation and associated conditions (Glutaricaciduria, type I; ORPHA:25) [local database]. Version dated 26 Sep 2026. Incorporates Orphanet birth prevalence data and GeneReviews NBK546575.
5. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: Accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61–8. doi:10.1002/humu.23348. Service accessed 7 Oct 2026 (GCDH, NM_000159.3 and NM_000159.4, GRCh38) at rest.variantvalidator.org.
6. Kölker S, Hoffmann GF. Cerebral Organic Acid Disorders and Other Disorders of Lysine Catabolism. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 22, sections 22.3, 22.4 (22.4.1–22.4.5, Table 22.1) and 22.6. doi:10.1007/978-3-662-63123-2_22.
7. Kölker S, Garbade SF, Greenberg CR, Leonard JV, Saudubray JM, Ribes A, et al. Natural history, outcome, and treatment efficacy in children and adults with glutaryl-CoA dehydrogenase deficiency. Pediatr Res. 2006;59(6):840–7. doi:10.1203/01.pdr.0000219387.79887.86. PMID:16641220.
8. Boy N, Mengler K, Thimm E, Schiergens KA, Marquardt T, Weinhold N, et al. Newborn screening: a disease-changing intervention for glutaric aciduria type 1. Ann Neurol. 2018;83(5):970–9. doi:10.1002/ana.25233. PMID:29665094.
9. Leadiant Biosciences. CARNITOR® (levocarnitine) tablets, oral solution; CARNITOR® SF sugar-free oral solution. FDA label (DailyMed setid d2133bc3-9c15-48bd-8b16-b8995a6a14cd, version dated 30 Oct 2024). DailyMed. Accessed 7 Oct 2026.
10. Leadiant Biosciences. CARNITOR® (levocarnitine) injection. FDA label (DailyMed setid cf801cc4-775e-433d-9d32-e5d9a98981d3, version dated 20 Nov 2023). DailyMed. Accessed 7 Oct 2026.
11. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Acidurie glutarique type 1. Filière de santé maladies rares G2M; Centre de référence des maladies héréditaires du métabolisme (Marseille). Saint-Denis La Plaine: HAS; April 2021 (published 23 Apr 2021). Available from: https://www.has-sante.fr/jcms/p_3262535
12. MetabERN — European Reference Network for Hereditary Metabolic Disorders. Subnetworks (including aminoacid and organic acids related disorders, AOA) [Internet]. metab.ern-net.eu. Accessed 7 Oct 2026.
13. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. Chapter 9, Glutaric aciduria (type I), p. 70–81. Local chapter.
14. Boy N, Mengler K, Heringer-Seifert J, Hoffmann GF, Garbade SF, Kölker S. Impact of newborn screening and quality of therapy on the neurological outcome in glutaric aciduria type 1: a meta-analysis. Genet Med. 2021;23(1):13–21. doi:10.1038/s41436-020-00971-4. PMID:32981931. PMCID:PMC7790745.
15. Goodman SI, Markey SP, Moe PG, Miles BS, Teng CC. Glutaric aciduria; a "new" disorder of amino acid metabolism. Biochem Med. 1975;12(1):12–21. doi:10.1016/0006-2944(75)90091-5. PMID:1137568.
16. Goodman SI, Kohlhoff JG. Glutaric aciduria: inherited deficiency of glutaryl-CoA dehydrogenase activity. Biochem Med. 1975;13(2):138–40. doi:10.1016/0006-2944(75)90149-0. PMID:1191271.
17. Morton DH, Bennett MJ, Seargeant LE, Nichter CA, Kelley RI. Glutaric aciduria type I: a common cause of episodic encephalopathy and spastic paralysis in the Amish of Lancaster County, Pennsylvania. Am J Med Genet. 1991;41(1):89–95. doi:10.1002/ajmg.1320410122. PMID:1951469.
18. Greenberg CR, Prasad AN, Dilling LA, Thompson JR, Haworth JC, Martin B, et al. Outcome of the first 3-years of a DNA-based neonatal screening program for glutaric acidemia type 1 in Manitoba and northwestern Ontario, Canada. Mol Genet Metab. 2002;75(1):70–8. doi:10.1006/mgme.2001.3270. PMID:11825066.
19. Strauss KA, Puffenberger EG, Robinson DL, Morton DH. Type I glutaric aciduria, part 1: natural history of 77 patients. Am J Med Genet C Semin Med Genet. 2003;121C(1):38–52. doi:10.1002/ajmg.c.20007. PMID:12888985.
20. Kölker S, Christensen E, Leonard JV, Greenberg CR, Burlina AB, Burlina AP, et al. Guideline for the diagnosis and management of glutaryl-CoA dehydrogenase deficiency (glutaric aciduria type I). J Inherit Metab Dis. 2007;30(1):5–22. doi:10.1007/s10545-006-0451-4. PMID:17203377.
21. Kölker S, Christensen E, Leonard JV, Greenberg CR, Boneh A, Burlina AB, et al. Diagnosis and management of glutaric aciduria type I—revised recommendations. J Inherit Metab Dis. 2011;34(3):677–94. doi:10.1007/s10545-011-9289-5. PMID:21431622.
22. Boy N, Mühlhausen C, Maier EM, Heringer J, Assmann B, Burgard P, et al. Proposed recommendations for diagnosing and managing individuals with glutaric aciduria type I: second revision. J Inherit Metab Dis. 2017;40(1):75–101. doi:10.1007/s10545-016-9999-9. PMID:27853989.
23. Heringer J, Boy SP, Ensenauer R, Assmann B, Zschocke J, Harting I, et al. Use of guidelines improves the neurological outcome in glutaric aciduria type I. Ann Neurol. 2010;68(5):743–52. doi:10.1002/ana.22095. PMID:21031586.
24. Märtner EMC, Thimm E, Guder P, Schiergens KA, Rutsch F, Roloff S, et al. The biochemical subtype is a predictor for cognitive function in glutaric aciduria type 1: a national prospective follow-up study. Sci Rep. 2021;11(1):19300. doi:10.1038/s41598-021-98809-9. PMID:34588557.
25. Preisner F, Garbade SF, Harting I, Wortmann SB, Mühlhausen C, Heiland S, et al. Peripheral neuropathy expands the neurological phenotype in glutaric aciduria type 1. J Inherit Metab Dis. 2026;49(1):e70131. doi:10.1002/jimd.70131. PMID:41485880.
26. Shanghai Vitalgen BioPharma. An open-label, dose-escalation and dose-expansion phase I clinical study to evaluate the tolerability, safety and efficacy of VGM-R02b in patients with glutaric acidemia type I. ClinicalTrials.gov NCT06217861. Last updated 13 May 2026. Registry. Accessed 7 Oct 2026.
27. Segur-Bailach E, Mateu-Bosch A, Bofill-De Ros X, Parés M, da Silva Buttkus P, Rathkolb B, et al. Therapeutic AASS inhibition by AAV-miRNA rescues glutaric aciduria type I severe phenotype in mice. Mol Ther. 2025;33(10):4820–33. doi:10.1016/j.ymthe.2025.07.022. PMID:40682274.
28. Richert AC, Zhang Y, Bharathi SS, Hernandez A, Dodatko T, Bons J, et al. Odd-chain dicarboxylic acid feeding produces a glutaric aciduria type 1-like metabolic signature in mice. FASEB J. 2025;39(21):e71215. doi:10.1096/fj.202502381R. PMID:41196119.
29. Busquets C, Merinero B, Christensen E, Gelpí JL, Campistol J, Pineda M, et al. Glutaryl-CoA dehydrogenase deficiency in Spain: evidence of two groups of patients, genetically, and biochemically distinct. Pediatr Res. 2000;48(3):315–22. doi:10.1203/00006450-200009000-00009. PMID:10960496.
30. Christensen E, Ribes A, Merinero B, Zschocke J. Correlation of genotype and phenotype in glutaryl-CoA dehydrogenase deficiency. J Inherit Metab Dis. 2004;27(6):861–8. doi:10.1023/B:BOLI.0000045770.93429.3c. PMID:15505393.
31. van der Watt G, Owen EP, Berman P, Meldau S, Watermeyer N, Olpin SE, et al. Glutaric aciduria type 1 in South Africa—high incidence of glutaryl-CoA dehydrogenase deficiency in black South Africans. Mol Genet Metab. 2010;101(2–3):178–82. doi:10.1016/j.ymgme.2010.07.018. PMID:20732827.
32. Basinger AA, Booker JK, Frazier DM, Koeberl DD, Sullivan JA, Muenzer J. Glutaric acidemia type 1 in patients of Lumbee heritage from North Carolina. Mol Genet Metab. 2006;88(1):90–2. doi:10.1016/j.ymgme.2005.12.008. PMID:16466958.
33. Selamioğlu A, İdrisoğlu M, Kılıç Ş, Güneş D, Karaca M, Balcı MC, et al. Glutaric acidemia type 1 in a non-newborn-screened cohort: clinical, biochemical, and molecular features and neurologic outcomes. Ital J Pediatr. 2026;52(1):117. doi:10.1186/s13052-026-02284-9. PMID:42210298.
34. Boy N, Garbade SF, Heringer J, Seitz A, Kölker S, Harting I. Patterns, evolution, and severity of striatal injury in insidious- vs acute-onset glutaric aciduria type 1. J Inherit Metab Dis. 2019;42(1):117–27. doi:10.1002/jimd.12033. PMID:30740735.
35. Vester ME, Visser G, Wijburg FA, van Spronsen FJ, Williams M, van Rijn RR. Occurrence of subdural hematomas in Dutch glutaric aciduria type 1 patients. Eur J Pediatr. 2016;175(7):1001–6. doi:10.1007/s00431-016-2734-6. PMID:27246831.
36. Vester ME, Bilo RA, Karst WA, Daams JG, Duijst WL, van Rijn RR. Subdural hematomas: glutaric aciduria type 1 or abusive head trauma? A systematic review. Forensic Sci Med Pathol. 2015;11(3):405–15. doi:10.1007/s12024-015-9698-0. PMID:26219480.
37. Boy N, Mohr A, Garbade SF, Freisinger P, Heringer-Seifert J, Seitz A, et al. Subdural hematoma in glutaric aciduria type 1: high excreters are prone to incidental SDH despite newborn screening. J Inherit Metab Dis. 2021;44(6):1343–52. doi:10.1002/jimd.12436. PMID:34515344.
38. Boy N, Heringer J, Brackmann R, Bodamer O, Seitz A, Kölker S, Harting I. Extrastriatal changes in patients with late-onset glutaric aciduria type I highlight the risk of long-term neurotoxicity. Orphanet J Rare Dis. 2017;12(1):77. doi:10.1186/s13023-017-0612-6. PMID:28438223.
39. Lokar M, Georgiev KJ, Rus T, Gregorič N, Peterlin B, Pušenjak MŠ, Georgiev D. Glutaric aciduria type 1 presenting with Parkinsonism with presynaptic dopaminergic dysfunction in an adult female: a case report. Neurogenetics. 2026;27(1):70. doi:10.1007/s10048-026-00940-0. PMID:42742749.
40. Kölker S, Valayannopoulos V, Burlina AB, Sykut-Cegielska J, Wijburg FA, Teles EL, et al. The phenotypic spectrum of organic acidurias and urea cycle disorders. Part 2: the evolving clinical phenotype. J Inherit Metab Dis. 2015;38(6):1059–74. doi:10.1007/s10545-015-9840-x. PMID:25875216.
41. Serrano Russi A, Donoghue S, Boneh A, Manara R, Burlina AB, Burlina AP. Malignant brain tumors in patients with glutaric aciduria type I. Mol Genet Metab. 2018;125(3):276–80. doi:10.1016/j.ymgme.2018.08.006. PMID:30217722.
42. Lang SH, Shanghvi D, Divin K, Donaruma M, Sutton VR, Bachim A. Glutaric aciduria type 1 as a mimicker of abusive head trauma in the era of newborn screening. Child Abuse Negl. 2026;179:108212. doi:10.1016/j.chiabu.2026.108212. PMID:42413254.
43. Garcia P, Martins E, Diogo L, Rocha H, Marcão A, Gaspar E, et al. Outcome of three cases of untreated maternal glutaric aciduria type I. Eur J Pediatr. 2008;167(5):569–73. doi:10.1007/s00431-007-0556-2. PMID:17661081.
44. Grillet PE, Marelli C, Mondésert E, Francois-Heude MC, Roubertie A, Sabourdy F, et al. Incidental maternal glutaric aciduria type I detection through newborn screening: a case report. Mol Genet Metab Rep. 2026;46:101300. doi:10.1016/j.ymgmr.2026.101300. PMID:41783484.
45. Wang S, Meng X, Zhang X, Li L, Tang Y, Kong Y, Wu Q. Fetal cranial biometric abnormalities: the value of routine ultrasound for early diagnosis of glutaric acidemia type 1. Prenat Diagn. 2026. Advance online publication. doi:10.1002/pd.70167. PMID:42108166.
46. Spenger J, Maier EM, Wechselberger K, Bauder F, Kocher M, Sperl W, et al. Glutaric aciduria type I missed by newborn screening: report of four cases from three families. Int J Neonatal Screen. 2021;7(2):32. doi:10.3390/ijns7020032. PMID:34207159. PMCID:PMC8293111.
47. American College of Medical Genetics and Genomics. Newborn Screening ACT Sheet [Elevated C5-DC Acylcarnitine]: Glutaric Acidemia Type I (GA I) and algorithm “Glutaric Acidemia Type I (GA I): Elevated C5-DC”. Content updated: July 2022. In: Newborn Screening ACT Sheets and Algorithms. Bethesda: ACMG/NCBI Bookshelf (NBK55827). ACT Sheet; Algorithm. Accessed 7 Oct 2026.
48. American College of Medical Genetics and Genomics. Newborn Screening ACT Sheets and Algorithms [Internet]. NCBI Bookshelf NBK55827 (index of conditions and analytes). NBK55827. Accessed 7 Oct 2026.
49. Mujamammi AH, Mazi TA, AlMalki RH, Sabi EM, Al Mogren M, Alwazae M, et al. Untargeted LC-HRMS of dried blood spots reveals metabolic alterations and candidate biomarkers in glutaric aciduria type-1. Metabolites. 2026;16(3):214. doi:10.3390/metabo16030214. PMID:41893363.
50. ClinGen Actionability Working Group. Glutaric Acidemia I — Actionability Adult Summary Report (AC142). Literature search dated 19 Sep 2017; status «Released – Under Revision». Report. Accessed 7 Oct 2026.
51. Saudubray JM, García-Cazorla Á. Clinical Approach to Inborn Errors of Metabolism in Paediatrics. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1, section 1.5.5 and Table 1.22. doi:10.1007/978-3-662-63123-2_1.
52. Strauss KA, Williams KB, Carson VJ, Poskitt L, Bowser LE, Young M, et al. Glutaric acidemia type 1: treatment and outcome of 168 patients over three decades. Mol Genet Metab. 2020;131(3):325–40. doi:10.1016/j.ymgme.2020.09.007. PMID:33069577.
53. Laboratório Medinfar – Produtos Farmacêuticos, S.A. Disocor 1000 mg/10 ml solução oral (levocarnitine): summary of product characteristics. INFARMED, Infomed; SmPC approved 29/12/2023. Infomed. Accessed 26 Oct 2026.
54. U.S. Food and Drug Administration. Drugs@FDA: NDA 018948, NDA 019257 and NDA 020182 (Carnitor®, levocarnitine). Data obtained through the openFDA drugsfda API. Drugs@FDA. Accessed 7 Oct 2026.
55. European Medicines Agency. Medicines output — medicines JSON report (list of human medicines assessed by EMA). Report. Accessed 7 Oct 2026.
56. Bene Farmacêutica, Lda. Ben-u-ron 40 mg/ml xarope (paracetamol): summary of product characteristics. INFARMED, Infomed. Infomed. Accessed 26 Oct 2026.
57. Viatris Healthcare, Lda. Brufen Sem Açúcar 20 mg/ml suspensão oral (ibuprofen): summary of product characteristics. INFARMED, Infomed. Infomed. Accessed 26 Oct 2026.
58. British Inherited Metabolic Disease Group. Glutaric aciduria type 1 – acute decompensation (standard version). Emergency protocol. © BIMDG 2008; revised September 2016. PDF. Accessed 7 Oct 2026.
59. Sociedade Brasileira de Triagem Neonatal e Erros Inatos do Metabolismo. Carta emergencial em caso de descompensação aguda: acidúria glutárica tipo 1. PDF. Accessed 7 Oct 2026.
60. British Inherited Metabolic Disease Group. Adult emergency management: glutaric aciduria type 1. BIMDG_ADULT_GA1, 2015 revision, 2018 update. PDF. Accessed 7 Oct 2026.
61. Atnahs Pharma Netherlands B.V. Valium 5 mg e 10 mg comprimidos (diazepam): summary of product characteristics. INFARMED, Infomed; SmPC approved 25/02/2022. Infomed. Accessed 26 Oct 2026.
62. Roche Laboratories. VALIUM (diazepam) tablets: prescribing information, NDA 013263; SPL version of 01/12/2025. openFDA. Accessed 26 Oct 2026.
63. Slate Run Pharmaceuticals. Tetrabenazine tablets: prescribing information, ANDA 213316; SPL version of 22/10/2020. openFDA. Accessed 6 Nov 2026.
64. Marques L, Lipari Pinto P, Loureiro Cadilha H, Janeiro P, Gaspar A. Subdural hematoma in an infant with glutaric aciduria type 1: a case report on conservative management. Acta Med Port. 2025;38(8):496–501. doi:10.20344/amp.23093. PMID:40775234.
65. Kölker S, Garbade SF, Boy N, Maier EM, Meissner T, Mühlhausen C, et al. Decline of acute encephalopathic crises in children with glutaryl-CoA dehydrogenase deficiency identified by newborn screening in Germany. Pediatr Res. 2007;62(3):357–63. doi:10.1203/PDR.0b013e318137a124. PMID:17622945.
66. Healy L, O'Shea M, McNulty J, King G, Twomey E, Treacy E, et al. Glutaric aciduria type 1: diagnosis, clinical features and long-term outcome in a large cohort of 34 Irish patients. JIMD Rep. 2022;63(4):379–87. doi:10.1002/jmd2.12302. PMID:35822093.
67. Saad C, Jung-Klawitter S, Dimitrov B, Aguilar-Pimentel JA, Becker L, da Silva-Buttkus P, et al. Aminoadipate-semialdehyde synthase, a potential target for substrate reduction therapy in glutaric aciduria type 1. Sci Rep. 2026;16(1):10995. doi:10.1038/s41598-026-44377-9. PMID:41917075.
68. Kim SJ, Nam YH, Joo EY, Park J, Park S, Jung SC, Jin DK. Systemic AAV-hGCDH gene therapy alleviates glutaric acid accumulation and attenuates chronic brain vacuolation in a novel mouse model of glutaric aciduria type I. Int J Mol Sci. 2026;27(12):5569. doi:10.3390/ijms27125569. PMID:42353287.
69. Naughten ER, Mayne PD, Monavari AA, Goodman SI, Sulaiman G, Croke DT. Glutaric aciduria type I: outcome in the Republic of Ireland. J Inherit Metab Dis. 2004;27(6):917–20. doi:10.1023/B:BOLI.0000045777.82784.74. PMID:15505400.
70. Vilarinho L, Rocha H, Sousa C, Marcão A, Fonseca H, Bogas M, Osório RV. Four years of expanded newborn screening in Portugal with tandem mass spectrometry. J Inherit Metab Dis. 2010;33(Suppl 3):S133–8. doi:10.1007/s10545-010-9048-z. PMID:20177789.
71. Spenger J, Maier EM, Wechselberger K, Bauder F, Kocher M, Sperl W, et al. Correction: Spenger et al. Glutaric aciduria type I missed by newborn screening: report of four cases from three families. Int J Neonatal Screen 2021, 7, 32. Int J Neonatal Screen. 2022;8(1):2. doi:10.3390/ijns8010002. PMID:35076458. PMCID:PMC8788418.