American Staffordshire Terrier DVL2 Screw Tail: Helsinki CT Study of 285 Dogs

SamMy AmStaff’s tail has a crook near the tip. Is that the same screw tail Bulldogs have?

Elena MarshNot necessarily, and that uncertainty is why this gene matters in your breed. A 2021 Helsinki team found crooked tails in AmStaffs carrying no copy of the variant at all.

SamI may breed her someday. Would a DNA test actually change anything I decide?

Elena MarshIn an AmStaff it can, which is not true in every breed. In the same Finnish sample of 285 AmStaffs, most dogs had no copy, so there were real choices to make when picking a mate.

SamHonestly I’m more worried about her back and her breathing than about how her tail looks.

Elena MarshThat is the right worry, and the research gives a mixed but useful answer. CT scans of 19 AmStaffs of known genotype covered the spine, skull and palate, not just the tail.

Bottom line: In the American Staffordshire Terrier, the screw tail variant DVL2 c.2044delC is still segregating, so a test result can genuinely change a breeding plan. A University of Helsinki study of 285 AmStaffs found 79 carriers (27.7%) and 5 dogs with two copies (1.8%); CT scans showed every two-copy dog had malformed tail vertebrae (milder than a Bulldog screw tail) and a shorter hard palate, but no thoracic hemivertebrae. A 2026 study of US primary-care records linked two copies to cleft palate risk. Pairing two carriers gives each puppy a 25% chance of two copies. The test is risk information, not a diagnosis: hind-leg weakness, wobbling or trouble urinating or defecating means a same-day visit to your veterinarian.

Neither the Staffordshire Terrier Club of America nor the OFA’s CHIC program asks AmStaff breeders to test for this variant. Testing happens only when an owner or breeder decides to order it, often under a trade name that does not mention DVL2 at all. Labs sell it as “Screwtail”, “Screw tail”, “Robinow-like syndrome” or the “ST locus”. Below: what is known for AmStaffs, what is borrowed from Bulldog studies, and where the data run out.

Fixed in Bulldogs, still a choice in AmStaffs

The variant was identified in 2018 by a UC Davis research group. In Mansour and colleagues’ whole-genome study in PLOS Genetics, the team sequenced 100 dogs from 21 breeds and tested more than 12.4 million variants. The strongest signal anywhere in the genome (p = 4.37 x 10^-37) was a single deleted base in exon 15 of the DVL2 gene on chromosome 5. The resulting protein, p.Pro684LeufsX26, ends 23 amino acids short and carries 26 altered amino acids at its tail end. In cell experiments it responded less to WNT signaling, a pathway that shapes the developing skeleton.

The same paper genotyped 667 dogs from 49 breeds. All 79 French Bulldogs and all 33 Bulldogs had two copies, and 59 of 65 Boston Terriers did. In those breeds the variant is fixed or nearly so. Every puppy inherits it, which means no test-and-select program can breed it out without bringing in outside dogs. The Helsinki press release on the follow-up work says it plainly: in some breeds “there is no longer any variation, which makes it impossible to improve the situation with current breeding programmes”.

The AmStaff picture is different. In Niskanen and colleagues’ 2021 Human Genetics paper, from the University of Helsinki and Folkhälsan, 1,954 privately owned Finnish dogs from 15 breeds were genotyped. The 285 AmStaffs split 201 with no copy, 79 with one and 5 with two. Converted to an allele count, roughly 15.6% of the AmStaff chromosomes in that sample carried the deletion (our calculation, 89 of 570). Genotype proportions in the breed matched what random mating predicts, so there was no sign of hidden selection for or against the variant.

So the same lab result reads very differently by breed. For a Frenchie owner, “two copies” describes nearly every dog. For an AmStaff breeder, it describes fewer than 2 dogs in 100 in the only published sample.

The breed standard pushes in the same direction. The AKC standard for the American Staffordshire Terrier (approved June 10, 1936) calls for a tail “short in comparison to size, low set, tapering to a fine point; not curled or held over back”, and lists “tail too long or badly carried” among faults. A corkscrew is part of the Bulldog and Boston look. In an AmStaff it moves a dog away from the written ideal.

Inside the Helsinki CT study: 19 scans, three two-copy dogs

SamSo did the two-copy AmStaffs look like Bulldogs once they were scanned?

Elena MarshOnly partly. In the Helsinki CT work, all two-copy dogs had malformed tail bones, but none had the thoracic wedge vertebrae that turn up in most scanned French Bulldogs.

The Helsinki group did something the Bulldog literature cannot do. Because AmStaffs come in all three genotypes, they could compare dogs with zero, one and two copies directly. They CT-scanned 19 AmStaffs (8 with no copy, 8 carriers, 3 with two copies) and took radiographs of 23 more.

The tail results were consistent. Four two-copy dogs could be evaluated, and every one had abnormal caudal vertebrae: 1, 8, 10 and 14 malformed vertebrae respectively. Their tails averaged 26 cm, against 33 cm in the 8 carriers and 34 cm in the 8 dogs without the variant. The number of tail vertebrae was not reduced, though; each dog still had 20 to 21. So the two-copy AmStaff tail is a shorter, kinked version of a normal tail rather than the truncated screw tail seen in Bulldogs. The university’s summary put it directly: the AmStaff tail abnormalities “were less severe than the screw tails typically seen in English Bulldogs, French Bulldogs and Boston Terriers”.

The spine results were the surprise. None of the AmStaffs, in any genotype, had thoracic or lumbar vertebral malformations. One two-copy dog had a malformed sacrum. Compare that with the 2018 Davis data, where thoracic malformations showed up in 92.1% of two-copy French Bulldogs, 45.5% of Boston Terriers and all 7 Bulldogs examined. The Helsinki authors offered two explanations: either the deletion does not cause thoracic malformations on its own, or its penetrance varies even more than the 45 to 100% seen in the screw-tail breeds. With three scanned dogs, nobody can yet say which.

A shorter hard palate and the BOAS question

The AmStaff is not usually listed as a brachycephalic breed, and that is what made the skull measurements interesting. In the same Helsinki CT set, hard palate length tracked genotype: shortest in two-copy dogs, longer in carriers and longest in dogs with no copy. The carrier versus no-copy difference was itself statistically significant (P = 0.007). Two-copy dogs also had a shorter skull base, a higher facial index and a smaller soft palate height, and both carriers and two-copy dogs had a lower ratio of hard palate to skull base.

Soft palate length, however, did not differ by genotype (P = 0.80). The authors’ reasoning: if the bony palate shortens but the soft tissue does not, there is relatively more soft palate in a smaller space, which could add to the risk of brachycephalic obstructive airway syndrome (BOAS). That is a hypothesis, not a measured breathing outcome, and it rests on only three two-copy dogs. It does mean one point needs stating carefully: in this study carriers were not identical to dogs without the variant on every skull measure.

It also helps to keep the gene in proportion. The 2018 Davis paper found that the screw-tail breeds carry other skull-shortening variants as well, including SMOC2 and BMP3, and its authors described head shape as a complex trait influenced by several loci, among them BMP3 and SMOC2. A short muzzle is a polygenic trait, and DVL2 is one contributor among several.

Two of the three scanned two-copy AmStaffs also had a mild heart murmur, and one was diagnosed with aortic stenosis. The authors called a cardiac link speculative, and the press release labeled it “a preliminary finding that requires further study”. It belongs in a list of open questions, not in a list of known effects.

A cleft palate signal from US primary-care records

SamWas the cleft palate link found in AmStaffs specifically, or just in Bulldogs?

Elena MarshAmStaffs specifically. In the 2026 Animal Genetics study, 19 of 27 AmStaff and AmStaff-mix cleft palate cases had two copies, and none of 15 unaffected controls did.

The newest piece of evidence comes from the United States. Donner and colleagues’ 2026 paper in Animal Genetics linked Wisdom Panel genotypes to Banfield primary-care records from the US and Mexico. A genome-wide comparison of 266 dogs with cleft palate and 506 controls produced a signal on chromosome 5. A separate scan of mixed-breed dogs (103 cases, 205 controls) pointed to the same region.

Whole-genome sequencing of three cleft palate cases, an English Bulldog, an AmStaff and a Boston Terrier, put DVL2 c.2044delC at the top of the list of high-impact variants. Follow-up genotyping in mixed breeds gave the variant the lowest p-value of anything tested (9.03 x 10^-21).

For AmStaffs and AmStaff mixes, the case-control split was stark. Among 27 cleft palate cases, 2 had no copy, 6 were carriers and 19 had two copies. Among 15 controls, 11 had no copy, 4 were carriers and none had two copies (Fisher p = 7.3 x 10^-7). The authors call the variant a risk factor for cleft palate. They do not say every two-copy puppy gets one, and a case-control design cannot tell you the proportion that do. What it does tell a breeder is that the two-copy genotype, which only arises when both parents carry the variant, is heavily overrepresented among puppies born with this defect.

One practical oddity: the Wisdom Panel microarray used in that research did not directly include the DVL2 variant, and the Wisdom Panel Premium product page (USD 127.99 on sale, USD 159.99 regular, as listed October 18, 2026) does not list a screw tail result among its traits. It does list SMOC2 “Short Snout” and the TBXT “Short Tail” bobtail variant, a different gene that has nothing to do with this one.

Crooked tails without the variant, and clear results that prove less than they seem

A positive result does not explain every crooked tail, and a negative one does not certify a normal spine. The Helsinki data show both sides of that.

Among the CT-scanned AmStaffs, 3 of 8 carriers and 2 of 8 dogs with no copy had caudal vertebral malformations too. Across the study’s cohorts, 12 carriers and 13 dogs without the variant had some kind of tail anomaly, and a Staffordshire Bull Terrier with no copy had a short kinked tail with only 11 caudal vertebrae. The authors concluded that other, still unknown variants also cause tail kinks. A crook near the tip of an AmStaff’s tail is therefore not, on its own, evidence of DVL2.

The single-carrier result needs the same care. In the CT group, carriers were not distinguishable from no-copy dogs on tail length, but they did sit between the two homozygous groups on hard palate length. The commercial labs describe carriers differently, as covered below, and the honest summary is that one copy appears to have small measurable effects on skull shape in AmStaffs and no consistent effect on the tail.

The broader hemivertebra literature mostly concerns Bulldogs, Frenchies and Pugs, and none of these studies typed DVL2. A 2017 Royal Veterinary College and Glasgow study found hemivertebrae in 93.5% of neurologically normal French Bulldogs, and an Australian CT study by Brown and colleagues (2021) found that the presence, number and type of malformation did not reliably predict neurological signs. A 2024 review in The Veterinary Journal reached a similar view: many of these malformations are incidental in neurologically normal dogs. OMIA classifies canine hemivertebrae as multifactorial, with no single key variant. None of that is AmStaff data, and the Helsinki scans found no thoracic malformations in the breed at all.

Pugs are a useful reminder of why gene and appearance should not be confused. All 29 Pugs in the 2018 study had no copy of the variant, and their curled tails were full-length with no caudal malformations. The Czech lab Genomia makes the same point on its test page: a Pug’s curl “is not a screw tail”.

Recessive or incomplete dominant? Four labels for one deletion

US buyers meet the same deletion under different names, different coordinates and, more confusingly, different inheritance claims. A side-by-side reading of the lab pages on October 18, 2026:

• Animal Genetics sells it as “Screwtail” in its canine trait tests, not its disease panel. Results come as ST/ST, n/ST or n/n, and the page says a carrier “may result in a partially kinked or shortened tail”.
• Paw Print Genetics listed it as “ST Locus (Screw Tail, Bulldog and Terrier Type)”, test 415, with inheritance given as “Autosomal Incomplete Dominant with Incomplete Penetrance and Variable Expressivity” and the variant written as chr5:32195044 (canFam3), a 1 bp deletion. Our copy is a December 2024 archive; the live address now redirects to a notice that Paw Print is merging with Orivet, so current availability is unconfirmed.
• Genomia calls it “Screw tail”, writes the variant as c.2044delC, says it is “a recessively inherited disease” and that carriers “are without clinical signs”.
• DogWellNet, the International Partnership for Dogs’ harmonization database, records the Genomia test as c.2051del and describes inheritance as “suggested to be autosomal dominant, with incomplete penetrance”.

The coordinate differences are not separate mutations. OMIA lists the published c.2044delC as c.2051del and notes only that the name was changed to HGVS nomenclature. (As a general HGVS convention, not something OMIA spells out, a deletion inside a repeat is written at its 3′-most position.) Same base, different bookkeeping. The OMIA entry for screw tail in dogs lists the mode of inheritance as autosomal recessive.

The inheritance disagreement is partly about which trait you look at. For the malformed tail in the 2018 and 2021 data, two copies were needed: carriers in the Davis segregation analysis had no thoracic or caudal malformations, and all two-copy dogs had caudal ones. For skull measurements in AmStaffs, carriers fell in between. A lab that frames the variant as “incomplete dominant” is leaning on the second kind of finding; one that says “recessive” with symptom-free carriers is leaning on the first. For breeding, the arithmetic is the same either way.

The carrier-to-carrier math, and the gap in AmStaff screening

SamIf neither the club nor OFA asks for it, is this test really worth paying for?

Elena MarshFor a pet, rarely. For a breeding dog it is cheap insurance: one AmStaff in four was a carrier in the Helsinki sample, and two carriers give each puppy a 25% chance of two copies.

The useful math is simple. Breed two carriers and each puppy has a 25% chance of two copies, a 50% chance of one and a 25% chance of none. Breed a carrier to a dog with no copy and no puppy gets two copies, though about half will be carriers. Testing both partners turns a hidden gamble into a planned pairing without dropping carriers from the gene pool.

The Staffordshire Terrier Club of America’s health page recommends testing for hips, congenital cardiac disease, thyroid, NCL-A cerebellar ataxia and ALPP (juvenile laryngeal paralysis and polyneuropathy), with eyes and elbows optional. CHIC requirements match that list. Neither mentions DVL2, screw tail or vertebral screening, and the OFA breed statistics served on October 18, 2026 show no DVL2 row for the breed. For scale, the same OFA data show 4,304 AmStaff hip evaluations and 966 NCL tests. US breeders do submit DNA results; they just are not asked about this one.

Breed size matters here too. The American Staffordshire Terrier ranked No. 97 in the AKC’s 2025 registration ranking. A modestly sized registered population means individual stud dogs can have a large reach. A popular sire who carries the variant can spread it widely before anyone notices kinked tails, because carriers in the Helsinki data did not have shorter tails.

No published DVL2 frequency exists for American AmStaffs. The 27.7% carrier figure is Finnish, and US lines may differ in either direction. Treat it as a reason to test, not as a US estimate.

Where to order, as listed on October 18, 2026 (prices before tax and shipping):

• Animal Genetics: USD 45.00 (GBP 40.00 on its UK pricing), “Breeds: All”, turnaround stated as 3-5 or 7-10 business days.
• Genomia (Pilsen, Czech Republic, ships internationally with cheek swab kits): USD 56.00 without VAT, usual turnaround 12 business days, AmStaff named among 11 listed breeds.
• LABOKLIN: “Robinow-like syndrome (DVL2)”, service 8747, AmStaff listed, 1-2 weeks after arrival; price not shown on the product page.
• Paw Print Genetics ST Locus: price and post-merger availability not confirmed.
• UC Davis VGL discovered the variant, but we could not confirm that it sells a DVL2 test (its site blocked our requests). Embark did not turn up a DVL2 result in our search either.

Lab reports vary in notation. Whatever the letters, confirm which result means no copy, one copy and two copies before you make a pairing decision. For a breeder outside the US, England’s 2018 dog breeding licensing regulations say no dog may be kept for breeding if it can reasonably be expected, “on the basis of its genotype, phenotype or state of health”, that breeding from it could harm its offspring’s health or welfare. A DNA result is exactly the kind of genotype information that wording anticipates.

Tail-fold infections, wobbly hind legs and insurance timing

Most AmStaffs with a kinked tail need nothing more than a look from their regular vet. VCA Animal Hospitals’ guide to corkscrew tails notes that the most common problem is recurrent infection in the tail fold, that diagnosis starts with a physical exam and sometimes X-rays, and that treatment choices are made case by case. If you see redness, odor or discharge around a kinked tail, book a regular appointment.

Neurological signs are a different category. Hind-leg weakness or wobbling, sudden paralysis, or trouble urinating or defecating need a veterinarian the same day, whatever the dog’s DVL2 result. Spinal problems are diagnosed with imaging, and VCA notes that board-certified neurologists, who list “DACVIM (Neurology)” after their name, use CT and MRI to look at the nervous system. Your vet can refer you. Bring the DNA report along, but expect it to be one piece of context rather than the answer.

For a sense of cost, Embrace’s 2021 article on hemivertebrae gave USD 1,000 to 3,000 as a typical range for advanced imaging and USD 1,500 to 4,000 for spinal surgery at one location. Those figures describe the Bulldog-type thoracic scenario, which the Helsinki scans did not find in AmStaffs.

Insurance timing is where a visible kink matters most. The NAIC Pet Insurance Model Act defines a preexisting condition to include signs or symptoms present before the policy’s effective date or during its waiting period, and separately defines congenital (present from birth) and hereditary (genetically transmitted) disorders. A tail kinked since birth is easy to see, so it may count as a pre-existing sign under that kind of wording. Embrace says it covers hereditary and congenital conditions that are not pre-existing, counts any irregularity you or your vet noted before the waiting period ends, uses a 2-week illness waiting period, and does not pay for DNA testing. Pumpkin and Figo list hereditary conditions as covered while excluding pre-existing ones, and Nationwide’s coverage page says pre-existing conditions are not covered. Figo may re-cover “curable” pre-existing conditions after 12 symptom-free months, but a vertebral malformation is not curable in the usual sense. Enroll a puppy early, and read the definitions before assuming a tail-related claim will be paid.

UK readers should check breed lists as well: Animal Friends excludes “American/Irish Staffordshire Bull Terriers” and their crosses. Whether that wording covers the American Staffordshire Terrier by name is not clear from the page, so ask before you buy.

Frequently asked questions

Q. Does a DVL2 test diagnose hemivertebrae or spinal disease in my AmStaff?
No. It reports how many copies of one variant your dog has. Spinal problems are diagnosed by a veterinarian with a neurological exam and imaging. In the Helsinki study, no AmStaff of any genotype had thoracic or lumbar malformations on CT.

Q. My AmStaff tested n/n but has a kinked tail. Was the test wrong?
Probably not. In the same research, several AmStaffs with no copy of the variant had tail vertebra malformations, and the authors concluded other unknown genetic causes exist. A clear result rules out this one deletion, nothing more.

Q. Can I breed a carrier?
Yes, if the partner has no copy. That pairing produces no two-copy puppies. Carrier-to-carrier pairings give each puppy a 25% chance of two copies, the genotype linked to malformed tail vertebrae and, in the 2026 US study, to cleft palate risk.

Q. Is a carrier AmStaff completely unaffected?
Carriers in the Helsinki CT group had normal tail length on average, but their hard palate was shorter than in dogs without the variant. Whether that has any health meaning for a carrier is not known.

Q. Will a two-copy AmStaff have breathing or heart problems?
Not necessarily. The shorter palate is a suggested BOAS risk based on three two-copy dogs, and the heart finding was called speculative by the authors. If your dog snores heavily, tires quickly or faints, see your veterinarian.

Q. Is this the same as Robinow syndrome in people?
It is similar, not the same. Human Robinow syndrome involves related genes such as DVL1, DVL3, ROR2 and WNT5A. The dog condition is called “Robinow-like” because the gene and pathway are closely related.

References

• Mansour TA, Lucot K, Konopelski SE, Dickinson PJ, Sturges BK, Vernau KL, Choi S, Stern JA, Thomasy SM, Doring S, Verstraete FJM, Johnson EG, York D, Rebhun RB, Ho HYH, Brown CT, Bannasch DL (2018). Whole genome variant association across 100 dogs identifies a frame shift mutation in DISHEVELLED 2 which contributes to Robinow-like syndrome in Bulldogs and related screw tail dog breeds. PLoS Genet 14(12):e1007850.
• Niskanen JE, Reunanen V, Salonen M, Bannasch D, Lappalainen AK, Lohi H, Hytonen MK (2021). Canine DVL2 variant contributes to brachycephalic phenotype and caudal vertebral anomalies. Hum Genet 140(11):1535-1545.
• Donner J, Hytonen M, Freyer J, Lohi H, Foran RC, Forman OP (2026). Association of the Disheveled 2 (DVL2) gene c.2044delC variant with increased risk of canine cleft palate. Anim Genet 57(3):e70101.
• Ryan R, Gutierrez-Quintana R, Ter Haar G, De Decker S (2017). Prevalence of thoracic vertebral malformations in French bulldogs, Pugs and English bulldogs with and without associated neurological deficits. Vet J 221:25-29.
• Brown JD, Podadera J, Ward M, Goldsmid S, Simpson DJ (2021). The presence, morphology and clinical significance of vertebral body malformations in an Australian population of French Bulldogs and Pugs. Aust Vet J 99(9):378-387.
• De Decker S, Rohdin C, Gutierrez-Quintana R (2024). Vertebral and spinal malformations in small brachycephalic dog breeds: current knowledge and remaining questions. Vet J 304:106095.
• OMIA 002186-9615: Screw tail in Canis lupus familiaris (dog). Online Mendelian Inheritance in Animals, University of Sydney.
• University of Helsinki (2021-02). Researchers discover new features of a gene defect that affects the length of the muzzle and caudal vertebrae in dogs. Press release.
• American Kennel Club. Official Standard of the American Staffordshire Terrier (approved June 10, 1936).
• Staffordshire Terrier Club of America. Health testing recommendations. Accessed 2026-10-18.
• Genomia s.r.o. (Pilsen, Czech Republic). Screw tail. Accessed 2026-10-18.
• DogWellNet / International Partnership for Dogs. Harmonization of Genetic Testing for Dogs: Screw tail. Accessed 2026-10-18.
• National Association of Insurance Commissioners. Pet Insurance Model Act (MO-633), 2022.

About the author

Elena Marsh

Editor & writer (not a veterinarian)

A writer with a molecular-biology background and a lifelong dog and cat owner. Not a veterinarian — she translates peer-reviewed genetics research and primary data into plain language, always as information rather than diagnosis.