Brachycephaly in Dogs: Health Evidence, Measurement and the Demand Problem

1. What Brachycephaly Is

1.1 The Conformation

Brachycephaly is a heritable shortening of the facial skeleton, produced through selection for craniofacial morphology in breeds including the Bulldog, Pug and French Bulldog (Packer, Hendricks, Tivers & Burn, 2015). The skull base is shortened while the soft tissue of the airway is not, which is the mechanical core of everything that follows.

1.2 The Measurement

The standard metric is the craniofacial ratio: muzzle length divided by cranial length. Muzzle length is measured from the dorsal tip of the nasal planum to the stop; cranial length from the stop to the occipital protuberance (Packer et al., 2015).

Values run from roughly 0.08 in extreme breeds such as the Pug to around 0.5 in moderate ones — a range produced by selection on appearance (as the breed evidence illustrates), and one whose developmental basis is polygenic rather than a single switch (with gene–environment interaction covered separately). A ratio of 0.5 or below is the conventional threshold for classifying a breed as brachycephalic.

1.3 What BOAS Refers To

Brachycephalic obstructive airway syndrome describes the condition in which soft tissue obstructs the airway in dogs with this conformation. The primary abnormalities are stenotic nostrils, aberrant nasal turbinates, an elongated soft palate, macroglossia and a hypoplastic trachea; secondary changes include everted laryngeal ventricles and progressive laryngeal collapse.

Clinical signs range from snoring and exercise intolerance to syncope, and discomfort of this kind is easily missed (as visceral pain illustrates). The range matters, because the mild end is what gets normalised (much as behavioural signs get reinterpreted elsewhere).

2. The Epidemiological Finding

2.1 The Study

Using primary-care veterinary clinical data, researchers extracted all disorders recorded during 2016 from a random sample of 22,333 dogs within the VetCompass Programme, drawn from a sampling frame of 955,554 dogs under UK veterinary care (O'Neill, Pegram, Crocker, Brodbelt, Church & Packer, 2020).

Univariable and multivariable logistic regression were used; the multivariable models account for confounders such as differing age and neutering structures.

2.2 What It Established

Brachycephalic dogs showed predispositions across a range of disorders, with an elevated disorder count relative to non-brachycephalic dogs. The disorder groups documented as conformation-related include respiratory disease, eye disease, dystocia, spinal disease, heat stroke and pneumonia (O'Neill et al., 2020) — several involving chronic discomfort rather than acute events (with documented behavioural consequences).

2.3 Why the Design Matters

Earlier work relied on single-disorder studies, or on prevalence within breeds without a comparator group. A US pet insurance analysis had compared both groups across disorders, but drew on a biased insured subset without multivariable adjustment (O'Neill et al., 2020).

The VetCompass design addresses both problems, which is why it carries the epidemiological weight (the same methodological point that decides other canine questions).

3. The Measurement Dispute

3.1 The First Finding

Two studies were run: an exploratory sample of 700 dogs across diverse breeds and conformations, and a confirmatory study of 154 brachycephalic dogs. BOAS risk increased sharply and non-linearly as relative muzzle length shortened, and — the specific finding that has travelled furthest — BOAS only occurred in dogs whose muzzles comprised less than half their cranial lengths (Packer et al., 2015).

Thicker neck girth independently increased risk in both populations, a factor known from human sleep apnoea and not previously described in dogs.

3.2 The Contradicting Finding

Two years later, 604 Pugs, French Bulldogs and Bulldogs were assessed with soft tape measurements, breed-specific models and validation against a BOAS index derived from whole-body barometric plethysmography (Liu, Troconis, Kalmar, Price, Wright, Adams, Sargan & Ladlow, 2017).

Stenotic nostrils were associated with BOAS in all three breeds, and higher body condition scores raised risk in Pugs and Bulldogs — the one risk factor an owner can influence after acquisition (much as feeding structure is modifiable where genetics are not). Craniofacial ratio, however, showed minimal to no association with BOAS severity within these breeds — and several measurements that did reach significance had poor inter-observer agreement.

3.3 How the Two Fit Together

This is where secondary coverage goes wrong. The studies asked different questions at different levels: Packer examined variation across the full brachycephalic range, from Pugs at a craniofacial ratio around 0.08 to moderately brachycephalic breeds around 0.5. Liu examined variation within three extreme breeds, where the available range is narrow.

A predictor can be strong across a wide range and weak within a narrow band of it. Liu's own paper frames the contribution as addressing whether the across-breed measurements have practical implications in specific breeds — not as a refutation.

3.4 What Follows Practically

For breeding policy at breed level, muzzle proportion remains defensible. For predicting whether an individual Pug will develop BOAS, nostril stenosis and body condition are the better indicators. Both are true, and treating either as the whole answer misrepresents the literature (as level-of-analysis confusions do generally).

4. Beyond the Airway

4.1 Why the Airway Dominates

BOAS is visible, audible and dramatic, which makes it the reference case. Framing brachycephaly as a breathing problem nonetheless understates the conformational burden.

4.2 The Wider Disorder Range

The disorder groups linked to brachycephalic conformation include ocular disease — shallow orbits leaving the globe exposed — obstetric difficulty requiring caesarean section, spinal disease associated with screw-tail conformation, heat stroke, and pneumonia (O'Neill et al., 2020).

4.3 The Thermoregulation Problem

Dogs dissipate heat primarily through panting, which depends on airflow across the nasal mucosa. A shortened, obstructed airway degrades the mechanism that would otherwise handle heat load.

The consequence is quantified. Analysing veterinary records covering 905,543 dogs, brachycephalic skull shape was associated with 1.32 times the odds of exertional heat-related illness, 2.36 times the odds of environmental and 3.07 times the odds of vehicular heat-related illness, compared with mesocephalic dogs (Hall, Carter & O'Neill, 2020). Notably, 74.2 percent of events were exertional rather than the parked-car scenario that dominates public messaging (with arousal adding a further physiological load).

4.4 Behavioural Consequences

A dog that cannot breathe comfortably is under sustained physiological load (with the stress systems described separately), and sleep is disrupted in obstructive airway disease — while canine sleep architecture is directly involved in memory consolidation and emotional regulation (as the neurophysiology shows; with measurable effects on learning).

These connections are plausible and have not been quantified in brachycephalic dogs specifically — a gap section 9 returns to.

5. The Demand Problem

5.1 The Paradox

Brachycephalic breeds have risen rapidly in popularity while the evidence linking their conformation to chronic and severe health conditions has accumulated (Packer, Murphy & Farnworth, 2017). Any account of this as an information deficit has to explain that timing.

5.2 What Owners Actually Weigh

Comparing owners of three popular brachycephalic breeds with owners of popular non-brachycephalic breeds, the factor with the strongest influence on the purchase decision was appearance, with perceived breed health less influential for brachycephalic purchasers than for the comparison group (Packer et al., 2017).

The finding is not that these owners are indifferent to health. It is that health weighed less against appearance than it did for owners choosing other breeds.

5.3 Normalisation

The mechanism that makes this self-sustaining is the reclassification of clinical signs as breed characteristics, and it has been quantified. In a questionnaire study of owners of 285 dogs across 68 breeds, more than half of the owners whose dogs showed clinical signs of BOAS did not recognise that their dog had a breathing problem, frequently describing the signs as normal for the breed (Packer, Hendricks & Burn, 2012).

Snoring, snorting, exercise intolerance and sleeping in positions that keep the airway open are described as typical rather than as symptoms.

An owner who has never seen the breed breathe otherwise has no comparison, and pain in dogs is under-recognised even where it is not normalised by breed expectation (a documented pattern in behaviour caseloads). A dog whose signs are normalised does not get presented (a reading error with the same structure as those in behaviour).

6. Regulatory Approaches

6.1 The Shared Statutory Test

Several European animal welfare acts share a structure. Section 11b(1) of the German Animal Welfare Act prohibits breeding vertebrates where breeding knowledge gives reason to expect that offspring will hereditarily lack body parts or organs needed for species-appropriate use, or have them unsuited or altered, and thereby experience pain, suffering or damage (Tierschutzgesetz, § 11b). Norway's Animal Welfare Act contains a comparable provision in section 25.

The difficulty follows directly from section 3: establishing what is to be expected requires a predictor, and the one that works across breeds performs poorly within the most extreme. What follows describes how these provisions have been applied; it is not legal advice.

6.2 The Netherlands: Six Criteria, Not One Threshold

The Dutch approach is the most concrete and more sophisticated than its usual summary. Article 3.4 of the Animal Keepers Decree has since 2014 prohibited breeding companion animals in ways that harm the welfare or health of parents or offspring; in March 2019 the Ministry issued enforcement criteria drawn up by Utrecht University's expertise centre for companion animal genetics, and a policy rule of 2023 gave them formal shape.

Six criteria apply, only one conformational: normal breathing sound at rest, open nostrils, a craniofacial ratio of at least 0.3, no nose fold, limited visible sclera, full eyelid closure. A ratio of 0.5 or above passes; between 0.3 and 0.5 breeding is permitted only if every other criterion is met; below 0.3 the dog is excluded (NVWA; Staatscourant 2023, no. 23619).

That structure matters for section 3. The Dutch did not build the regulation on the craniofacial ratio alone — they embedded it in a set that is otherwise functional, which is close to what the measurement dispute would recommend. Enforcement falls to the food safety authority and the animal welfare inspectorate, and a later court ruling extended the logic to pedigree registration.

6.3 Norway: A Case-by-Case Judicial Route

Norway took the opposite path — litigation rather than thresholds — and the outcome is instructive because it changed twice.

In early 2022 the Oslo District Court ruled that breeding English Bulldogs and Cavalier King Charles Spaniels violated section 25. Later that year the Borgarting Court of Appeal reversed the Bulldog finding as an overreach, and in October 2023 the Supreme Court confirmed that outcome: breeding Cavaliers violates the Act, while breeding English Bulldogs under a defined health-screening programme does not (Supreme Court of Norway, 2023).

Two things follow. Courts distinguished between breeds on evidence rather than conformation category — syringomyelia in the Cavalier was decisive, respiratory conformation in the Bulldog manageable through screening. And a judgment can be reported worldwide as a breeding ban while ending as a screening requirement.

The United Kingdom, by contrast, has no equivalent conformational threshold in breeding regulation; oversight runs through general welfare legislation and licensing, with breed standards handled by the Kennel Club — whose stated position has been that blanket bans are hard to enforce and risk driving breeding underground. That objection is not merely self-interested: unenforceable prohibitions displace demand rather than reduce it, and the Norwegian case drew exactly that criticism from veterinary geneticists.

6.4 Status of This Section

The positions above were verified against primary sources in August 2026: the German Act as then in force, the Norwegian Supreme Court judgment of 10 October 2023, the Dutch policy rule in the Staatscourant, and the enforcement criteria published by the Netherlands food safety authority.

Unlike the findings in sections 2 to 5, this material has a short half-life — thresholds are revised, judgments are appealed, and the Norwegian case changed twice within eighteen months. Readers consulting it later should verify the current position.

6.5 The Alternative: Grade the Function

Functional grading assesses respiratory function directly rather than inferring it from proportions — through exercise tolerance testing and, in research settings, whole-body barometric plethysmography (Liu et al., 2017).

That measures the outcome rather than the predictor, which is what the Norwegian Bulldog judgment effectively requires and what the Dutch criteria partly incorporate. It needs equipment and trained assessors, which is why tape measurements persist.

7. What This Means in Practice

7.1 For an Existing Dog

The conformation is fixed; the modifiable factors are not. Body condition raised BOAS risk in Pugs and Bulldogs independently of conformation, which makes weight management the clearest modifiable risk factor identified in that study (Liu et al., 2017).

Heat avoidance has the clearest evidential basis, and the exertional finding sharpens it: the risk sits in exercise more than in parked cars (Hall et al., 2020). Anything that raises arousal also raises respiratory demand, which makes calm handling a physiological measure rather than a stylistic one (with anxiety compounding it).

Avoiding pressure on the neck is a reasonable inference from the mechanics rather than a finding — no canine study has compared collar and harness in brachycephalic dogs specifically. Surgical correction exists, and plethysmographic work indicates that dogs which have undergone corrective surgery may remain affected by systemic consequences of the disease (Liu et al., 2017).

7.2 What Not to Assume

That snoring is benign. That an absence of collapse means an absence of BOAS — the syndrome is a spectrum, and the mild end is where normalisation operates. And that a dog which has never had a problem is unaffected: the Packer threshold describes population risk, not an individual guarantee in either direction.

7.3 For Advice

A professional asked about acquiring one of these breeds is in an uncomfortable position, because the honest answer conflicts with a decision usually already made. The defensible version is factual rather than moral: this conformation carries a documented disorder burden, these are the specific risks, and this is what reduces them.

Whether that framing changes acquisition decisions has not been tested directly. What is established is narrower and still relevant: over half of owners with affected dogs did not recognise a breathing problem at all (Packer et al., 2012), so naming the signs addresses a documented gap even where its effect on purchasing is unknown (and information that contradicts an existing decision is generally resisted).

8. Summary at a Glance

• The population-level evidence is strong and consistent — Across 22,333 dogs from a sampling frame of 955,554 under UK veterinary care, brachycephalic dogs showed predispositions across multiple disorder groups (O'Neill et al., 2020).

• The threshold finding — In the Packer et al. study, BOAS occurred only in dogs whose muzzles comprised less than half their cranial length, with risk rising sharply and non-linearly (Packer et al., 2015).

• Neck girth matters independently — Thicker necks raised risk in both study populations, paralleling human sleep apnoea (Packer et al., 2015).

• Within extreme breeds, the ratio stops predicting — Across 604 Pugs, French Bulldogs and Bulldogs, craniofacial ratio showed minimal to no association with BOAS severity (Liu et al., 2017).

• Nostril stenosis predicts within breeds — Associated with BOAS in all three breeds studied, alongside body condition in Pugs and Bulldogs (Liu et al., 2017).

• The two studies are not in contradiction — One examined variation across the brachycephalic range, the other within a narrow band of it.

• More than half of owners with affected dogs saw no problem — Among owners of dogs showing clinical signs of BOAS, more than half did not identify a breathing problem, describing the signs as normal for the breed (Packer et al., 2012).

• Heat risk is quantified and mostly exertional — 1.32 times the odds of exertional, 2.36 environmental and 3.07 vehicular heat-related illness versus mesocephalic dogs, with 74.2 percent of events exertional (Hall et al., 2020).

• Appearance outweighed health at purchase — The strongest influence on the decision, with perceived health less influential than for non-brachycephalic purchasers (Packer et al., 2017).

• Body condition is modifiable — Independently associated with BOAS risk in two of three breeds studied, and within an owner's control (Liu et al., 2017).

9. Research Gaps and Critical Appraisal

The conformation dispute is unresolved. Both landmark studies are methodologically sound and asked different questions. Reporting either as having refuted the other misrepresents the literature, and a measurement that works for breed-level policy while failing for individual prediction is an awkward but real result.

Inter-observer reproducibility is a weakness. Several conformational measurements showed poor agreement between observers (Liu et al., 2017), which limits what tape-measure thresholds can carry in enforcement.

The epidemiology rests on primary-care records. VetCompass captures what was presented and recorded, and normalisation works against it: an owner who considers snoring typical does not present the dog, so the true burden is more likely underestimated.

Disorder counts ignore severity and duration. The authors say so themselves: a count treats a mild dermatological problem and chronic airway obstruction as equivalent units (O'Neill et al., 2020).

Behavioural consequences are under-studied. The link from respiratory compromise to sleep disruption and behaviour has not been quantified in this population.

Owner-reported satisfaction measures the wrong thing. High reacquisition intent is regularly cited in defence of these breeds; it describes the owner's experience, not the dog's respiratory function (a limitation that constrains owner-reported outcomes throughout).

Longitudinal data are scarce. Whether breeding for longer muzzles produces measurable health gains over generations is the question that matters for policy, and that follow-up has not been done.

10. Conclusion

Brachycephaly is the clearest case in canine welfare of evidence and practice moving in opposite directions. The elevated disorder burden is well supported at population level: across more than twenty-two thousand dogs in primary-care records, flat-faced breeds carry predispositions spanning respiratory disease, eye disease, dystocia, spinal disease, heat stroke and pneumonia — and in the study that quantified the relationship, the airway syndrome that gives the group its name occurred only in dogs whose muzzles were less than half their cranial length. What is genuinely open is narrower and more technical than the public argument suggests: craniofacial ratio predicts risk well across the brachycephalic range and poorly within the most extreme breeds, where nostril stenosis and body condition do the work instead. Those two findings are usually presented as one having demolished the other, and they have not; they were measured at different levels and both hold at their own. The part that no measurement resolves is the demand. Popularity rose while the evidence accumulated, appearance outweighed perceived health in the purchase decision, and the clinical signs became breed characteristics in the description before they became symptoms in the consultation. That is not an information problem in the ordinary sense, which is why information alone has not fixed it — and why the most useful thing a professional can say is not that the breed is wrong, but that the snoring is not normal (much as behavioural signs are misread as temperament).

Key Insights (Takeaways)

• The epidemiological finding is robust and rests on primary-care data. Across a random sample of 22,333 dogs from a sampling frame of 955,554 under UK veterinary care, multivariable analysis found brachycephalic dogs predisposed across multiple disorder groups (O'Neill et al., 2020). The multivariable design matters: it accounts for the age and neutering differences that had confounded earlier comparisons.

• In the Packer et al. sample, BOAS occurred only below a craniofacial ratio of 0.5. Across 700 dogs of diverse conformations and a confirmatory sample of 154 brachycephalic dogs, risk rose sharply and non-linearly as relative muzzle length shortened, with thicker neck girth an independent risk factor (Packer et al., 2015). This is a finding within that dataset, not an established diagnostic cut-off.

• Within extreme breeds, that ratio stops working. Across 604 Pugs, French Bulldogs and Bulldogs, craniofacial ratio showed minimal to no association with BOAS severity, while stenotic nostrils predicted in all three breeds and body condition added risk in two (Liu et al., 2017). This is a level-of-analysis difference, not a refutation — the across-breed range is wide, the within-breed range is narrow.

• Appearance outweighed health in the purchase decision. Comparing brachycephalic with non-brachycephalic owners, appearance was the strongest influence and perceived health mattered less for brachycephalic purchasers (Packer et al., 2017). Popularity rose while the evidence against the conformation accumulated, which rules out a simple information deficit.

• More than half of owners of dogs showing clinical signs did not recognise a breathing problem. Across 285 dogs from 68 breeds, owners whose dogs showed clinical signs of BOAS frequently attributed them to the breed rather than identifying a problem (Packer et al., 2012). This is the mechanism that keeps the disorder burden invisible and the demand intact.

• Body condition is the variable an owner can actually change. Higher body condition scores raised BOAS risk in Pugs and Bulldogs independently of conformation (Liu et al., 2017). The skull is fixed; the weight is not, and it is the intervention with the clearest evidence behind it for a dog that already exists.

References

Hall, E. J., Carter, A. J., & O'Neill, D. G. (2020). Incidence and risk factors for heat-related illness (heatstroke) in UK dogs under primary veterinary care in 2016. Scientific Reports, 10(1), 9128. https://doi.org/10.1038/s41598-020-66015-8

Höyesterett (Supreme Court of Norway) (2023). Judgment of 10 October 2023, HR-2023-1901-A. Breeding of the Cavalier King Charles Spaniel is a violation of the Animal Welfare Act, while breeding of the English Bulldog under a specific breeding programme is not. https://www.domstol.no/en/supremecourt/rulings/2023/supreme-court-civil-cases/HR-2023-1901-A/

Liu, N.-C., Troconis, E. L., Kalmar, L., Price, D. J., Wright, H. E., Adams, V. J., Sargan, D. R., & Ladlow, J. F. (2017). Conformational risk factors of brachycephalic obstructive airway syndrome (BOAS) in pugs, French bulldogs, and bulldogs. PLoS ONE, 12(8), e0181928. https://doi.org/10.1371/journal.pone.0181928

Ministerie van Landbouw, Natuur en Voedselkwaliteit. Besluit houders van dieren, Article 3.4. Enforcement criteria for brachycephalic dogs, policy rule published in the Staatscourant, 2023, no. 23619. https://zoek.officielebekendmakingen.nl/stcrt-2023-23619.html

Nederlandse Voedsel- en Warenautoriteit (NVWA). Information for veterinarians on breeding with short-muzzled dogs. https://www.nvwa.nl/onderwerpen/dier/honden-en-katten/informatie-voor-dierenartsen-over-kortsnuiten

O'Neill, D. G., Pegram, C., Crocker, P., Brodbelt, D. C., Church, D. B., & Packer, R. M. A. (2020). Unravelling the health status of brachycephalic dogs in the UK using multivariable analysis. Scientific Reports, 10(1), 17251. https://doi.org/10.1038/s41598-020-73088-y

Packer, R. M. A., Hendricks, A., & Burn, C. C. (2012). Do dog owners perceive the clinical signs related to conformational inherited disorders as normal for the breed? A potential constraint to improving canine welfare. Animal Welfare, 21(S1), 81–93. https://doi.org/10.7120/096272812X13345905673809

Packer, R. M. A., Hendricks, A., Tivers, M. S., & Burn, C. C. (2015). Impact of facial conformation on canine health: Brachycephalic obstructive airway syndrome. PLoS ONE, 10(10), e0137496. https://doi.org/10.1371/journal.pone.0137496

Packer, R. M. A., Murphy, D., & Farnworth, M. J. (2017). Purchasing popular purebreds: Investigating the influence of breed-type on the pre-purchase motivations and behaviour of dog owners. Animal Welfare, 26(2), 191–201. https://doi.org/10.7120/09627286.26.2.191