BOAS Beyond Bulldogs: What the 2026 Cambridge Study Found in 14 Flat-Faced Breeds

For more than a decade, Brachycephalic Obstructive Airway Syndrome (BOAS) has been the headline welfare problem of three breeds: the Bulldog, French Bulldog and Pug. But anyone who has heard a Pekingese snore from across a room, or watched a Boston Terrier wheeze on a hot pavement, has wondered the obvious question: what about the rest of the flat-faced breeds?

A new 2026 study published in PLOS ONE by Tomlinson, Liu, Sargan and Ladlow at the University of Cambridge finally answers it. Over three years they assessed 898 dogs across 14 brachycephalic breeds with the same Respiratory Function Grading (RFG) system used by the UK Kennel Club. The picture that emerges is sobering — and breed-specific in ways that matter for every flat-faced-dog owner.

Which 14 breeds were tested?

The Cambridge team deliberately stretched beyond the usual suspects to capture the full range of brachycephalic phenotypes — extreme flat faces, moderate flat faces, and even short-but-wide skulls like the Staffordshire Bull Terrier. The fourteen breeds were:

Affenpinscher, Boston Terrier, Boxer, Cavalier King Charles Spaniel, Chihuahua, Dogue de Bordeaux, Griffon Bruxellois, Japanese Chin, King Charles Spaniel, Maltese, Pekingese, Pomeranian, Shih Tzu and Staffordshire Bull Terrier.

Each dog was over 12 months old, had no history of BOAS surgery, and was scored from Grade 0 (clinically clear) to Grade 3 (severe respiratory distress) after a 3-minute trot-to-gallop exercise test.

The headline result: almost every breed had affected dogs

Only two breeds — the Maltese and the Pomeranian — had zero dogs with clinically significant disease. Every other breed in the study had at least some Grade 2 or Grade 3 dogs.

The two worst offenders were striking:

• Pekingese — only 10.9% of dogs were Grade 0.
• Japanese Chin — only 17.4% were Grade 0.

In other words, more than 8 in 10 Pekingese in this UK pet population had at least mild detectable upper-airway obstruction. The traditional "big three" (Bulldog, French Bulldog, Pug) are not even in this study — yet several of the smaller, less-discussed flat-faced breeds appear to be in just as much trouble.

Three risk factors held up across every breed

When the team ran multiple logistic regression across the entire 898-dog dataset, three variables independently predicted BOAS status:

1. Higher body condition score (BCS) — fatter dogs breathed worse.
2. Nostril stenosis — narrow, slit-like nares.
3. Lower craniofacial ratio (CFR) — a more extreme flat face (snout length ÷ cranial length).

Together those three factors explained about 20% of the variation in BOAS status. That is a lot for a disease this complex — and crucially, two of the three are things owners can influence today: weight, and choosing a puppy with open nostrils and a longer relative muzzle. The same weight-and-conformation conversation already drives our reporting on the 2026 raw-meat-vs-kibble obesity study, and it applies just as much here.

The surprises

Two findings will reshape breed-club conversations:

• Extreme flat face does not always mean severe BOAS. The King Charles Spaniel — anatomically one of the most facially hypoplastic breeds in the study — had lower rates of clinically significant disease than expected. The authors suggest different breeds compensate differently, and that genetic background matters as much as skull shape.
• "Brachycephalic" is not one phenotype. The Staffordshire Bull Terrier is brachycephalic by skull index but has a relatively long muzzle, and its BOAS profile looks nothing like a Pekingese. This is why the authors argue strongly for breed-specific cut-offs, not one-size-fits-all rules.

That mirrors what we saw in the 2026 UBC study on breed-clade learning differences: treating "all dogs" as one population hides the most important signal.

What this means for owners — practically

If you live with a flat-faced dog, the 2026 paper turns into a short, evidence-based checklist:

• Get a Respiratory Function Grade. The same scheme used in the study is now offered for several breeds via the UK Kennel Club / University of Cambridge. Even one grading visit gives you a baseline.
• Keep your dog lean. BCS was one of only three independent predictors. The structural problems are fixed; the fat layer is not. (For a deeper dive on body composition and breed, see our coverage of why large-breed dog brains age slower than their bodies.)
• Look at the nostrils, not just the face. Stenotic nares are the most fixable conformational problem. Many vets correct them at the same time as a routine neuter, often with dramatic improvement.
• Mind heat, exercise and sleep. Noisy sleep is not "just how the breed is" — it is sleep-disordered breathing, and it is a Grade 1+ sign.
• If you breed, use the data. The authors call explicitly for breed-specific selection criteria. A "moderate" Pekingese is not the same animal as a "moderate" Boston Terrier.

How this fits the wider 2026 picture

This study is part of a growing wave of breed-aware veterinary research we have been tracking — from spinal canal size predicting disc disease in small breeds to whether heartworm-positive dogs can safely have surgery and how lifetime sports keep a dog's brain young. The thread that links all of them is the same: averages across "all dogs" hide the answer; the truth lives at the breed and the individual level.

For BOAS, that truth is now hard to ignore. The disease is not a Bulldog-French-Bulldog-Pug problem — it is a flat-faced-dog problem, and the next decade of welfare improvement will depend on breeders, vets and owners using breed-specific evidence like this study to drive selection.

---

Source: Tomlinson F, Liu N-C, Sargan DR, Ladlow JF (2026). A cross-sectional study into the prevalence and conformational risk factors of BOAS across fourteen brachycephalic dog breeds. PLoS ONE 21(2): e0340604. doi.org/10.1371/journal.pone.0340604