Flat-faced dogs face biggest breathing risks from skull shape, weight and nostril width

The appeal of flat-faced dogs is easy to understand: compact bodies, round heads, large eyes, and a look that owners often describe as perpetually quizzical or endearingly grumpy. The health cost embedded in that face shape is less easy to see. A study published February 18, 2026 in PLOS ONE quantifies that cost across 14 brachycephalic breeds in more detail than previous research managed, identifying the physical traits most predictive of serious breathing obstruction and pinpointing which breeds carry the greatest risk.

The condition and why it matters

Brachycephalic Obstructive Airway Syndrome (BOAS) is the clinical name for the breathing difficulties that arise from the shortened skull shape called brachycephaly. When the skull is compressed front-to-back, soft tissue structures - including the palate, the throat lining, and the cartilage of the nose - become crowded into a space too small for them. Nostrils collapse, airways narrow, and the dog must work substantially harder to move air with every breath. Symptoms range from snoring and exercise intolerance to severe respiratory distress requiring surgery. Heat and stress make all of it worse.

Study design: 898 dogs, 14 breeds

Francesca Tomlinson of the University of Cambridge and colleagues measured 898 dogs from 14 breeds including Boxers, Cavalier King Charles Spaniels, King Charles Spaniels, Chihuahuas, Pekingese, Japanese Chins, Pomeranians, and Shih Tzus, as well as the three reference breeds already well established as high-risk: Pugs, French Bulldogs, and Bulldogs. Each dog was assessed for skull and nose dimensions, body proportions, and neck measurements. Dogs were graded on a BOAS scale from zero (few symptoms) to three (difficulty exercising and breathing adequately at rest).

By comparing the 14 study breeds against the Pug/French Bulldog/Bulldog baseline, the researchers could estimate relative disease burden and identify which physical measurements drove the grading scores.

Three traits dominate the risk profile

A flat face geometry - specifically, a wider and shorter skull shape relative to body size - was the strongest structural predictor of BOAS severity across the dataset. Narrowed nostrils were the second key factor; collapsed nares restrict airflow at the first point of entry and compound whatever downstream narrowing exists in the throat. Body weight was the third consistent predictor, with overweight dogs showing substantially worse scores than lean dogs of the same breed.

In some breeds, additional factors appeared relevant. Shorter, higher-set tails - which correlate with spinal compression in some brachycephalic lines - and thicker necks around the airway were associated with worse scores in specific breed groups. These findings suggest that BOAS involves the entire body conformation, not just the face.

Pekingese and Japanese Chins at highest risk among less-studied breeds

While Pugs, French Bulldogs, and Bulldogs have long been identified as the most severely affected breeds, the study found that Pekingese and Japanese Chins carry BOAS burdens comparable to or exceeding those reference breeds. Both are ancient breeds with extremely flat skull profiles that have been selected for over centuries. They have received far less systematic veterinary research attention than the more commercially popular Frenchie or Bulldog, which means many owners and even some veterinarians may underestimate their risk.

Within-breed variation is substantial

One of the study's practically important findings is that BOAS varies considerably within individual breeds, not just between them. Even within breeds at high population-level risk, individual dogs with less extreme skull shapes, lower body weight, and wider nostrils showed substantially lower BOAS grades. This has direct implications for breeding: selecting within high-risk breeds for less extreme physical features could reduce disease burden without abandoning the breed type entirely.

The study involved 898 dogs, which is a meaningful sample, but the distribution across 14 breeds means some individual breed samples were modest. The BOAS grading was conducted by trained veterinary staff, reducing subjectivity, but the assessment remains somewhat clinical rather than quantified via objective airway measurements such as whole-body barometric plethysmography, which some studies use. The sample was drawn from dogs presented to veterinary or research settings, which may oversample affected animals relative to the true breed population.