Brachycephalic dogs are disproportionately affected by ocular surface diseases (1–5). D. G. O’Neill et al. showed that brachycephalic dogs have a 3.63 times higher risk of developing keratoconjunctivitis sicca (KCS) than mesocephalic breeds (4). This, along with their overrepresentation in veterinary ophthalmology (6), is due to anatomic and physiologic changes that are collectively referred to as brachycephalic ocular syndrome (BOS) (1, 3, 7). However, there is discrepancy regarding the most common ocular surface disease in the various brachycephalic breeds (5–13) (Table 1).

One of the brachycephalic breeds most commonly seen in specialist eye clinics is the French Bulldog. In 2024, 15.2% of all patients in a specialist eye clinic were French Bulldogs (14). Not only has their popularity increased in recent years, with registrations at the British Kennel Club rising by about 5.5 times between 2014 and 2021 (15), but they were the most popular dog breed in the United States in 2022 (16). In a retrospective study of 809 French Bulldogs, 76.6% showed keratitis (spontaneous chronic corneal epithelial defects (SCEED), 31%; melting ulcers, 19%; stromal keratitis, 16.3%; pigmentary keratitis, 5.3%; granulomatous keratitis, 5.2%) and 45% showed tear film disorders (quantitative KCS, 16.2%; qualitative KCS, 28.8%), underscoring the severity and frequency of ocular surface disease in this breed (17). Quantitative changes (aqueous deficient dry eye, ADDE) and qualitative changes in the tear film (evaporative dry eye disease, EDED) directly contribute to the development of ocular surface diseases (1). We hypothesized that French Bulldogs suffer disproportionately from qualitative tear film disorders.

The main objective of this study was to collect normative data on the characteristics of the ocular surface of French Bulldogs in order to advance research in the field of breed differences. The focus was on the examination of the tear film, which was evaluated using the I.C.P. OSA-Vet® and a series of additional tear film tests. These data were compared with those of non-brachycephalic breeds. Another objective was to investigate the usefulness of measuring osmolarity as an additional parameter for distinguishing between ADDE and EDED in veterinary medicine.

The study population of this prospective study consisted of a group of healthy French Bulldogs and a control group of healthy non-brachycephalic breeds. French Bulldogs were included in the study if at least one eye was free of ocular disease and no topical eye medications had been administered within 5 days prior to the examination. The dogs were recruited either as inpatients with unilateral eye disease (with examination of the unaffected eye) or as clinically healthy pets from the same household. The control group consisted of mesocephalic and dolichocephalic dogs recruited from routine hereditary eye examinations, patients with unilateral eye disease, or healthy pets of the same household. In all cases, only eyes without clinical abnormalities and without recent topical treatment were included.

The owner’s consent for the evaluation of the anonymized data of their dog(s) and the photos taken for research purposes and publication was obtained prior to the examination. All examinations were carried out in accordance with the GERVO Declaration on the Use of Animals in Eye and Vision Research and with the approval of the Berlin State Secretariat for Health and Social Affairs under the number StN 024/22.

All dogs underwent a complete ophthalmic examination, including menace response, dazzle reflex, pupillary light reflex, palpebral reflex, slit lamp biomicroscopy (Kowa SL-17; Kowa, Japan), rebound tonometry (TonoVet, iCare, Finland), indirect ophthalmoscopy (Video Omega 2C; HEINE Optotechnik GmbH & Co. KG, Germany), a complete tear film analysis with the I.C.P. OSA-Vet® (SBM Sistemi, Italy), and a series of diagnostic tests for tear film deficiency, including osmolarity measurement, punctate fluorescein staining, lissamine green staining, Schirmer tear test I and Schirmer tear test II (all strips Madhu Instruments Pvt., Ltd., India). In 36 French Bulldogs and all non-brachycephalic dogs, tear film osmolarity was measured using the ScoutPro® osmolarity system (Trukera Medical, Southlake, USA). The tear film analysis with the I.C.P. OSA-Vet® was conducted first, followed by a 10 min rest period before the remaining examination.

To our knowledge, this is the first study to perform a comprehensive tear film analysis exclusively in French Bulldogs. While previous studies have examined certain brachycephalic breeds such as the Shih Tzu (8, 9) in great detail, the French Bulldog, despite being one of the most popular breeds in many countries (15, 16), has not been the subject of systematic tear film investigation to date. A number of studies (2, 4, 6, 7, 27) have included individual French Bulldogs in larger, heterogeneous cohorts. However, these data do not allow conclusions to be drawn about specific breeds. In contrast, the present study provides the first targeted insight into the tear film characteristics of this breed and highlights the clinical relevance of qualitative tear film abnormalities in otherwise ophthalmologically healthy dogs.

In this study, French Bulldogs showed a significantly higher STT-I value (p = 0.01) compared to non-brachycephalic breeds. The average STT-I values were 19.05 ± 4.00 mm/min in French Bulldogs and 16.88 ± 3.54 mm/min in the control group. This result contrasts with previous studies, which found STT-I values to be approximately 14% lower in brachycephalic dogs without clinical eye disease compared to non-brachycephalic dogs (1, 28). This discrepancy in these studies was primarily attributed to reduced corneal sensitivity and its negative effect on the afferent part of the tear reflex in brachycephalic breeds (1, 28). However, the results of this study suggest that this trend is not transferable to French Bulldogs. The increased STT-I in this breed is more likely related to increased reflex tear flow than to increased basal tear production. This assumption is supported by simultaneous evidence of a shortened NIBUT, reduced interferometry values, and a significantly increased TMH, all of which indicate EDED (1). It has been demonstrated that the process of increased tear evaporation leads to local dehydration of the cornea, activating both nociceptors and cold thermoreceptors in the cornea. This in turn stimulates the tear reflex (29, 30). The STT-I values observed in this study in French Bulldogs were similar to those reported by Li et al. for brachycephalic dogs (19.05 ± 4.00 mm/min vs. 18.2 ± 3.9 mm/min) (27), but lower than previously described for Shih Tzus (22.0 ± 5.5 mm/min) (9) and other brachycephalic dogs (20.1 ± 3.4 mm/min) (28). In contrast, STT-I values in the non-brachycephalic group showed a high correlation with data from the dolichocephalic subgroup (16.88 ± 3.54 mm/min vs. 16.8 ± 3.4 mm/min) from Li et al. (27).

A significant increase in TMH was observed in French Bulldogs compared to the control group (0.64 ± 0.20 mm vs. 0.31 ± 0.22 mm, p = 9 × 10⁻¹⁰), resulting in more than a twofold increase in TMH compared to control values. These results confirm the report by Li et al. on higher TMH in brachycephalic breeds (27). An increase in TMH could be a result of an impairment in the tear drainage system (31). However, although the tear drainage system of brachycephalic dogs is anatomically different, they consistently have an additional drainage into the oral cavity (32), thus supporting a normal tear drainage. The suggested normal range for TMH, measured with the I.C.P. OSA-Vet® in healthy dogs with normal STT-I, normal eyelid shape, and normal tear drainage system, is 0.53 ± 0.11 mm (18). This value is higher than the data reported by Li et al. of 0.27 ± 0.12 mm for mesocephalic breeds and 0.26 ± 0.09 mm for dolichocephalic breeds (27), 0.41 ± 0.21 mm by Kim et al. for Beagles, 0.32 ± 0.14 mm (33) and 0.36 ± 0.18 mm by Sonego et al. in their groups of non-brachycephalic breeds (34), the median values measured by Yoon et al. using optical coherence tomography of the anterior segment of the eye (0.32 mm and <0.32 mm) (35) and 0.31 ± 0.22 mm in this study. A study by Silva et al. found a TMH of 0.59 ± 0.29 mm in the right eye and 0.88 ± 0.27 mm in the left eye of healthy Shih Tzus (36). In a study on qualitative and quantitative tear tests in various brachycephalic breeds, Voitena et al. reported a TMH of 0.51 ± 0.17 (37). Zwiauer-Wolfbeisser et al. reported a TMH of 0.5 ± 0.2 mm, with the study population consisting of 58.3% brachycephalic breeds treated for distichiasis, making the population incomparable to the other studies (38). Peruccio et al. did not disclose the number of dogs and breeds examined (18). The available data on TMH shows that TMH is increased in brachycephalic breeds compared to non-brachycephalic breeds (27, 33–38). Our results suggest that the significant increase in TMH observed in French Bulldogs, combined with the significant decrease in NIBUT and lower interferometry values, provides convincing evidence of underlying EDED in this breed. A substantial variation in reported TMH is observed among studies involving a brachycephalic cohort. This variations includes lower (27, 37) and higher (36) values compared to the values reported in our study. This heterogeneity is likely attributable to variations in cohort composition. These findings underscore the importance of establishing breed-specific reference values for tear film diagnostics, particularly in brachycephalic dogs. The NIBUT was significantly lower in French Bulldogs than in the control group (7.39 ± 3.32 s. vs. 14.74 ± 5.92 s., p = 3.5*10⁻⁹). This result is consistent with other studies comparing brachycephalic breeds with non-brachycephalic breeds (27, 39). The shorter NIBUT could contribute to the higher STT-I, as reported by Faghihi et al. (39). The value for the control group was at the lower end of the reported normal range for tear film breakup time (TBUT) in healthy dogs (19.7 ± 5 to 21.53 ± 7.42 s.) (40). It was lower than the NIBUT in healthy beagles (33). Li et al. (27) and Voitena et al. (37) reported lower than previously described NIBUT values as well. Li et al. discussed that their lower NIBUT values may be due to environmental influences that contribute to a shorter break-up time, especially in the summer months (27). As reported in the results section, relative humidity was higher in the control group, whereas room temperature did not differ between groups. Because higher humidity reduces evaporative loss (41), this would be expected to prolong NIBUT in the control group and thereby exaggerate the observed between group difference. Although temperature can influence the tear film, temperature-related bias is unlikely in in this study due to comparable temperatures. Since all ambient values were within the typical indoor range (42), the influence is likely to be minor and probably not responsible for the general pattern of significantly shorter NIBUT in French Bulldogs. Nevertheless, residual bias cannot be completely ruled out.

The IF grade was significantly lower in French Bulldogs (p = 9*10⁻⁸) than in the control group, with the majority of French Bulldogs achieving an IF grade of B and the control group achieving a grade of E. The IF grade allows a direct estimation of the lipid layer thickness. The lipid layer is the outermost layer of the tear film, which prevents evaporation of the aqueous phase and stabilizes the tear film (43). It is produced by the Meibomian glands (43). Increased loss of the Meibomian glands or Meibomian gland dysfunction (MGD) leads to a reduction in the lipid layer (22) and thus to a low IF grade. This explains the correlation between a higher MGL and a lower NIBUT. A possible change in the composition of meibum due to MGD may also increase evaporative stress and provoke reflex tearing, which contributes to the higher STT-I and TMH, respectively (44). Another factor that must be taken into account is allergies, which are common in brachycephalic breeds (45). Studies in human medicine have shown that seasonal allergies alter the morphology of the Meibomian glands, promoting a change in lipid secretion and instability of the tear film (46, 47). In accordance with the previously mentioned findings, preliminary data reported by Marques et al. (ESVO 2025) revealed that atopic dogs suffering from allergic conjunctivitis exhibited lower IF grades in comparison with healthy controls. This observation supports to the hypothesis that ocular allergies are related to impairment of the lipid layer.

We have documented a higher MGL value in French Bulldogs, which could partly explain the lower IF grade. However, some patients had both a low IF grade and a low MGL percentage. Consequently, the MGL percentage is not the only factor influencing the IF grade. MGD is a multifactorial disease in which not only the end stage of glandular loss must be considered, but also qualitative and quantitative changes in the secreted lipid content. This may be due to hyposecretory, obstructive, or hypersecretory mechanisms (48, 49). As Ofri et al. have shown, there are differences in the amount of meibum secreted in different dog breeds (50). This also explains the prevalence of low IF grades in French Bulldogs. Another possible cause is incomplete closure of the eyelids during blinking, and the lower blink rate often observed in brachycephalic breeds (22). The reduced nerve fiber density of the cornea in brachycephalic dogs (51) is thought to be the cause of the reduced blink rate in these breeds (1). This leads to meibum not being distributed evenly across the cornea, which promotes destabilization of the tear film. Possible MGD in the French Bulldogs requires further investigation.

There are conflicting statements in the human literature as to whether only the upper eyelid (52), the lower eyelid (53), or both eyelids (54, 55) need to be measured for the correct diagnosis of Meibomian gland disease. Also, in the veterinary literature there is no consensus on which eyelid should be examined using meibography. Kim et al. performed meibography on both eyelids of Beagles and found it difficult to adequately expose the Meibomian gland area of the lower eyelid (33). Kitamura et al. did not examine the lower eyelid because this was found to be uncomfortable for their cohort of Shih Tzus (56). Li et al. examined the lower eyelids only because the patients showed discomfort during the examination of the upper eyelid (27). In our study, we selected the upper eyelid for meibography during the ophthalmological examination, as most dogs were more comfortable with it.

In this study, tear osmolarity was measured using the ScoutPro®. Veterinary clinical validation data for this device are limited (57). The osmolarity values of both groups were lower than previously reported normal values of TearLab® and I-PEN® Vet (58) and the ScoutPro® (57). Although reduced NIBUT and LLT values are typically associated with increased evaporation and increased osmolarity (29), the lower osmolarity of the French Bulldogs is presumably a consequence of the significantly increased STT-I and TMH values. This interpretation is supported by García-Resúa et al., who demonstrated a negative correlation between TMH and osmolarity (59). Given the lower osmolarity values in the control group, it can be concluded that a tear reflex triggered by handling can be ruled out, as no significant difference was found between the osmolarity of basal tears and reflex tears (60).

The clinical interpretability of osmolarity values remains unclear due to a lack of more published data for the ScoutPro® in dogs. After considering the above factors, it was determined that osmolarity within our specific cohort did not exhibit sensitivity to evaporation-related instability. Our observation shows that there was no difference in osmolarity between French Bulldogs and the control group. Consequently, osmolarity is considered an inadequate single marker for qualitative tear film deficiency in dogs. Further validation of the device and standardization of protocols are deemed necessary.

There were no differences between French Bulldogs and the control group regarding PFS and lissamine green staining. This is to be expected as STT-I was not decreased in animals included in this study. High PFS scores correlate with a lower STT-I (20). The weak negative correlation between PFS and IF grade aligns with the general findings, as PFS is regarded as a parameter for ocular health (22). Smith et al. reported higher lissamine green staining scores in dogs with lower STT-I values and shorter TBUT (21).

When interpreting the results, the limitations of the study must be taken into account. The assessment of the Meibomian glands was limited to non-contact infrared meibography. Image quality was partially limited by anatomical conditions or pigmentation. Although allergies are an important factor in MGD in humans, allergy testing was not performed in French Bulldogs, as this was beyond the scope of this study. All examinations and image analyses were performed by a single examiner without formal assessments of intra- or interobserver variability, which may limit reproducibility and introduce potential observer bias.

This study provides strong evidence that all clinically normal French Bulldogs in this cohort showed consistent indicators of EDED. Despite the absence of obvious clinical symptoms, the dogs showed a combination of shortened NIBUT, reduced LLT, increased TMH, and increased STT-I - characteristic features of a qualitative tear film disease. These findings suggest that EDED is not an isolated occurrence but rather a widespread condition within this breed. It is essential that routine ophthalmic examinations of French Bulldogs consider this underlying pathology.