Canine brachycephalic obstructive airway syndrome as a model for studying the effects of obstructive sleep apnea during pregnancy Authors: Roberto Rodrigues da Rosa Filho, Thayná Pantoja Gardés, Maíra Morales Brito, Leticia Lima de Almeida, Camila Infantosi Vannucchi Published: 2026 ABSTRACT AbstractBrachycephalic obstructive airway syndrome (BOAS) in dogs shares pathophysiological features with human obstructive sleep apnea, presenting different degrees of airway obstruction. Thus, the severity of the disorder may result in negative pregnancy consequences differently. This study aimed to compare the effects of BOAS severity throughout gestation on maternal clinical status, uteroplacental hemodynamics, and fetal development. Bitches were grouped according to BOAS symptomatic manifestation as mild (n = 5), severe BOAS (n = 5), and non‐BOAS groups (control; n = 5). Along gestation (early, mid, and late stage), females were examined for clinical evaluations, arterial blood acid–base balance, and obstetric B‐mode and uterine and umbilical artery Doppler velocimetry. Mild and severe BOAS had lower blood pressure, considered hypotension, and remained tachycardiac. All bitches had tachypnea, but mild BOAS group had higher (p < 0.05) respiratory rate. Peripheral oximetry decreased (p < 0.05) with time in severe BOAS, in a manner that females presented severe hypoxia (87.8% ± 4.6%) and hypocapnia (30.7 ± 2.37 mmHg) in late pregnancy. Blood glucose concentration was higher (p < 0.05) in mild BOAS (123.8 ± 5.79 mg/dL), in addition to persistent hyperglycemia until mid‐pregnancy (121.8 ± 7.37 mg/dL). Severe BOAS bitches had higher (p < 0.05) lactatemia and hyperlactatemia throughout gestation. Fetal abdominal diameter, femur length, and humerus length were lower (p < 0.05) in severe BOAS compared to control. Uterine artery hemodynamic indexes, indicative of vascular resistance, decreased (p < 0.05) over time only in control, whereas BOAS degree groups remained unchanged. Severe BOAS had higher (p < 0.05) umbilical artery resistance in mid and late pregnancy (0.84 ± 0.02 and 0.77 ± 0.01) compared to mild BOAS (0.76 ± 0.01 and 0.64 ± 0.04, respectively). In conclusion, the degree of respiratory impairment in BOAS dogs has an adverse impact on maternal health and the fetoplacental unit, despite tentative attempts of physiological adaptations during pregnancy to maintain gestational homeostasis. INTRODUCTION In dogs, brachycephalic obstructive airway syndrome (BOAS) is one of the most common conditions affecting specific breeds (e.g., English bulldog, French bulldog, and Pug), characterized, among other signs, by hypoxemia secondary to upper airway obstruction. 1 BOAS involves primary abnormalities of cranial structures and soft tissues of respiratory tract, as well as secondary dysfunctions that alter airflow dynamics. 2 Consequently, a set of clinical manifestations is established, including stridor, stertor, dyspnea, exercise intolerance, sleep apnea, cyanosis, and collapse. 3 Notably, BOAS in dogs shares pathophysiological features with human obstructive sleep apnea (OSA). 4 Among sleep‐related breathing disorders, obstructive apnea and central sleep apnea are primary types that lead to repeated interruptions in breathing during sleep. 5 Significant risk factors for development of these disorders include increased body mass index associated with obesity, gestational obesity, or relative weight gain during pregnancy. 6 , 7 These characteristics make BOAS pregnant bitches a relevant animal model for the study of human respiratory sleep disorders. 8 During pregnancy, sleep apnea can lead to significant maternal and/or gestational alterations, with intermittent hypoxia responsible for oxidative imbalances and inflammation, 9 , 10 as well as morphophysiological changes in vascular bed. 11 Indeed, women with sleep apnea exhibit alterations in placental architecture and morphology, in addition to increased oxidative stress. 10 However, despite oxygenation fluctuations associated with sleep apnea, the placenta has the capacity to mitigate negative effects of hypoxia‐reoxygenation phenomena. 12 On the contrary, Sanapo et al. 13 reported that sleep‐disordered breathing may affect fetal growth, resulting in neonates that are either small or large for gestational age. Moreover, sleep‐disordered breathing has been associated with prematurity and stillbirth, 14 suggesting compromised fetal end. Therefore, pregnancy outcomes may be linked not only to sleep‐disordered breathing but also to severity of the underlying respiratory dysfunction. Accordingly, using the animal model of pregnant BOAS bitches, we hypothesize that severity of respiratory dysfunction during gestation differentially impacts maternal health and fetoplacental unit. Thus, we aimed to compare effects of BOAS severity throughout gestation on maternal clinical status, uteroplacental hemodynamics, and fetal development. MATERIALS AND METHODS Animals and experimental groups As described in detail previously, using the same study population, 15 a prospective and observational study was conducted with 15 female dogs of various breeds, small to large sized, aged between 2 and 6 years, either primiparous or multiparous. All females used in this experiment were breeding females from commercial kennels or private owners and were healthy at the beginning of the experiment (at the time of pregnancy diagnosis), with no history of concurrent diseases and adequately dewormed and immunized with ethical vaccines. None of the females received medications other than those routinely administered. Females had body condition scores ranging from 4 to 6 (on a 1–9 scale), and bitches with superfecundation or excessively large litters were excluded from the experiment. All females were mated by stud dogs of the same breed; therefore, the litters consisted of puppies of defined breeds. To enroll females into the experimental groups and avoid employing exercise stress tests for ethical reasons, physical examination and an owner questionnaire were completed to identify the occurrence of the five main signs of BOAS, which were additionally rated using scores of 0, 1, or 2, and the final score (a sum of 0–10) was considered to be the BOAS severity degree, based on an adaptation of the Poncet–Dupré score 16 (Supplementary Material). In addition, to homogenize the experimental groups in terms of weight and size of the canine breeds, for each bitch included in one of the groups, another female of similar size was included in the opposite experimental group. To determine the ideal number of animals in each experimental group for statistical relevance, the power analysis was performed using the SAS Power and Sample Size (PSS) application (SAS Institute Inc., Cary, NC, USA), employing means and standard errors of selected variables (lactatemia, uterine artery resistance index, and heart rate), demonstrating a power test of 0.94, 0.74, and 0.93, respectively. Considering an acceptable statistical power >0.80, the number of animals used in this experiment ensures the scientific value of the findings. Then, bitches were grouped according to the symptomatic manifestation of BOAS: Non‐BOAS or control group (n = 5): dolicho or mesocephalic bitches without any BOAS clinical manifestations.Mild BOAS group (n = 5): only bitches with a severity score ≤5.Severe BOAS group (n = 5): bitches with a severity score ≥6. Non‐BOAS or control group (n = 5): dolicho or mesocephalic bitches without any BOAS clinical manifestations. Non‐BOAS or control group (n = 5): dolicho or mesocephalic bitches without any BOAS clinical manifestations. Mild BOAS group (n = 5): only bitches with a severity score ≤5. Mild BOAS group (n = 5): only bitches with a severity score ≤5. Severe BOAS group (n = 5): bitches with a severity score ≥6. Severe BOAS group (n = 5): bitches with a severity score ≥6. Additionally, all bitches were evaluated at three distinct time points at pregnancy: early gestation (33–36 days postovulation; estimated via vaginal cytology at proestrus and estrus and/or serum progesterone levels [>4 ng/mL], when matings or artificial inseminations started until diestrus), mid gestation (43–46 days of gestation), and late gestation (56–59 days of gestation). Clinical evaluations Bitches underwent clinical evaluations at the location where they were housed, including measurements of heart rate, respiratory rate, and body temperature. Systolic, mean, and diastolic blood pressure was measured using a noninvasive multiparametric monitor (Dixtal, DX2021) with an oscillometric method previously standardized for dogs, using cuffs adjusted to each dog's size. Peripheral oxygen saturation was assessed using a pulse oximeter (OxiMax N65) coupled to a SpO2 forehead sensor Max‐Fast, placed on the femoral artery region. Laboratory evaluations Blood samples were collected at the location where each bitch was housed without any prior sedation. Arterial blood was obtained by femoral artery puncture using preheparinized syringes. Blood gas analysis was performed using a portable clinical analyzer (i‐STAT, Abbott), with temperature correction, to assess pH, partial pressure of oxygen (pO2), carbon dioxide (pCO2), bicarbonate concentration (HCO3 −), base excess (BE), oxygen saturation (sO2), and lactatemia. After approximately 12 h of fasting, venous blood samples were obtained to measure blood glucose levels using a glucometer (Accutrend Plus—Roche Diagnostics). Ultrasonographic evaluation Ultrasonographic evaluations were performed transabdominally using a Mindray M5 VET ultrasound system equipped with convex and linear multifrequency transducers (from 5 to 7.5 MHz), adjusted based on the size of the female dog. Fetal biometry was conducted using conventional B‐mode ultrasound, focusing on a single fetus per bitch, preferably the one located at the cranial extremity of the uterine horns. The following parameters were measured after the respective structures were identified: embryonic vesicle diameter, placental thickness, biparietal diameter, femur length, humerus length, abdominal diameter, and thoracic diameter, according to Feliciano et al. 17 Placental vascularization was semiquantitatively assessed using color Doppler and scored from 1 to 3 (1 = minimal, 2 = moderate, 3 = maximal vascularization). Maternal‐fetal hemodynamic assessment was performed using color Doppler to map and identify the uterine and umbilical arteries (near the placental surface). Hemodynamic indices were obtained using pulsed (spectral) Doppler and automatically calculated by the system's software. For each artery, three stable waveforms were recorded and averaged. Sample volume was set between 1 and 2 mm, centrally positioned within the vessel, and insonation angle was manually adjusted to remain below 60°. The following parameters were obtained: systolic/diastolic ratio (S/D), resistance index (RI), pulsatility index (PI), peak systolic velocity (PS), end‐diastolic velocity (ED), and time‐averaged maximum velocity (TAMAX). Statistical analysis Data analysis was performed using the SAS System for Windows (SAS Institute Inc., USA). Data were tested for normality of residuals and homogeneity of variances. When assumptions were not met, appropriate transformations were applied. If transformations were not successful, the NPAR1WAY procedure for nonparametric analysis of variance was used. Differences between treatments were analyzed by means of parametric and nonparametric tests, according to the residual normality (Gaussian distribution) and homogeneity of variance. As a longitudinal study, that is, measuring the same subjects at multiple time points, we used the repeated‐measures analysis of variance (ANOVA) for parametric variables and the Friedman test for nonparametric variables. The post‐hoc least significant difference (LSD) test was used to compare differences over time and between groups for parametric variables. For nonparametric variables, the Wilcoxon test with Bonferroni correction was applied (nonparametric variables: oxygen saturation sO2, base excess, diastolic blood pressure, and oximetry). Pearson's correlation analysis with Bonferroni correction was performed for response variables, considering the experimental group as the main factor and merging all time points. Only correlations maintaining a corrected p‐value <0.05 alpha level were considered statistically significant. Results were described as untransformed mean ± standard error of the mean (SEM) for parametrical variables and medians (quartil Q1–quartil Q3) for nonparametrical variables, with a significance level set at 5%. RESULTS Clinical evaluation At early pregnancy, mild and severe BOAS bitches had higher heart rate than the control group, without any further differences throughout gestation (Table 1). However, BOAS bitches remained tachycardiac throughout gestation, whereas tachycardia in non‐BOAS bitches occurred only from mid‐pregnancy onward. All bitches had tachypnea throughout the gestational period, but the mild BOAS group had consistently higher respiratory rate in early and late pregnancy, not different from severe BOAS group at final term of pregnancy (Table 1). TABLE 1Mean ± standard error (95% CLM*) and median (interquartile range—Q1–Q3) of clinical examination and blood acid–base of bitches of the non‐brachycephalic, mild, and severe brachycephalic obstructive airway syndrome (BOAS) groups throughout gestation.Moment of evaluationNon‐BOAS groupMild BOAS groupSevere BOAS groupReference rangeBody temperature (°C)Early pregnancy 38.0 ± 0.23 (37.3–38.6) 38.2 ± 0.18 (37.6–38.7) 38.1 ± 0.16 (37.6–38.5) 37.5–39.2Mid‐pregnancy 37.9 ± 0.15 (37.4–38.3) 38.0 ± 0.12 (37.6–38.3) 37.7 ± 0.22 (37.1–38.3) Late pregnancy 37.8 ± 0.19 (37.2–38.2) 38.2 ± 0.23 (37.6–38.9) 37.7 ± 0.22 (37.0–38.2) Respiratory rate (mpm)Early pregnancy 50 ± 6.4b (32.6–68.1) 86 ± 17.1a (38.1–133.0) 48 ± 4.69b (3.7–60.8) 10–30Mid‐pregnancy 46 ± 4.31 (34.4–58.3) 100 ± 21.8 (39.3–160.2) 51 ± 2.52 (44.4–58.4) Late pregnancy 42 ± 8.99b (17.4–67.3) 81.8 ± 15.1a (39.9–123.6) 54 ± 8.04ab (31.6–76.3) Heart rate (bpm)Early pregnancy 106 ± 5.48b (90.9–121.4) 145 ± 10.8a (115.6–175.9) 141 ± 10.2a (112.6–169.7) 70–120Mid‐pregnancy 125 ± 10.57 (95.6–1254.3) 148 ± 17.8 (98.8–197.9) 128 ± 6.93 (109.1–147.6) Late pregnancy 124 ± 2.33 (118.3–131.2) 146 ± 19.9 (90.9–201.8) 130 ± 10.6 (100.8–159.9) Glucose concentration (mg/dL)Early pregnancy 97.6 ± 8.74b (73.3–121.8) 123.8 ± 5.79a (107.7–139.8) 102.0 ± 9.00ab (76.9–127.0) 66–120Mid‐pregnancy 106.0 ± 5.89 (89.6–122.3) 121.8 ± 7.37 (101.3–142.2) 108.6 ± 4.84 (95.1–122.0) Late pregnancy 107.6 ± 6.79 (88.7–126.4) 107.2 ± 2.22 (101.0–113.3) 100.0 ± 4.67 (87.0–112.9) Bicarbonate concentration (mmol/L)Early pregnancy 20.1 ± 1.31 (15.9–24.3) 16.1 ± 1.46 (15.1–18.1) 16.1 ± 1.58 (11.0–21.1) 19–26Mid‐pregnancy 18.9 ± 1.31 (15.3–22.5) 16.7 ± 0.65 (14.9–18.5) 16.7 ± 1.37 (12.9–20.5) Late pregnancy 17.1 ± 0.82 (14.8–19.3) 16.6 ± 0.52 (12.1–20.2) 17.3 ± 1.08 (14.3–20.3) Oxygen saturation (sO2; %)Early pregnancy95.5 (95–96.5)96 (96–97)96 (93–97)>95Mid‐pregnancy95.5 (95–96)95 (94–95)95 (95–95)Late pregnancy93 (93–94)94 (94–95)92 (88–94)pHEarly pregnancy 7.41 ± 0.005 (7.39–7.43) 7.37 ± 0.02 (7.30–7.43) 7.36 ± 0.01 (7.31–7.41) 7.35–7.46Mid‐pregnancy 7.39 ± 0.01 (7.35–7.42) 7.36 ± 0.01 (7.31–7.40) 7.36 ± 0.01 (7.31–7.41) Late pregnancy 7.37 ± 0.01 (7.32–7.41) 7.39 ± 0.02 (7.32–7.45) 7.36 ± 0.01 (7.32–7.39) Partial pressure of CO2 (mmHg)Early pregnancy 31.9 ± 1.69 (26.4–37.3) 28.0 ± 1.77 (23.1–32.9) 28.4 ± 3.05 (18.6–38.1) 30.8–42.8Mid‐pregnancy 31.3 ± 2.07 (25.5–37.1) 29.5 ± 0.23 (28.9–30.2) 29.6 ± 2.31 (23.2–26.1) Late pregnancy 30.1 ± 2.36 (23.5–36.7) 27.5 ± 0.99 (24.8–30.3) 30.7 ± 2.37 (24.1–37.3) Partial pressure of O2 (mmHg)Early pregnancy 81.2 ± 3.59 (69.8–92.7) 85.6 ± 7.76 (64.0–107.1) 79.2 ± 9.62 (48.6–109.8) 80.9–103.3Mid‐pregnancy 82.0 ± 5.81 (65.8–98.1) 81.2 ± 3.95 (70.2–92.1) 79.2 ± 1.06 (76.2–82.1) Late pregnancy 77.2 ± 5.33 (62.3–92.0) 76.4 ± 4.73 (63.2–89.5) 63.6 ± 7.12 (43.8–83.3) Note: a,bDifferent superscript lowercase letters in the same line represent significant (p < 0.05) difference between groups within the same moment of evaluation. Marked table cells represent data outside the reference range for dogs.Abbreviation: CLM*, confidence limits for the mean. Mean ± standard error (95% CLM*) and median (interquartile range—Q1–Q3) of clinical examination and blood acid–base of bitches of the non‐brachycephalic, mild, and severe brachycephalic obstructive airway syndrome (BOAS) groups throughout gestation. 38.0 ± 0.23 (37.3–38.6) 38.2 ± 0.18 (37.6–38.7) 38.1 ± 0.16 (37.6–38.5) 37.9 ± 0.15 (37.4–38.3) 38.0 ± 0.12 (37.6–38.3) 37.7 ± 0.22 (37.1–38.3) 37.8 ± 0.19 (37.2–38.2) 38.2 ± 0.23 (37.6–38.9) 37.7 ± 0.22 (37.0–38.2) 50 ± 6.4b (32.6–68.1) 86 ± 17.1a (38.1–133.0) 48 ± 4.69b (3.7–60.8) 46 ± 4.31 (34.4–58.3) 100 ± 21.8 (39.3–160.2) 51 ± 2.52 (44.4–58.4) 42 ± 8.99b (17.4–67.3) 81.8 ± 15.1a (39.9–123.6) 54 ± 8.04ab (31.6–76.3) 106 ± 5.48b (90.9–121.4) 145 ± 10.8a (115.6–175.9) 141 ± 10.2a (112.6–169.7) 125 ± 10.57 (95.6–1254.3) 148 ± 17.8 (98.8–197.9) 128 ± 6.93 (109.1–147.6) 124 ± 2.33 (118.3–131.2) 146 ± 19.9 (90.9–201.8) 130 ± 10.6 (100.8–159.9) 97.6 ± 8.74b (73.3–121.8) 123.8 ± 5.79a (107.7–139.8) 102.0 ± 9.00ab (76.9–127.0) 106.0 ± 5.89 (89.6–122.3) 121.8 ± 7.37 (101.3–142.2) 108.6 ± 4.84 (95.1–122.0) 107.6 ± 6.79 (88.7–126.4) 107.2 ± 2.22 (101.0–113.3) 100.0 ± 4.67 (87.0–112.9) 20.1 ± 1.31 (15.9–24.3) 16.1 ± 1.46 (15.1–18.1) 16.1 ± 1.58 (11.0–21.1) 18.9 ± 1.31 (15.3–22.5) 16.7 ± 0.65 (14.9–18.5) 16.7 ± 1.37 (12.9–20.5) 17.1 ± 0.82 (14.8–19.3) 16.6 ± 0.52 (12.1–20.2) 17.3 ± 1.08 (14.3–20.3) 7.41 ± 0.005 (7.39–7.43) 7.37 ± 0.02 (7.30–7.43) 7.36 ± 0.01 (7.31–7.41) 7.39 ± 0.01 (7.35–7.42) 7.36 ± 0.01 (7.31–7.40) 7.36 ± 0.01 (7.31–7.41) 7.37 ± 0.01 (7.32–7.41) 7.39 ± 0.02 (7.32–7.45) 7.36 ± 0.01 (7.32–7.39) 31.9 ± 1.69 (26.4–37.3) 28.0 ± 1.77 (23.1–32.9) 28.4 ± 3.05 (18.6–38.1) 31.3 ± 2.07 (25.5–37.1) 29.5 ± 0.23 (28.9–30.2) 29.6 ± 2.31 (23.2–26.1) 30.1 ± 2.36 (23.5–36.7) 27.5 ± 0.99 (24.8–30.3) 30.7 ± 2.37 (24.1–37.3) 81.2 ± 3.59 (69.8–92.7) 85.6 ± 7.76 (64.0–107.1) 79.2 ± 9.62 (48.6–109.8) 82.0 ± 5.81 (65.8–98.1) 81.2 ± 3.95 (70.2–92.1) 79.2 ± 1.06 (76.2–82.1) 77.2 ± 5.33 (62.3–92.0) 76.4 ± 4.73 (63.2–89.5) 63.6 ± 7.12 (43.8–83.3) Note: a,bDifferent superscript lowercase letters in the same line represent significant (p < 0.05) difference between groups within the same moment of evaluation. Marked table cells represent data outside the reference range for dogs. Abbreviation: CLM*, confidence limits for the mean. At late pregnancy, severe BOAS bitches had lowest systolic blood pressure, considered hypotension, whereas mild BOAS group had lower diastolic and medium blood pressure (hypotension) (Figure 1A,B). Peripheral oximetry decreased continuously with time in severe BOAS bitches, with the lowest SpO2 observed at late pregnancy, indicative of severe hypoxia (Figure 1C). In addition, bitches severely affected by BOAS had peripheral hypoxia throughout pregnancy, whereas hypoxemia (sO2) occurred as early as mid‐pregnancy in BOAS bitches compared to a hypoxemic status only at late pregnancy in nonaffected females (Table 1). (A) Mean ± standard error of bitches' systolic and medium blood pressure during late pregnancy. (B) Median (interquartile range—Q1–Q3) of the diastolic blood pressure during late pregnancy. (C) Median (interquartile range—Q1–Q3) of the oximetry in non‐BOAS, mild, and severe BOAS groups throughout gestation. Blue rectangle bounded by dotted lines represents the normal range in dogs for each variable. a,bDifferent superscript lowercase letters represent significant (p < 0.05) difference between groups within the same moment of evaluation. A,BDifferent superscript capital letters represent significant (p < 0.05) difference between moments of evaluation within the group. Laboratory evaluation At early pregnancy, blood glucose concentration was higher in mild BOAS bitches, but not different from the severe BOAS group, in addition to a slight hyperglycemic state persisting until mid‐pregnancy (Table 1). Both at early and late pregnancy, severe BOAS bitches had higher lactatemia, except for the control group at term pregnancy (Figure 2A). Severely affected bitches had hyperlactatemia throughout the entire gestational period. Interestingly, blood base excess decreased continuously over time in the control group, whereas it remained unchanged in mild and severe BOAS groups, despite being above the reference range irrespective of the pregnancy stage (Figure 2B). However, bitches' blood bicarbonate concentration remained below the reference range during pregnancy, except for early‐pregnant non‐BOAS bitches (Table 1). Although severe BOAS bitches had hypoxemia (pO2), they also presented hypocapnia (pCO2) along all pregnancy stages, so as mild‐affected females (Table 1). Conversely, hypoxemia was verified in mild and non‐BOAS bitches only at late pregnancy, accompanied by hypocapnia (Table 1). (A) Mean ± standard error of bitches' blood lactate concentration and (B) median (interquartile range—Q1–Q3) of the blood base excess in non‐BOAS, mild, and severe BOAS groups throughout gestation. Blue rectangle bounded by dotted lines represents the normal range in dogs for each variable. a,bDifferent superscript lowercase letters represent significant (p < 0.05) difference between groups within the same moment of evaluation. A,BDifferent superscript capital letters represent significant (p < 0.05) difference of base excess between groups at early‐pregnancy. Ultrasonographic evaluation Severe BOAS bitches had consistently the lowest fetal abdominal diameter, femur length, and humerus length. However, mild BOAS degree was only different from control bitches regarding humerus length at mid‐pregnancy (Table 2). TABLE 2Mean ± standard error (95% CLM*) of uterine and umbilical artery Doppler velocimetry, and fetal and placental biometry in bitches of the control, mild, and severe brachycephalic obstructive airway syndrome (BOAS) groups throughout gestation.Moment of evaluationNon‐BOAS groupMild BOAS groupSevere BOAS groupUterine artery systolic/diastolic ratioEarly pregnancy 3.1 ± 0.19A (2.5–3.6) 2.5 ± 0.31 (1.5–3.4) 3.4 ± 0.55 (1.6–5.2) Mid‐pregnancy 2.7 ± 0.04AB (2.5–2.8) 2.2 ± 0.10 (1.9–2.4) 2.8 ± 0.63 (0.1–5.5) Late pregnancy 2.5 ± 0.19B (2.0–3.1) 2.2 ± 0.28 (0.9–3.4) 2.6 ± 0.18 (2.1–3.1) Uterine artery peak systolic velocity (cm/s)Early pregnancy 130.3 ± 10.18a (101.9–158.5) 148.5 ± 14.49a (108.7–188.2) 79.5 ± 13.69b (35.9–123.0) Mid‐pregnancy 111.0 ± 16.89 (57.2–164.8) 127.4 ± 10.89 (97.1–157.6) 95.4 ± 21.17 (4.2–186.4) Late pregnancy 146.7 ± 10.22 (118.3–175.1) 110.7 ± 14.69 (47.5–173.9) 109.1 ± 25.78 (37.5–180.6) Umbilical artery pulsatility indexEarly pregnancy 1.3 ± 0.03b (1.1–1.4) 1.4 ± 0.18a (0.6–2.1) 1.3 ± 0.02b (1.3–1.4) Mid‐pregnancy 1.4 ± 0.03 (1.2–1.6) 1.4 ± 0.07 (1.2–1.6) 1.1 ± 0.26 (0.9–1.8) Late pregnancy 1.3 ± 0.13ab (0.9–1.7) 1.1 ± 0.12b (0.7–1.5) 1.5 ± 0.09a (1.3–1.6) Umbilical artery resistance indexEarly pregnancy 0.84 ± 0.01A (0.7–0.9) 0.83 ± 0.02A (0.7–0.9) 0.87 ± 0.01A (0.8–0.9) Mid‐pregnancy 0.77 ± 0.02Abb (0.7–0.8) 0.76 ± 0.01Ab (0.7–0.8) 0.84 ± 0.02Aa (0.7–0.9) Late pregnancy 0.70 ± 0.02Bab (0.6–0.8) 0.64 ± 0.04Bb (0.5–0.8) 0.77 ± 0.01Ba (0.7–0.8) Umbilical systolic/diastolic ratioEarly pregnancy 6.66 ± 0.62A (3.9–9.3) 6.36 ± 0.95A (2.2–10.4) 8.09 ± 0.74A (4.8–11.3) Mid‐pregnancy 4.84 ± 0.64ABab (3.0–6.6) 4.42 ± 0.40Ab (3.3–5.5) 6.97 ± 1.09Aa (3.4–10.4) Late pregnancy 3.57 ± 0.36Bab (2.5–4.6) 2.89 ± 0.29Bb (1.9–3.8) 4.42 ± 0.30Ba (3.5–5.2) Umbilical artery end‐diastolic velocity (cm/s)Early pregnancy 2.3 ± 0.26B (1.2–3.4) 5.9 ± 2.16B (3.4–15.2) 2.9 ± 0.53B (0.6–5.2) Mid‐pregnancy 5.4 ± 1.17A (2.1–8.6) 6.1 ± 1.36B (2.2–9.8) 6.1 ± 1.70A (0.7–11.5) Late pregnancy 6.6 ± 0.76Ab (4.4–8.6) 14.0 ± 3.41Aa (3.1–24.8) 9.3 ± 1.55Aab (4.9–13.6) Fetal abdominal diameter (cm)Mid‐pregnancy 3.06 ± 0.19Aa (2.5–3.6) 2.57 ± 0.25Aab (1.8–3.3) 2.45 ± 0.09Ab (2.2–2.7) Late pregnancy 4.10 ± 0.22Ba (3.4–4.7) 3.67 ± 0.12Bab (3.3–4.0) 3.48 ± 0.20Bb (2.9–4.0) Fetal femur length (cm)Mid‐pregnancy 1.11 ± 0.12Aa (0.7–1.4) 0.88 ± 0.07Aab (0.7–1.1) 0.67 ± 0.06Ab (0.5–0.8) Late pregnancy 1.71 ± 0.09Ba (1.4–1.9) 1.45 ± 0.15Bab (1.0–1.9) 1.26 ± 0.13Bb (0.8–1.7) Fetal humerus length (cm)Mid‐pregnancy 0.95 ± 0.08Aa (0.7–1.2) 0.77 ± 0.05Ab (0.6–0.9) 0.69 ± 0.01Ab (0.6–0.7) Late pregnancy 1.50 ± 0.08Ba (1.3–1.7) 1.30 ± 0.16Bab (0.8–1.8) 1.07 ± 0.08Bb (0.8–1.3) Placenta thickness (cm)Early pregnancy 0.48 ± 0.05B (0.3–0.6) 0.47 ± 0.04B (0.3–0.6) 0.48 ± 0.03C (0.4–0.6) Mid‐pregnancy 0.60 ± 0.03AB (0.5–0.7) 0.63 ± 0.03A (0.5–0.7) 0.64 ± 0.02B (0.5–0.7) Late pregnancy 0.65 ± 0.04Ab (0.5–0.8) 0.64 ± 0.03Ab (0.5–0.7) 0.77 ± 0.04Aa (0.6–0.9) Placenta vascularization score (0–3)Early pregnancy 1.4 ± 0.24B (0.7–2.1) 1.25 ± 0.25C (0.4–2.0) 1.2 ± 0.2B (0.6–1.7) Mid‐pregnancy 2.25 ± 0.25A (1.4–3.0) 2 ± 0B — 1.75 ± 0.25B (0.9–2.5) Late pregnancy 3 ± 0A — 2.75 ± 0.25A (1.9–3.5) 2.6 ± 0.24A (1.9–3.3) Note: a,bDifferent superscript lowercase letters in the same line represent significant (p < 0.05) difference between groups within the same moment of evaluation.Abbreviation: CLM*, confidence limits for the mean. Mean ± standard error (95% CLM*) of uterine and umbilical artery Doppler velocimetry, and fetal and placental biometry in bitches of the control, mild, and severe brachycephalic obstructive airway syndrome (BOAS) groups throughout gestation. 3.1 ± 0.19A (2.5–3.6) 2.5 ± 0.31 (1.5–3.4) 3.4 ± 0.55 (1.6–5.2) 2.7 ± 0.04AB (2.5–2.8) 2.2 ± 0.10 (1.9–2.4) 2.8 ± 0.63 (0.1–5.5) 2.5 ± 0.19B (2.0–3.1) 2.2 ± 0.28 (0.9–3.4) 2.6 ± 0.18 (2.1–3.1) 130.3 ± 10.18a (101.9–158.5) 148.5 ± 14.49a (108.7–188.2) 79.5 ± 13.69b (35.9–123.0) 111.0 ± 16.89 (57.2–164.8) 127.4 ± 10.89 (97.1–157.6) 95.4 ± 21.17 (4.2–186.4) 146.7 ± 10.22 (118.3–175.1) 110.7 ± 14.69 (47.5–173.9) 109.1 ± 25.78 (37.5–180.6) 1.3 ± 0.03b (1.1–1.4) 1.4 ± 0.18a (0.6–2.1) 1.3 ± 0.02b (1.3–1.4) 1.4 ± 0.03 (1.2–1.6) 1.4 ± 0.07 (1.2–1.6) 1.1 ± 0.26 (0.9–1.8) 1.3 ± 0.13ab (0.9–1.7) 1.1 ± 0.12b (0.7–1.5) 1.5 ± 0.09a (1.3–1.6) 0.84 ± 0.01A (0.7–0.9) 0.83 ± 0.02A (0.7–0.9) 0.87 ± 0.01A (0.8–0.9) 0.77 ± 0.02Abb (0.7–0.8) 0.76 ± 0.01Ab (0.7–0.8) 0.84 ± 0.02Aa (0.7–0.9) 0.70 ± 0.02Bab (0.6–0.8) 0.64 ± 0.04Bb (0.5–0.8) 0.77 ± 0.01Ba (0.7–0.8) 6.66 ± 0.62A (3.9–9.3) 6.36 ± 0.95A (2.2–10.4) 8.09 ± 0.74A (4.8–11.3) 4.84 ± 0.64ABab (3.0–6.6) 4.42 ± 0.40Ab (3.3–5.5) 6.97 ± 1.09Aa (3.4–10.4) 3.57 ± 0.36Bab (2.5–4.6) 2.89 ± 0.29Bb (1.9–3.8) 4.42 ± 0.30Ba (3.5–5.2) 2.3 ± 0.26B (1.2–3.4) 5.9 ± 2.16B (3.4–15.2) 2.9 ± 0.53B (0.6–5.2) 5.4 ± 1.17A (2.1–8.6) 6.1 ± 1.36B (2.2–9.8) 6.1 ± 1.70A (0.7–11.5) 6.6 ± 0.76Ab (4.4–8.6) 14.0 ± 3.41Aa (3.1–24.8) 9.3 ± 1.55Aab (4.9–13.6) 3.06 ± 0.19Aa (2.5–3.6) 2.57 ± 0.25Aab (1.8–3.3) 2.45 ± 0.09Ab (2.2–2.7) 4.10 ± 0.22Ba (3.4–4.7) 3.67 ± 0.12Bab (3.3–4.0) 3.48 ± 0.20Bb (2.9–4.0) 1.11 ± 0.12Aa (0.7–1.4) 0.88 ± 0.07Aab (0.7–1.1) 0.67 ± 0.06Ab (0.5–0.8) 1.71 ± 0.09Ba (1.4–1.9) 1.45 ± 0.15Bab (1.0–1.9) 1.26 ± 0.13Bb (0.8–1.7) 0.95 ± 0.08Aa (0.7–1.2) 0.77 ± 0.05Ab (0.6–0.9) 0.69 ± 0.01Ab (0.6–0.7) 1.50 ± 0.08Ba (1.3–1.7) 1.30 ± 0.16Bab (0.8–1.8) 1.07 ± 0.08Bb (0.8–1.3) 0.48 ± 0.05B (0.3–0.6) 0.47 ± 0.04B (0.3–0.6) 0.48 ± 0.03C (0.4–0.6) 0.60 ± 0.03AB (0.5–0.7) 0.63 ± 0.03A (0.5–0.7) 0.64 ± 0.02B (0.5–0.7) 0.65 ± 0.04Ab (0.5–0.8) 0.64 ± 0.03Ab (0.5–0.7) 0.77 ± 0.04Aa (0.6–0.9) 1.4 ± 0.24B (0.7–2.1) 1.25 ± 0.25C (0.4–2.0) 1.2 ± 0.2B (0.6–1.7) 2.25 ± 0.25A (1.4–3.0) 2 ± 0B — 1.75 ± 0.25B (0.9–2.5) 3 ± 0A — 2.75 ± 0.25A (1.9–3.5) 2.6 ± 0.24A (1.9–3.3) Note: a,bDifferent superscript lowercase letters in the same line represent significant (p < 0.05) difference between groups within the same moment of evaluation. Abbreviation: CLM*, confidence limits for the mean. As expected, placenta thickness increased over time, but significantly from mid‐pregnancy onward for BOAS degree groups, whereas control‐placenta hypertrophied homogenously throughout gestation (Table 2). It is worth noting that severe BOAS group had greatest placenta thickness at late pregnancy. Color Doppler placental evaluation revealed an increased vascularization pattern throughout pregnancy. However, a significant increase occurred in control and severe BOAS groups at mid‐pregnancy, whereas mild BOAS bitches showed an additional increment in placental vascularization at late pregnancy (Table 2). Uterine artery hemodynamic indexes decreased over time only in the control group, whereas BOAS degree groups remained unchanged (Table 2). It is interesting to note that severe BOAS bitches had higher uterine artery PI at early pregnancy compared to the mild BOAS group. In addition, blood velocimetry of the uterine artery was higher in mild than in severe BOAS bitches at early pregnancy, and was not different from the control group (Table 2). Umbilical artery RI and S/D showed a progressive vasodilation pattern throughout gestation, however with an earlier occurrence in the control group (Table 2). Severe BOAS group had highest umbilical artery resistance, pulsatility, and S/D at mid‐ and/or late pregnancy compared to mild BOAS bitches, not different from the control group (Table 2). Conversely, umbilical artery blood flow velocity at late pregnancy was higher in mild and severe BOAS groups compared to the control bitches and also an earlier significant increase in ED at mid‐pregnancy (Table 2). Correlation analysis In the mild BOAS group, positive correlations were found between blood pCO2 and uterine artery PS (r = 0.65; p = 0.02), EDV (r = 0.77; p = 0.003) and TAMAX (r = 0.73; p = 0.006). Additionally, RI and S/D of the umbilical artery were positively correlated with medium blood pressure (r = 0.67; p = 0.01 and r = 0.70; p = 0.01, respectively). In the severe BOAS group, uterine artery PS, EDV, and TAMAX were positively correlated with blood pH (r = 0.61, r = 0.60, and r = 0.60, respectively; p = 0.03). DISCUSSION This study presents the effects of severity of BOAS on pregnancy, evaluating maternal and fetal health in dogs, as a comparative model for studying obstructive airway conditions in pregnant women, more specifically OSA. Obstructive airway syndrome in brachycephalic dogs primarily manifests with respiratory clinical signs. However, it is a progressive condition that can result in multiorgan dysfunction underlined by hypoxia. 18 Similarly, OSA is caused by intermittent oxygen deprivation resulting from upper airway blockage during sleep and is associated with numerous dysfunctions, including neurological cognitive deficits. However, in both BOAS and OSA, depending on the extent of airway obstruction (partial or complete), the impact on intermittent hypoxia can be of different degrees, with negative pregnancy consequences differently. In fact, mild BOAS females had a slightly hyperglycemic profile until mid‐pregnancy, higher than that of nonaffected ones. Although alteration in energy homeostasis can occur punctually during pregnancy, altered blood glucose concentration was not observed in severely affected bitches in this study, suggesting that BOAS severity implies differentiated metabolic response in pregnant bitches. During pregnancy, OSA women have a higher risk of insulin intolerance and type 2 diabetes 19 ; however, the influence of different degrees of severity remains unknown for glucose control. However, situations of stress during pregnancy can result in metabolic disturbance, which can be the case of both the hypoxic stress of BOAS in dogs and the decreased quality of sleep in OSA women. On the contrary, the chronic effects of severe BOAS in bitches can lead to a situation of tolerance to stress hyperglycemia. Thus, it is possible that pregnant women share the same mechanism of stress tolerance, with those mildly affected by OSA having a worse prognosis for metabolic decompensation compared to those severely affected. Gestation itself may have a negative impact on blood oxygenation and blood acid–base balance, as shown by a tachypneic state throughout pregnancy, along with hypoxia, low blood bicarbonate at late pregnancy, and a reducing BE profile in non‐BOAS bitches. Interestingly, bitches with mild BOAS clinical signs managed to maintain adequate pO2 through hyperventilation, a compensatory reflex mechanism not demonstrated efficiently by severely affected bitches. It was notable that severely affected bitches exhibited decreasing oximetry throughout pregnancy to the point that, at final gestation, peripheral oxygen was extremely lower. It is thus suggested that the degree of gestational hypoxia also varies based on the severity of the anatomical abnormality that impairs upper airway airflow. Severe changes in respiratory patterns were not sufficient to maintain acid–base balance, as bitches of the severe BOAS group had marked hyperlactatemia throughout gestation, and also higher lactatemia compared to mild BOAS females at early and late pregnancy. The primary origin of hyperlactatemia as an acid–base disorder in dogs is respiratory disease coupled with hypoxia 20 ; even so, no female presented acidemia during pregnancy. Therefore, blood buffering mechanisms in dogs are thought to be efficient in maintaining acid–base balance, at the expense of greater blood bicarbonate consumption. Conversely, OSA women are prone to respiratory acidosis during gestation, 21 highlighting a species‐specific difference. We can thus assume that pregnant bitches with severe signs of BOAS face similar hypoxemic challenges as OSA women, but present a more efficient compensatory reflex of acidosis through hyperventilation and secondary hypocapnia. Another difference between women and dogs regards to systemic cardiovascular effects. Pregnant OSA women experience systemic hypertension resulting from intermittent hypoxia, 22 whereas BOAS bitches faced medium and systolic hypotension at late pregnancy, as well as tachycardia. Arteriolar relaxation resulting in low vascular resistance and blood pressure in BOAS bitches suggests a vascular rearrangement that prioritizes blood flow to pregnant uterus, in an attempt to prevent negative effects of maternal hypoxia to the developing fetoplacental unit. Positive correlation between BOAS maternal blood pressure and umbilical artery resistance reinforces the hypothesis that hypotension increases the blood volume reaching the growing fetuses. In fact, although bitches lacking signs of BOAS increased placenta thickness only at late gestation, placental development occurred earlier at mid‐pregnancy in BOAS‐affected females. Interestingly, severe BOAS had greater placental thickness at late pregnancy, suggesting that placental hypertrophy may be considered a compensatory mechanism for severe BOAS hypoxia. However, even with all systemic vascular changes and placental remodeling, BOAS severity had a negative impact on fetal growth from mid‐pregnancy onward. These data reinforce the suggestion that the severity of the respiratory disorder affects gestational development differently, being indeed more worrying in severely affected bitches. On the contrary, data correlating OSA women and fetal growth restriction or low‐birth‐weight neonates are inconsistent. Although there is insufficient scientific evidence supporting an association between OSA and impaired fetal growth, 23 studies suggest negative impacts of OSA on fetal weight and growth. 10 , 24 , 25 Thus, our findings may constitute an additional scientific contribution to the hypothesis of impaired fetal growth during gestation in females affected by obstructive airway disorder, although further studies are warranted to accurately elucidate the underlying mechanisms. Altered blood flow within fetal–placental unit can be a possible explanation behind the mechanism by which BOAS bitches present fetal growth restriction. Although no differences were observed in placental vascularization among groups, timely placental vascular changes occurred differently. Placental vascularization score increased significantly earlier in non‐BOAS bitches compared to late‐pregnancy augmentation in severe BOAS. Throughout normal pregnancy, the uterine artery shows a progressive decline in vascular resistance as a means to support fetal and placental growth through increased blood flow. Indeed, such uterine hemodynamic profile was observed in non‐BOAS throughout gestation but not in mild and severe BOAS groups. Both BOAS and OSA are described as disorders with intermittent hypoxia, leading to repeated cycles of hypoxia and reoxygenation, which are the main cause of tissue oxidative stress, including vascular injury. 26 Altered uteroplacental blood flow in OSA was indeed related to oxidative stress, pro‐inflammatory cytokines, and endothelial dysfunction. 27 Positive correlation between maternal blood pH and pCO2 and uterine artery blood flow velocity in BOAS bitches indicates a direct effect of hypoxemia to increase blood flow to the pregnant uterus. However, among vascular sites of injury, uterine arteries are particularly susceptible during pregnancy. 28 Thus, we can assume that vascular oxidative and inflammatory lesions of BOAS prevent uterine artery from efficient remodeling to achieve a vasodilation status during pregnancy. Indeed, intermittent hypoxia promotes reduced endothelium‐dependent vasodilation and increased vasoconstriction of uterine artery in mouse model. 29 It is also worth pointing that severely affected BOAS bitches exhibited a pattern of uterine artery hypertension early in pregnancy, mainly compared to the mildly affected ones. These data lead to the assumption that ongoing vascular injury generated by intermittent and severe hypoxia is a preexisting alteration that ultimately causes adverse maternal and fetal outcomes during late pregnancy. At late pregnancy, severe BOAS bitches had increased umbilical artery vascular resistance (RI and S/D) compared to mild BOAS, despite no differences in umbilical artery blood flow velocity (ED). Considering that umbilical artery resistance is strongly associated with fetal well‐being and pregnancy outcomes, 30 our results may indicate that when BOAS symptoms are severe, vascular resistance in the umbilical artery increases, decreasing blood flow to the fetoplacental unit, eventually causing adverse effects on fetal development. There are several potential limitations to this study. It is constrained by a relatively small sample size and does not include a larger number of bitches from different breeds to account for the interbreed variability of the canine species. Therefore, our data should be interpreted with caution, as the bitches in this study presented breed‐related structural differences that need to be taken into account. Another potential impact on the present findings is the absence of a systematized and objective method for classifying the intensity of BOAS, such as the exercise tolerance test. For this reason, the condition of females based solely on the owners' perception, and clinical manifestations may be underestimated. In conclusion, severity of respiratory obstruction of BOAS has a parallel negative impact on maternal health and fetoplacental unit, despite systemic physiological adaptations aimed at maintaining gestational homeostasis. Albeit the data presented herein do not allow for definitive statements about specific mechanisms, some speculations can be made, but these definitely need confirmation in future studies. AUTHOR CONTRIBUTIONS Roberto Rodrigues da Rosa Filho: Data curation; formal analysis; validation; writing – original draft. Thayná Pantoja Gardés: Conceptualization; data curation; formal analysis; investigation. Maíra Morales Brito: Investigation; methodology. Leticia Lima de Almeida: Data curation; investigation. Camila Infantosi Vannucchi: Conceptualization; funding acquisition; project administration; resources; supervision; writing – review and editing. FUNDING INFORMATION This work was financially supported by FAPESP (São Paulo State Foundation) [grant number 2017/11844‐6]. CONFLICT OF INTEREST STATEMENT The authors have no relevant financial or nonfinancial interests to disclose. ETHICS STATEMENT This study was approved by the Animal Ethics Committee of the School of Veterinary Medicine and Animal Science at University of São Paulo (protocol number: 5264070317). Supporting information Table S1. Scorea of BOAS signs to evaluate the severity degree, adapted from the Poncet–Dupré score. 16