Interpretation
Prevalence was high in our study. The prevalence of OSA in a general population of pregnant women, regardless of BMI, is not known. It has been estimated, depending on the study, at 1.4–16.9% among women of child-bearing age.2,14,15 In studies of populations of pregnant women at risk of OSA, its prevalence was higher, as in our study. Rice et al. sought to compare the risk of OSA as a function of BMI in a population of pregnant women.23 They found that the odds ratio for risk of OSA was 3.69 (95% CI; 1.82–7.50) for overweight women (BMI 25–29.9 kg/m2) and 13.23 (95% CI, 6.25–28.01) for obese women (BMI ≥ 30 kg/m2), compared with normal-weight women (BMI <25 kg/m2). Louis et al. studied 175 pregnant women with BMI ≥30 kg/m2 and found a 15.4% prevalence of OSA (13 mild, 9 moderate, and 5 severe).24 Their prevalence was lower than ours, perhaps explained by their BMI inclusion criterion (BMI >30 kg/m2), also lower than ours. In addition, although the polysomnographic criteria for an OSA diagnosis were similar to those we used, the term of pregnancy at the time of their testing and diagnosis was not specified; this difference too could explain the difference in prevalence between their results and ours; prevalence of OSA increases with the term of pregnancy. Work by Facco et al. in a population of pregnant women at risk of OSA (BMI ≥30 or chronic hypertension) showed that in the first trimester of pregnancy, the prevalence rates of mild, moderate, and severe OSA were respectively 21%, 6%, and 3%, while in the third trimester, they were 35%, 7%, and 5% (P <.001).25 These two studies are consistent with our results in finding a higher prevalence of OSA in a population at risk, especially due to obesity, for mild and moderate cases.
We did not find significantly different pregnancy, delivery, and neonatal complications between the two groups except that gestational diabetes was more frequent in women with this sleep disorder. Our results are not in complete accord with the literature. In 2012, Chen et al. conducted their randomized study of 791 pregnant women with and 3955 without OSA, and found significantly higher risks of preeclampsia, FGR, cesarean delivery, preterm delivery, and low birth weight in the OSA group.26 The principal limitation of their study was the absence of adjustment for BMI in their Taiwanese population at risk of obesity. In 2014, a meta-analysis examining the consequences of OSA on pregnancy that did adjust for BMI found a risk of developing pregnancy-related hypertension and/or preeclampsia that was 2.34 (95% CI, 1.60–3.09) times higher in women with compared to without OSA, and a risk of gestational diabetes 1.86 (95% CI, 1.30–2.42) times higher.27 This study, like ours, found no increase in the risk of low birth weight. Another more recent meta-analysis in 2018 focused specifically on the consequences of OSA on delivery and on neonatal condition.18 After adjustment for age and BMI, this study found increased risks of cesarean and preterm delivery (<37 weeks) as well as of FGR (<2500 g). FGR may bias these results, as it is known to increase the risks of induced preterm birth and of cesarean delivery. These different studies did not target populations of pregnant women at risk of OSA because of high BMI. In their population of obese pregnant women, Louis et al. showed that the women in the OSA group had a higher risk of preeclampsia, cesarean delivery, and newborn transfer to the NICU.24 After adjustment for various criteria, and especially for BMI, they found that only the risk of preeclampsia was significantly higher. Our study observed that the rate of the composite criterion (of pregnancy-related vascular diseases) was higher in the OSA group, but not significantly so (34.5% vs 16.2%, P =.086). Contrary to our study, theirs found no difference between the groups for gestational diabetes.
Our results about the risk factors associated with OSA are consistent with those of the literature. Chronic hypertension, diabetes, high BMI, and age were all shown to be comorbidities associated with OSA in a general population of obese women.28,29 In a population of pregnant women with obesity, Louis et al. also showed that those with OSA had a higher BMI (46.8 ±12.2 vs 38.1± 7.5 kg/m2, P =.002) and more frequent chronic hypertension (55.6 vs 32.4%, P =.02).24 The 2018 meta-analysis also reported that the women with OSA were older (RR 1.66, 95% CI, 1.04–2.228) and had a higher BMI (RR 3.31, 95% CI 2.30–4.32).18