Relationship between maternal asthma, its severity and control and abortion

Are women with asthma, and more specifically those with severe or uncontrolled asthma, at higher risk of spontaneous and induced abortions? Pregnant women with asthma, notably when uncontrolled, are at higher risk of spontaneous abortion. Only one study has examined the association between asthma and spontaneous and induced abortions and revealed a modest increase in the risk of spontaneous abortions, particularly in women with more severe asthma and those with previous exacerbations, and a marginal decrease in the risk of induced abortions. A cohort of pregnancies from asthmatic (n 15 107) and non-asthmatic (n 34 331) women was reconstructed by linking three administrative databases from Quebec (Canada), between 1992 and 2002. The cohort included 7870 spontaneous abortions, 14 596 induced abortions and 26 972 live births. Pregnant women with and without asthma were analyzed. Asthma was defined by at least one asthma diagnosis and one dispensed prescription for an asthma medication in the 2 years prior to or during pregnancy. Asthma severity and control were assessed using validated indexes in the year before the 20th week of pregnancy or the termination of the pregnancy. Logistic polytomous regression models were used to estimate the relationship between asthma and asthma severity and control on the risk of abortion, while adjusting for potential confounders. The prevalence of spontaneous and induced abortions was 15.9 and 29.5, respectively. Maternal asthma was associated with an increased risk of a spontaneous abortion [odds ratio (OR) 1.41; 95 confidence interval (CI): 1.331.49] and a decreased risk of induced abortions (OR 0.92; 0.880.97). No association was observed between asthma severity and abortion, while uncontrolled asthma increased the risk of a spontaneous abortion by 26 (95 CI: 1441) and the risk of induced abortions by 11 (95 CI: 121). It is possible that the study results were confounded by imbalances between groups in variables that are not recorded in the databases, but that are known to be associated with spontaneous abortions, such as alcohol consumption, obesity or maternal smoking. However, we performed sensitivity analyses which revealed that these factors are unlikely to explain the observed increased risk for a spontaneous abortion. It is also possible that women with asthma are more likely to have abortions recorded in the databases, because subjects with a chronic disease tend to visit a physician more often than those without asthma. Therefore, our odds estimates for these outcomes may be overestimated when asthmatic women were compared with non-asthmatic women. A further limitation of the study is that it would have been more appropriate to measure the severity and control of asthma only during the pregnancy. Our cohort is less representative of women in the upper socio-economic level. This is not a threat to internal validity, but it could limit the external validity if the impact of asthma on the risk of abortion differed according to the socio-economic status of the mother. Despite the absence of supporting data, this possibility cannot be completely excluded. This study was funded by the Canadian Institutes of Health Research and Genentech. L.B. received research grants from Astra-Zeneca, Pfizer, sanofi-aventis, Novartis and Merck for research projects and co-chairs the Astra-Zeneca Endowment Pharmaceutical Chair in Respiratory Health. F.Z.K and A.F. have no competing interests to declare.