Adverse perinatal outcomes in the Australian Indigenous population, the role of asthma

Abstract

Asthma in pregnancy has been shown in a number of studies to be associated with adverse perinatal outcomes including low birth weight and perinatal death.1– 4 However, the majority of studies in this area have been in general populations in the US, Europe and Australia. It is therefore of interest that Clifton et al.5 recently published a study on perinatal outcomes in an Australian Indigenous population, especially as adverse perinatal outcomes in this population remain high, (although declining), with mortality and low birth weight rates double that of nonIndigenous rates.6 However, I do have some concerns about the results presented and their interpretation, namely around the choices of control group and confounders. In the abstract and results section the authors have stated they observed a twofold increase in neonatal deaths in Indigenous babies whose mothers had asthma compared to nonIndigenous women who did not have asthma. However, in Table 2 where odds ratios of Indigenous women with asthma have been presented the reference group is referred to as being ‘nonIndigenous women without asthma’. Given the existing risk of adverse perinatal outcomes in the Australian Indigenous population it is therefore not possible to tell what role asthma plays in these findings or whether the reported associations are in fact driven by Indigenous status (or rather, the risk factors and behaviours as a results of colonisation and racism that continue to drive Indigenous health outcomes). Further, the findings in Table 2 that asthma does not show any added effect on perinatal outcomes in nonIndigenous women supports the argument, that at least in this dataset, asthma may not be making a difference on perinatal outcomes. The other possible source of confounding is socioeconomic status. The authors have stated that the inclusion of confounders was determined by the univariate associations between covariates and outcomes. Table 1 shows that socioeconomic status measured as attained education level was significantly different between Indigenous and nonIndigenous mothers, P < 0.001. Given these findings and that it is well recognised that socioeconomic determinants are a strong driver of health,7 it is unclear then why the multivariate associations were not adjusted for education level. Although no other similar studies exist for Indigenous groups, studies comparing Black, Hispanic and White American women found that although Black women had a higher prevalence of asthma and were more likely to have lower socioeconomic determinants, asthma was not able to explain the increase in adverse perinatal outcomes in Black populations.8 These authors concluded that asthma does not contribute to racial disparities in obstetric and neonatal complications. It is therefore my suspicion that the results from Clifton et al. are driven in large part by Indigenous status and socioeconomic differences rather than by asthma. It may be that asthma is an effect modifier of the impact of Indigenous status on perinatal outcomes but the current analysis does not provide enough information for this assessment to be made, which brings into question the emphasis by the authors on a need for improved models of care due to asthma.