Inequalities in preterm birth among immigrants to Canada by race and time since immigration: A population‐based repeated cross‐sectional study

Preterm birth (birth <37 weeks' gestation) is the leading cause of infant mortality and morbidity.1 Differences in the risk of preterm birth between immigrants and the native-born population have been reported in many countries.2 However, most evidence is drawn from studies that treat immigrants as a singular group, potentially masking inequalities in perinatal outcomes within the immigrant population. Indeed, studies that disaggregated the immigrant population have found heterogeneities in birth outcomes according to maternal duration of residence, race, ethnicity, and resettlement pathway.2-4 Additionally, the health advantage ascribed to foreign-born individuals in receiving countries, also known as the healthy immigrant effect, has been demonstrated to dissipate over time and vary across immigrant subgroups and perinatal outcomes.5 Canada welcomes a significant number of immigrants annually and its immigrant population is projected to reach one-third of the total population in the next 20 years.6 Nonetheless, heterogeneity in the risk of preterm birth among the immigrant population remains poorly understood at a national level. We aimed to examine the risk of preterm birth in Canadian immigrants across multiple self-identified racial groups and years since immigration. We used the 2006 cycle of the Canadian Census Health and Environment Cohorts (CanCHEC), which contains sociodemographic characteristics from the Canadian long-form census linked to hospital discharge data from 2004 to 2016 in all provinces and territories except Quebec from the Discharge Abstract Database (DAD). We compared the risk of preterm birth between immigrants and Canadian-born women overall and by self-reported racial group (White, Black, East/Southeast Asian, Latinx, Middle Eastern, and South Asian). Preterm birth was defined as birth before 37 weeks, 0 days gestation. We considered all singleton live births throughout the study period and excluded those with implausible matching data or a missing outcome, as well as births to temporary residents, Indigenous women, and mothers who self-identified with multiple racial groups. The probability of preterm birth was modeled according to immigration status, for all women and by racial group, using survey-weighted logistic regression adjusted for maternal age, province/territory, and delivery year. For immigrant women we additionally modeled the probability of preterm birth as a function of time since immigration. Restricted cubic splines captured nonlinear relationships. Confidence intervals were estimated using bootstrap replicate weights, which accounted for multiple births to women. We report marginal risk differences (RD) for the probability of preterm birth between immigrants and non-immigrants and display the predicted probability of preterm birth as a function of years since immigration graphically. Immigrants accounted for 22% of the 2 355 270 (weighted) total pregnancies. Without accounting for time since immigration, the overall risk of preterm birth was similar between immigrants and non-immigrants (RD 0.002, 95% CI 0.000–0.005). However, within self-identified racial groups, immigrants generally had a lower risk of preterm birth compared to their Canadian-born counterparts (Table S1). This lower risk was driven by the large number of recent immigrants (<6 years in Canada) who had a lower probability of preterm birth compared to Canadian-born women of the same racial group, but the advantage diminished over time (Figure 1). By 10 years since arrival, the probability of preterm birth among all immigrants exceeded that of the Canadian-born population. When stratified according to racial group, most groups displayed comparable probabilities of preterm birth relative to the equivalent Canadian-born group after 15–20 years. However, for Black immigrants, the probability of preterm birth slightly surpassed that of the Canadian-born Black population around 16 years following immigration. A few limitations of this work should be noted. First, although we adjusted for factors including age, region, and delivery year to improve the comparability between immigrants and non-immigrants, we were unable to account for a comprehensive set of variables due to data limitations. CanCHEC does not collect information on overall health, and parity was not uniformly available across all provinces and territories throughout the study period. Additionally, socioeconomic characteristics were only available at the date of the census and might not reflect a person's socioeconomic profile at the time of delivery. However, a sensitivity analysis that additionally adjusted for maternal education level and income produced similar results. Finally, while a secondary analysis of the risk of spontaneous preterm birth (approximately two-thirds of the preterm births) demonstrated similar findings, we were unable to examine the risk of iatrogenic preterm birth due to small sample sizes when stratifying by race. Our findings align with previous work that demonstrated that duration of residence among immigrants is associated with an increasing risk of preterm birth.2 However, we additionally reveal racial differences within the observed phenomenon, further supporting the notion that the immigrant perinatal experience is not similar for all racial groups.4 While most groups exhibited a healthy immigrant effect that diminished over time, Black immigrant women were the only group whose risk of preterm birth eventually exceeded that of their Canadian-born counterparts. Although the etiology of preterm birth remains largely unknown, evidence from the United States suggests that Black women are at an increased risk due to social factors, such as structural racism and history of socioeconomic disadvantage, which contribute to an elevated allostatic load.7 Similar mechanisms might disproportionally affect Black immigrant women in Canada. Nevertheless, the precise causes of these inequalities warrant further investigation in perinatal health research. R.L. drafted the original manuscript and contributed to the statistical analysis. P.M.S. contributed to the statistical analysis and reviewed the manuscript. S.H. contributed to the statistical analysis and reviewed the manuscript. D.C.O. reviewed the manuscript. E.H. secured funding, was responsible for conceptualization and study design, contributed to the statistical analysis, and reviewed the manuscript. This work was supported by a Canadian Institutes for Health Research Postdoctoral Fellowship awarded to E.H. The authors have no conflicts of interest to declare. Ethical approval for this study was obtained from McGill University's Faculty of Medicine and Health Sciences Research Ethics Board. Research data are not shared. Table S1. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.