Socioeconomic disparities in low birth weight outcomes according to maternal birthplace in Québec, Canada

Abstract

Objective: Studies in the USA suggest that the association between maternal birthplace, socioeconomic status (SES), and low birth weight (LBW) can vary across different immigrant groups. Less is known outside the USA about these associations. Our study assesses the association of maternal birthplace and SES on the likelihood of LBW infants in Québec, Canada.

Methods: Using 2000 Québec birth registry data, logistic regression was used to examine differentials in LBW according to maternal birthplace and SES. Singleton infants born to Québec mothers (n=47,988) were grouped into nine regions based on maternal birthplace: (1) Canada; (2) the USA and western Europe; (3) eastern Europe; (4) Latin America; (5) the Caribbean; (6) Sub-Saharan Africa; (7) north Africa and Middle East; (8) South Asia; and (9) East Asia and Pacific. SES was classified into four categories according to maternal educational attainment: (1) low SES (<11 years); (2) medium–low SES (11–12 years); (3) medium–high SES (13–14 years); and (4) high SES (more than 14 years). Covariates included maternal age, gestational duration, and parity. LBW was defined as between 500 and 2499 g.

Results: Compared to a LBW prevalence of 4.5 for Canadian-born mothers, South Asian- and Caribbean-born mothers had prevalence percentages of 9.2 and 8.2, respectively. After adjusting for SES and other covariates, the likelihood (odds ratio (OR), 95% confidence intervals (CI)) of LBW outcomes remained greater for South Asian- (OR 2.84; 95% CI, 1.90–4.24) and Caribbean-born mothers (OR 1.52; 95% CI 1.11–2.10). After pooling these two groups and testing for moderation by SES, we found that high SES immigrant mothers (OR 3.82; 95% CI 2.33–6.25) had a higher likelihood of LBW infants than low SES mothers (OR 2.00; 95% CI 1.22–3.33) compared to high SES Canadian-born mothers.

Discussion: In Québec, the association between foreign-born status and LBW varies according to maternal birthplace.

Keywords:

• immigration
• low birth weight
• socioeconomic status
• Canada

Notes

1. Data on acculturation and ethnicity are not specifically collected in the Québec birth registry. We used the data available to create and test two proxy measures of acculturation and ethnicity. First, we constructed a dichotomous indicator of acculturation based on any reported household language other than French or English. This proxy indicator was not significant in any models. Neither did the indicator affect the significance level of the other study variables. Second, we constructed a proxy indicator of ethnicity based on maternal language. The indicator had a high degree of collinearity with maternal birthplace groupings, thus compromising our analyses. Moreover, we were critical of language as a marker of ethnicity given the diverse cultural traditions and histories contained within broad language families such as English.

2. To confirm our findings using maternal education as representing maternal SES, we obtained 2001 Canada census data on the average income of the mother's place of residence at the three-digit postal code level. The average income of the mother's area of residence correlated highly with maternal education. We created a factor score of maternal SES based on individual educational attainment and average income of the maternal residential area. The results using factor scores as SES indicators were statistically similar to and conclusions are the same as those obtained when using maternal education alone. We chose not to employ this measure of maternal SES since we viewed such a measure as conceptually and statistically mixing individual- and area-level indicators of SES with no additional value to the analysis.