Variability in birthweight, birthweight charts, and adverse outcomes: Is the “right size” the right question?

Abstract

In this issue of Paediatric and Perinatal Epidemiology, Choi and colleagues 1 present novel nationwide Australian data on adverse perinatal outcomes of immigrants. One objective was to evaluate whether infant mortality, morbidity, and size at birth varied by maternal country of birth. There is a substantial body of literature indicating that newborn outcomes do vary by maternal birthplace, as documented by the systematic reviews and meta-analyses cited by the authors. Therefore, their findings of maternal birthplace variations in newborn outcomes come as no surprise. What is less known, however, is that newborn outcomes of well-defined immigrant groups also vary according to country of destination. 2 This source of variability may contain valuable clues about environmental influences on perinatal outcomes of immigrants, including healthy adaptations to new environments and factors contributing to adverse outcomes. Methodological issues tied to variations in local population profiles, outcomes, practices, and policies in the countries of destination have contributed to inconsistent findings reported in the literature, however. The authors illustrate the inconsistency of previous findings with the example of Latin American/ Caribbean immigrants, who, compared with the native-born population, were at higher risk of low birthweight in Europe but at lower risk in the United States. Yet this result does not mean that this group fares worse in Europe than in the United States. The apparent inconsistency stems from the fact that the baseline rates of low birthweight among the (reference) native-born groups were higher in the United States than in Europe, thus affecting the disparities among Latin American/Caribbean immigrants, whose low birthweight rates were in between those of the two native-born groups. Shifting reference groups aside, other factors known to contribute to heterogeneity of results between migrant studies include different study populations and data sources, different definitions of migrant groups, different composition of source countries of immigration when immigrants are pooled into broad world regions (for instance, women from the Caribbean being more common in the UK and women from Latin America being more common in the United States), selective admission policies, and different analytic approaches. 3 Most perinatal research focusing on immigrants adopts a "domestic perspective" that compares different immigrant groups to the native-born population. While this ethnocentric approach helps to quantify health disparities at the local level, it undermines comparability of study results regarding specific immigrant groups in different settings. In this respect, one practical advantage of the prescriptive Intergrowth-21st (IG-21) standards 4,5 over population-based birthweight references is that using a single set of standards from intrauterine life to childhood and across geography and time allows comparability of results, among both specific immigrant groups and non-immigrant populations, providing com...

The 3G Multigenerational Cohort of Nova Scotian women and their mothers and offspring

“…Additionally, many of these cohorts are not necessarily representative of the general population as only women who agree to participate are included. Population‐based inter‐ and transgenerational cohorts have been established within administrative and perinatal databases in several places, including the Scandinavian countries and Canada, but some of these studies are limited with respect to the amount of antenatal and maternal behavioural data that is available. In comparison to experimental studies conducted in animals, associations observed in epidemiologic cohort studies may be attributable, separately or in combination, to epigenetic modifications influenced by in utero exposures, genetics, and environments or behaviours shared among family members.…”