The immigration paradox in cerebral palsy: More and better data are needed

Abstract

Sigholt et al.¹ observed lower prevalence of cerebral palsy (CP) among children born to immigrant mothers from high-income (HIC) (1.44 per 1000 livebirths) and low- and middle-income countries (LMICs) (1.71 per 1000) than among children born to non-immigrants mothers (2.11 per 1000). The study relied on a record linkage between two population-based, high-quality Norwegian registries, containing rich standardized clinical information and not affected by selection or recall bias. Although the finding that immigration is associated with lower CP prevalence is not new, these new data reignite the question about what makes immigrants less prone to CP. Delving into this question could help identify immigrant characteristics that provide clues about risk factors for CP.

The authors offer two potential explanations for the observed differences. First, they found that perinatal mortality was higher among immigrants from LMICs (7.5 per 1000 births) than among non-immigrants (5.4 per 1000). They suggest that if some of the children of immigrants who died in the perinatal period had survived, these children would have had a higher prevalence of CP, potentially explaining at least part of the difference. But the authors did not provide sensitivity analyses or further elaboration. If the 291 excess perinatal deaths in the LMIC group (2.17 per 1000) had survived, their CP prevalence would need to be 187 per 1000 to account for 54 additional CP cases needed in the LMIC group to close the gap with non-immigrants, which seems unlikely. The argument is also at odds with the findings related to HICs, which exhibited both the lowest prevalence of CP and of perinatal death among all three groups. Information on the causes or characteristics of perinatal deaths would have been a welcome addition.

Second, in line with previous studies, authors consider that the ‘healthy immigrant effect’ may help explain the lower CP prevalence among immigrant mothers, who exhibited lower rates of pre-pregnancy diagnoses than their Norwegian counterparts. On the other hand, immigrants from LMICs had more consanguinity, lower use of folate during pregnancy, presumably faced barriers in accessing and navigating the Norwegian health care system, and may have faced socioeconomic disadvantage. The presence of good perinatal outcomes despite such constellation of risk factors has been referred to as the ‘immigrant paradox’ in some populations,² which highlights the positive health selection of immigrants and their resilience.

A main limitation of the study is the lack of detailed sociodemographic and behavioral data to allow for subgroup analyses that may offer clues on protective factors. Grouping mothers according to their source country economies is not very informative, as it mixes mothers from different world regions and cultures and does not account for the heterogeneity within countries. Yet this rough classification detected a sizeable difference, which was not explained by the clinical variables. Larger differences would be expected to be detected in subgroups defined by meaningful pre- and post-migration characteristics, behavioral and socioeconomic factors, such as high maternal education,^{3, 4} which in turn may inform the design of studies testing specific causal pathways.

To sum up, the observed differences may reflect differences in risk and protective factors between the groups, but the observed variables do not explain them. Future studies are needed that expand the powerful population data linkage approach used in this study, with detailed data on maternal and paternal pre- and post-migration characteristics. These sociodemographic and behavioral factors are likely to account for the heterogeneity of the immigrant population and its differences with non-immigrants, which may reveal what is behind the association between immigration and CP.

No funding.

The authors have no conflict of interest.