Maternal Residential Mobility Between Births: A California Statewide Study.

Abstract

Background: Studies of perinatal and paediatric health often analyse data from consecutive pregnancies. However, little is known about the factors associated with maternal residential changes between births or how maternal mobility may affect the validity of epidemiological findings, particularly those involving geographic-based measurements.

Objectives: Using a statewide sibling cohort, we studied the pattern of residential mobility between births and its association with several major sociodemographic and perinatal factors. We discussed how maternal moving between births may affect the validity of epidemiologic research, focusing on sibling comparison analyses.

Methods: We geocoded maternal residential addresses extracted from the livebirth records of sibling pairs from 2007 to 2015 in California. We described moving patterns between births, and we estimated the risk ratio (RR) of maternal moving according to maternal sociodemographic, pregnancy, and neonatal factors of the preceding birth. We studied maternal social mobility by examining the changes in the neighbourhood Social Vulnerability Indexes (SVI) among the movers.

Results: Our sample included 995,342 sibling pairs. Overall, 49.8% of mothers moved between births; 34.9% to another ZIP code, and 9.3% to another county. Younger maternal age, smoking during pregnancy, self-reported non-Hispanic Black race, lower educational attainment, nulliparous status, no prenatal care or having used government insurance for prenatal care, and non-adverse birth outcomes at the precedent birth were associated with maternal moving between births. Maternal social mobility between births in the cohort also varied by age, race/ethnicity, and education levels.

Conclusions: Maternal residential mobility between births was more common within certain sociodemographic subgroups in California. We recommend measuring residential changes in sibling studies, and assessing the potential impacts on statistical power, time-varying confounding, carry-over effect, and generalisability of findings. Further research is warranted to better understand how maternal mobility influences various types of perinatal epidemiological studies.