The Impact of Maternal Smoking During Pregnancy on Risks of Obesity among Preschoolers

Child health outcomes of maternal smoking during pregnancy have been well-documented across multiple academic disciplines. Among the various short-term and long-term child health effects explored in the relevant literature, of particular interest to health researchers is the finding that children of mothers who smoke during pregnancy are at an elevated risk of obesity. More specifically, research focusing on children’s birth outcomes indicates that children whose mothers smoke during pregnancy are born with lower birth weight than non-smokers' children. However, relying largely on descriptive analyses, studies on later childhood outcomes find that the risk of obesity during pre-school years is higher among the former group. As such, findings from the two strands of the current literature suggest that despite having lower birth weight, children of smoker mothers catch-up with non-smokers’ children in terms of body-weight during later years, thereby demonstrating a higher susceptibility to risks of obesity. The current analysis is the first to empirically verify the conjectured relationship (‘catch-up’ hypothesis) between mothers’ smoking behavior during pregnancy and children’s weight using a longitudinal sample of children (considered from birth through preschool ages).

We construct our analysis samples by matching mothers’ information from the original cohort of National Longitudinal Survey of Youth with their biological children from the biennial Children and Young Adults surveys. We capture children's body-weight outcomes using reported measures of weight (pounds) and body mass index (BMI). To control for mother-specific unobserved heterogeneities that are correlated with mothers’ smoking behavior and may also affect child health, we control for mother fixed effects in our regressions that employ ‘within-mother and across childbirth’ variations in smoking behavior during pregnancy. The effects of maternal smoking are estimated by the differences in weight outcomes of siblings whose mothers’ smoking behavior during pregnancy varied across childbirth. We control for children’s demographic characteristics (race, ethnicity, birth order, sex, height), measures of mothers’ health, marital status, education, employment status, and prenatal behavior (substance use and doctor visits), and family-specific socio-economic conditions (family size and poverty status). Further, for greater comparability of regression results, we classify our analysis by child age and sex.

Consistent with findings in the aforementioned literature, our estimates suggest that maternal smoking during pregnancy has statistically significant negative effect on children's birth weight (estimated difference between smokers’ and non-smokers’ children is 0.27 pounds) and birth BMI (estimated difference is 0.57). Focusing on later years of preschool period, there are large heterogeneities in regression estimates across analysis samples. However, overall we find that children of smoker mothers do catch up with children whose mothers had not smoked during pregnancy. In particular, we find that mother's smoking during pregnancy has a negative effect on weight and BMI during early preschool ages (0-2 years), but the negative effects disappear in later years of preschool age (3-5 years). The regression estimates are supported by multiple sensitivity analyses. Our study provides comprehensive empirical support in favor of policy interventions that should aim at effectively minimizing the avoidable child health risks triggered by maternal smoking during pregnancy.