CRIS Scientific Seminar 2023-2024

Friday, June 7th 2024, 11:30 am
Sciences Po, Room K008 (1, St-Thomas-d'Aquin)

Estimating the Impact of Prenatal Opioid Exposure on Infant Health:
Evidence from Multiple Identification Strategies

Lawrence (Lonnie) Berger

Vilas Distinguished Achievement Professor
Associate Vice Chancellor for Research in the Social Sciences
Sandra Rosenbaum School of Social Work, University of Wisconsin-Madison

The adverse effects of the opioid epidemic in the United States are well documented. Moreover, high rates of prenatal opioid exposure, particularly among disadvantaged populations, has generated concern that a large number of children may be at risk of poor prenatal and, potentially, lifelong health and development. Whereas research has established that maternal opioid use during pregnancy is associated with a host of negative child outcomes, prenatal exposure is also strongly correlated with other socioeconomic and psychosocial risks to child development. As such, existing studies have yet to identify plausibly causal effects. This study uses longitudinal administrative data from Wisconsin and six identification strategies—standard (e.g., OLS) regressions with extensive controls; propensity score matched regressions; inverse probability (of prenatal exposure) weighted regressions; sibling fixed-effects regressions (comparing exposed and nonexposed siblings); cousin fixed-effects regressions (comparing exposed children to unexposed children of their mothers’ sisters), and instrumental variables regressions—to produce bounded estimates of the plausibly causal effects of prenatal opioid exposure on multiple domains of child health during the first year of life. Comparing estimates produced using each strategy provides new insights into whether associations of prenatal opioid exposure with specific aspects of infant health are likely causal and, if so, at what orders of magnitude. 

To find out more about Lawrence (Lonnie) Berger (cv)

Open Seminar. Please register here to join us!