Environment/Climate Change
Oil and Gas Development and Adverse Birth Outcomes: A Conceptual Framework, Literature Review, and Bias Analysis Marijke Rowse* Marijke Rowse Cassandra Clark Julia Bond Kaylin Vrkljan Erin Polka Erin Campbell Amelia Wesselink Amira Aker Mary Willis
Background: An estimated 17.6 million Americans reside within 1.6 km of active oil and gas development. Many studies have examined associations between oil and gas development (OGD) and adverse reproductive health outcomes. However, the strengths and weaknesses of this literature need further examination to improve future research. Building upon a recent systematic review (Aker et al. 2024), we identified the biases and gaps in the existing research, analyzed the potential influence of unmeasured confounding by income, and suggested potential research avenues for future studies.
Methods: Building on the previous review by Aker et al., we identified 18 epidemiologic studies of oil and gas development exposure and adverse birth outcomes. We extracted measures of association (e.g., risk ratios, odds ratios, hazard ratios) for the relation between OGD and specific birth outcomes from each study. For each study, we evaluated for pre-identified sources of bias that are common in reproductive health research (e.g., live birth bias). We will use quantitative bias analysis (QBA) to evaluate the potential impact of unmeasured confounding by socioeconomic status on the published associations.
Results: Among the 18 included studies, 83.3% relied on vital statistic records as their primary data source which can introduce misclassification and selection bias. The most common endpoints were birth weight (77.8% of studies), preterm birth (72.2%), and small-for-gestational-age (55.6%). QBA will help disentangle to what extent unmeasured confounding from income may be a key source of bias in this association.
Conclusion: Our detailed review of existing literature on the association between oil and gas development and birth outcomes highlights important sources of bias that should be considered when interpreting the results, including points of expansion for future research.