Oil and Gas Development and Risk of Preterm Birth: Evidence from San Joaquin Valley, California

Background. Prior studies report an association between adverse birth outcomes, including preterm birth, and proximity to oil and gas wells. Most attention has focused on unconventional natural gas developments, with little attention to conventional oil and gas extraction. Both conventional and unconventional oil and gas development have high levels of air pollution emissions.
Objective. Our objective was to determine whether maternal residence proximity to conventional oil and gas wells in active development increases the risk of preterm birth. Methods. We obtained health data from birth certificates linked to hospital discharge data from the Office of Statewide Health Planning and Development. Data on oil and gas wells came from the California Department of Conservation. After exclusion, the births dataset had 27,913 preterm cases and 197,461 term controls, with a total of 225,374 births, and the wells dataset included 11,972 oil and gas wells in active development. For each birth, we computed an exposure index by taking the sum of the inverse distance to all wells in active development within a 15 km radius of the maternal residence. We defined four preterm birth categories based on gestational length (in weeks): 20-23, 24-27, 28-31, and 32-36. We estimate crude and adjusted odds ratios for preterm birth using logistic regression. To test the robustness of our results, we used different radii and formulas to generate the exposure index and determined whether this changed the findings. To assess the effect of distance to wells, we generated an indicator for maternal residences with at least one well within a radius of 1 km, 3 km, 5 5 km, 7 km, or 10 km.
Results. Exposure to oil and gas wells in active development was associated with a small but significant increase in the risk of preterm birth. The risk was highest for the 32 to 36-week preterm birth category, with exposure throughout gestation associated with increased risk for this group. Several sensitivity analyses testing the assumptions of our exposure assessment method found similar results. In addition, the significant effects persisted for births that had at least one well within 1 km, 3 km, or 10 km for early preterm birth, and within 5 km or 7 km for late preterm birth.
Conclusions. Exposure to oil and gas wells in active development was associated with higher risk of preterm birth. More work is needed to infer whether a causal link exists between the exposure and adverse birth outcomes.