Miscarriage is the most common pregnancy complication in the United States, occurring in 15-20% of all clinically recognized pregnancies. It has devastating impacts on both parents, with elevated anxiety and depressive symptoms in the mother, and even post-traumatic stress disorder, lasting up to a year post-event. Identification of modifiable risk factors of pregnancy loss, thus, is of great public health importance. Ambient air pollution is an increasingly recognizable risk factor for pregnancy complications. Although some evidence exists linking air pollution to pregnancy loss, the need to identify pregnancy loss events through medical records is a large hurdle in comprehensively assessing this association. For the proposed work, we will employ a novel approach that uses the outcome metric of live birth-identified conceptions in a variation of the time series design to overcome this major limitation of all existing studies. To this end, we will use information on all live births recorded in Denmark between 1997 and 2017, and couple these with highly spatio-temporally resolved air pollution predictions from well-validated models. Specifically, we will (1) investigate the association between both particulate matter air pollution exposures, an indicator of the mixture of air pollution, and traffic- related air pollution, a component of ambient air pollution found to be especially toxic for pregnancy complications, and pregnancy loss; (2) identify the critical windows of vulnerability; (3) develop novel methodology to assess potential effect modification by multiple modifiers simultaneously while incorporating the highly resolved time-varying exposures during gestation; and (4) identify maternal- and neighborhood-level effect modifiers, and therefore, the most susceptible subpopulations to air pollution impacts. Our findings can greatly inform targeted regulations and policies, crafting of public health messages, and targeted interventions.
Miscarriage is the most common pregnancy complication in the US occurring in 15-20% of clinically recognized pregnancies; current analytic approaches, however, depend on early pregnancy losses being identified in the medical system. Ambient air pollution is an increasingly recognized risk factor for pregnancy complications, but quantification of the impact on pregnancy loss is limited by these events being identified through medical records. We propose to use a novel approach, using the outcome metric of live birth-identified conceptions and a variation of the time-series design, to (1) investigate the association between air pollution and pregnancy loss, (2) identify critical windows of vulnerability, and (3) identify important maternal- and neighborhood-level modifiers of this association to inform targeted policies and interventions.