Our hypotheses:

  • Environmental stressors, including urban heat and poor air quality will be associated with biomass of canopy and adverse health – specifically poor birth conditions.
  • Women exposed to more tree canopy will have healthier babies at birth


First, we will use descriptive analyses, which include crude bivariate analyses and multivariate analysesAnalyses will consider nonlinear functions, from which appropriate cut-points will be determined. Second, we will employ multi-level, multivariate regression modeling to evaluate the extent to which the HESI is associated to specific health outcomes and canopy structures.

City Comparison Icons (1).png

Our approach is innovative because a major focus of our project is to describempirical levels of exposure to environmental stressors and the role of tree canopy on human health at the sub-neighborhood scale. Previous research suggests that people below the poverty level bear a higher burden of impacts from air pollution1 (Wong et al., 2008), so we have chosen urban areas with similar percent of people at or below the nationally defined poverty level (< 24% and > 15%), similar amounts of impervious surface (< 45% and > 20%), yet vastly ranging in the amount of tree canopy (defined as vegetation above 5 meters high – LiDaR derived).


We will conduct newborn health, heat, and air quality research in 5 cities: Albuquerque, NM; Boise, ID; Portland, OR; Sacramento, CA; and Tacoma, WA.


In these 5 cities, we are seeking volunteers to host stationary air quality monitors at their homes, and we will host a 2-day mobile heat and air quality monitoring event during the peak heat period in each city over two years – Albuquerque and Sacramento in Summer of 2018, and Portland, Tacoma and Boise in Summer of 2019.

Previous Research
The bulk of the research previously conducted exploring environmental exposures and adverse pregnancy outcomes has been observational, which is the most widely used research approach in environmental epidemiology where exposures cannot be assigned and individual modeling is prohibitively expensive, especially on a population-level. Most of the study designs are geographically-based retrospective cohort studies, which is the design being employed in the proposed work. When individual-level outcome data are available, multilevel analyses have been used to assess the effect of the environmental exposure, even after adjusting for individual and area-level confounders, which allows for the estimation of the environmental effect. This is the approach we propose to undertake in this project.

Icons made by dDara, Freepik and Smashicons from licensed by CC 3.0 BY