Grantee Research Project Results
2023 Progress Report: Understanding Pediatric Susceptibility Across Temperature and Environment in New York (UPStATE NY)
EPA Grant Number: R840472Title: Understanding Pediatric Susceptibility Across Temperature and Environment in New York (UPStATE NY)
Investigators: Clougherty, Jane E. , Sheffield, Perry E , Schinasi, Leah H , Kinnee, Ellen J , Just, Allan C
Institution: Drexel University , University of Pittsburgh , Mount Sinai School of Medicine
EPA Project Officer: Hahn, Intaek
Project Period: September 1, 2022 through August 31, 2025
Project Period Covered by this Report: September 1, 2022 through August 31,2023
Project Amount: $1,350,000
RFA: Cumulative Health Impacts at the Intersection of Climate Change, Environmental Justice, and Vulnerable Populations/Lifestages: Community-Based Research for Solutions (2021) RFA Text | Recipients Lists
Research Category: Human Health , Environmental Justice
Objective:
We aim to identify community stressors and resilience-enhancing assets that shape relationships between climate change-related exposures [heat, ozone (O3), fine particulate matter (PM2.5)] and child health. We are particularly interested in assets which may bolster children’s resilience to climate-related exposures in environmental justice (EJ) communities across Upstate NY. We hypothesize that: (1) Many communities with high social stressors (e.g., poverty) also have healthpromoting assets (e.g., greenspace). (2) Health-relevant stressors and assets differ in urban vs. non-urban settings; (3) Some community assets buffer stressor-related susceptibilities to pollution and heat.
Experimental approach: We will work with Upstate EJ and climate partners to identify key assets – partners include Pediatric Environmental Health Specialty Units (PEHSU), the statewide Children’s Environmental Health Centers (NYSCHECK) network, and community and government partners. We leverage complete patient-level data on all child (age 0-17) emergency department visits and hospitalizations in NYS over 2005-19 from the Statewide Planning and Research Cooperative System (SPARCS), examine stressors identified by the NYS DEC Climate Justice Working Group, and apply the Resilience Activation Framework, organizing assets in five domains (human, economic, social, political, environmental capital). We have generated daily fine-scale exposure estimates for heat, O3, and PM2.5, and examine variation in exposure-health relationships using case-crossover methods, inherently adjusting for confounding by focusing on within-patient changes over time. Finally, we will apply climate projections to estimate impacts under climate scenarios, and, with partners, integrate findings into local and state climate adaptation and pediatric clinical initiatives.
Progress Summary:
Aim 1: Elicit input from community and diverse stakeholders to characterize and prioritize stressors and assets relevant to climate exposures and child health.
Aim 1.1: With Monarch of Infinite Possibilities, identify and engage with EJ and climate leaders from our 4 focal cities (Buffalo, Rochester, Syracuse, Albany), rural health and state partners, and clinicians serving vulnerable children, to prioritize stressors and assets, from among vulnerability indicators identified by NYS DEC Climate Justice Working Group, and assets in the Resilience Activation Framework, emphasizing policy/ programmatic/ clinical relevance.
The research team has met with community-engaged partners in three of the four focal cities (Buffalo, Rochester, and Syracuse) and also with partners in Binghamton. The partners included leadership of the New York State Children’s Environmental Health Centers in Buffalo, Rochester, and Syracuse and the Rural Health Network of South Central New York in Binghamton. The team also met with the Director of Community Engagement of the NIEHS Center for Environmental Health Research in Rochester. In these introductory meetings, the team provided an overview of the UPSTATE NY project and solicited input regarding the proposed assets using the Resilience Activation Framework. The initial strategy of encouraging open brainstorming of additional assets was modified to encourage critique of the proposed assets and a prioritization. This modified approach has helped focus the conversation and the efforts of the team in identifying datasets with statewide, spatially-resolved data that can be tested in our models.
We have identified a short list of community assets emphasized by our research partners. These assets, and possible indicators, include: School quality [crowding, buildings quality, educational quality (eg, student-teacher ratio), school spending as proportion of wealth or tax base]; Childcare access; Headstart programs; Food desert/ grocery stores; Greenspace/ recreational space; wealth (Median household income); educational attainment (percent with bachelors degree); Transportation (transit stops); Broadband access; Air conditioning data (energy use). We have compared these identified assets with the frameworks we originally proposed, and found reasonable correspondence, with each of the original five Abramson domains represented, as well as the relevant domains of the Child Opportunity Index 2.0.
Aim 1.2: Develop Geographic Information Systems (GIS)-based indicators of prioritized stressors, and assets. Examine spatial correlations among assets, stressors, and exposures, by race/ethnicity and poverty. Review with partners for interpretation, refinement.
We are now building GIS indicators for those assets emphasized, to date, by project partners and communities. We are putting some initial emphasis on school quality indicators, as a clear indicator of investment in children: data for these indicators is being drawn from NYS 5-year Building Condition Surveys (BCS) for 2005, 2010, 2015, and indicators from NYS School Report cards. Other asset indicators under development include tract-level grocery stores; data is drawn from National Neighborhood Data Archive, and indicators being developed annually for 2003-2017, to match cohort data, and assess changes in the food environment over time. Locations of public parks (ParkServe data) are being identified, and greenspace metrics developed using National Land Cover Database (NLCD) data for all years 2001-2021; land cover class is being finalized before calculating percent greenspace by tract.
The next steps in engagement and input is two-part: 1) A second round of conversations with the initial partners will build on the first conversation but focused on specific input about the identified asset data (i.e. we will “ground truth” how representative the community asset data are for the specific communities about which they have lived or professional experience). 2) We will invite youth to participate in a 3-part virtual dialogue, over the coming year, to orient them to the project, discuss assets and indicators, and discuss model findings.
Common spatial patterning in identified asset indicators, previously-identified stressors, and environmental hazards (heat, air pollution) will be examined using GIS. Bivariate spatial correlations between variables of interest will be computed using Simultaneous Autoregressive Models (SAR) to account for spatial dependence. An overall asset index will be generated using spatially-stratified principle component analysis (PCA), stratified by urbanicity (high density urban, low density urban, suburban/small town, rural).
Aim 2: Test whether community stressors and assets modify climate exposure-health associations. Use findings to project future climate-health burden and inform climate adaptation initiatives, including integration of climate into public health, environmental initiatives, and clinical care.
Aim 2.1: Examine associations of fine-scale daily climate exposures and ED/hospital visits, for children aged 0-17 for 2005-19 in Upstate NY, using case-crossover methods.
We obtained and geocoded all pediatric ED visits from 2005-2019 from New York Statewide Planning and
Research Cooperative Systems (SPARCS). Daily residential block-group level PM2.5 and temperature were estimated using fine-scale spatiotemporal models. We used a time-stratified case crossover design, with control days matched to case days by month and day of week, to estimate associations between daily block group PM2.5 and risk of all-cause and cause-specific ED visit using distributed lag nonlinear models (DLNM), adjusting for daily minimum temperature (Tmin). Associations were estimated across 6 lag days (lag0-lag5), and outliers [the top 1% of lag0 PM2.5 concentrations (> 28.6 µg/m3) were excluded].
Our dataset includes 22,660,029 all-cause ED visits among children aged 0-17, with most (65%) residing in high-density urban areas. Nonlinear exposure-response was observed for all-cause and cause-specific ED visits, with steeper increases in risk per unit increase in PM2.5 at low concentrations (< 6 ug/m3). We found particularly strong associations between PM2.5 and injury outcomes, even after adjusting for temperature . We found positive but weaker associations for other diagnoses (respiratory, nervous system, digestive diseases, infections, mental disorders, general symptoms). Effects from most outcomes were very acute, with significant positive associations for case day exposures, and weak or null associations for earlier lags.
Aim 2.2: Test variation in strength of exposure-health associations by stressors and assets, with emphasis on EJ areas, with on-going input from community, clinical, and government partners for interpretation and refinement. Test whether assets buffer stressor-related susceptibilities.
We tested effect modification in all-cause models described under 2.1 by urbanicity category (high-density urban, low-density urban, small town/ suburb, rural), and found that results were highly consistent across urbanicity categories, with strong positive associations on case day, and weaker associations thereafter.
In all urbanicity categories, we found larger relative risks among 5-13 year-olds than among young children and older teens. Per-unit risk were higher at lower PM2.5 concentrations, in all urbanicity categories.
Examining effect modification by poverty rate, we found stronger associations for PM2.5 in low-poverty (vs. high-poverty) tracts. Associations did not differ across quintiles of poverty rate in less densely-populated communities. We will likewise examine variation by quintiles of community assets, across urbanicity categories, once those indicators are prepared.
Aim 2.3: Use NYS future climate projections to estimate future child health impacts under different emissions scenarios to inform community engagement and integration.
The New York State Climate Impact Assessment report is expected to be released in December 2023, which will include updated climate projections which we can build into our community engagement work with both partners in focal cities as well as statewide youth.
Aim. 2.4: Integrate the stressor/asset findings into local and statewide climate adaptation and pediatric clinical initiatives. This work will occur as part of a future aim.
Challenges encountered & solutions: Dr. Allan Just changed institutions to Brown University, but maintains an adjunct role at Mt Sinai, and continues to supervise a programmer, to whom allocated funds were redirected, with no net impact on budget or subcontracts. There have been some delays in meeting with community partners to identify potential asset, those these meetings are now occurring more expediently, and in a more structured manner (e.g., Youth Climate Summit asset ‘voting’).
IRB approvals: The study was reviewed and approved by the IRB of the Icahn School of Medicine at Mt Sinai (STUDY-22-01227), which serves as the IRB of Record for this project. The Drexel University and University of Pittsburgh IRBs completed SMART IRB Reliance Agreements (Drexel IRB protocol #2302009749; Pitt IRB protocol #STUDY23070054). We also completed approvals for New York Statewide Planning and Research Cooperative (SPARCS), including data security protocols (SPARCS Amendment #2007-01).
Journal Articles:
No journal articles submitted with this report: View all 4 publications for this projectSupplemental Keywords:
Resilience, assets, community stressors, non-chemical stressors, child health, healthcare utilization, climate change, temperature, heat, ozone, PM2.5Relevant Websites:
PM2.5 exposure is associated with an increased risk of pediatric ED visits in New York State, particularly at low concentrations Exit
The perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.