A Spatial and Contextual Exposome-Wide Association Study of COVID-19 Hospitalizations in Florida, United States

Hui Hu Ph.D.

Assistant Professor of Medicine, Harvard Medical School

Associate Epidemiologist, Brigham and Women's Hospital

Channing Division of Network Medicine

September 20, 2022

Risk factors for severe COVID-19

• Known risk factors:
  • Age
  • Race/ethnicity
  • Underlying medical condition (e.g., asthma, cancer, etc.)
  • Physical inactivity
  • Smoking
     
• Spatial and contextual factors may be important determinants:
  • Air pollution
  • Green space
  • Neighborhood SES
     
• Gaps:
  • Most studies focused on individual exposures / exposures from a single class
  • Most existing studies on spatial and contextual factors were ecological and based on aggregated COVID-19 outcomes

Study aim

• To conduct a spatial and contextual exposome-wide association study (ExWAS) of COVID-19 hospitalization
   
  • ​​Individual-level electronic health records data of COVID-19 patients
     
  • Linked data from multiple sources to characterize patients’ long-term exposures to the spatial and contextual exposome based on their geocoded residential histories

• OneFlorida+ CRN:
  • 14 academic institutions and health systems
  • ~16 million patients
     
• We obtained data from 50,368 patients with:
  • ≥18 years old
  • A SARS-CoV-2 PCR/Antigen lab test or a COVID-19 diagnosis in 03/2020-10/2021
  • ≥1 inpatient encounter or 2 outpatient encounters (≥3 months apart) within 5 years before COVID-19 onset
  • Geocoded residential history
     
• COVID-19 hospitalization:
  • First hospital admission within -7 days to +15 days after patients’ first COVID-19 positive date

Spatial and contextual exposome measures

• Area- and time-weighted averages:
  • Patients' geocoded residential history within 5 years before COVID-19 positive date
  • 250m circular buffer around each address
     
• A total of 194 exposome factors covering 10 categories

Covariates

• Sociodemographic factors: age, gender, race/ethnicity, health insurance
• Underlying medical conditions:
  • Atherosclerotic cardiovascular disease
  • Myocardial infarction
  • Hypertension
  • Peripheral vascular disease
  • Cerebrovascular disease
  • Diabetes mellitus
  • Chronic obstructive pulmonary disease
  • Asthma
  • Cancer
  • Chronic kidney disease
  • Renal disease
  • Organ transplant
• Time-series county-level COVID-19 related factors:
  • ​Number of days since first COVID-19 case
  • County-level COVID-19 vaccination rates (i.e., at least one dose, fully vaccinated)
  • County-level hospital bed capacity

Statistical analysis

• Phase 1:
  • Mixed effect logistic regression models were fitted for each exposure after adjusting for all the potential confounders with a random intercept by county
  • Benjamin-Hochberg procedure was used to control the false discovery rate (FDR) at 5%
     
• Phase 2:
  • A multivariable mixed-effect logistic regression model including all significant exposures from Phase 1

• Four variables remained significant in Phase 2:
  • 2-Chloroacetophenone
  • Low food access
  • Neighborhood deprivation
  • Density of fitness and recreational sports center
     

• First spatial and contextual ExWAS of COVID-19 hospitalization

• Confirmed a previously reported factor: neighborhood deprivation

• Also generated unexpected predictors that may warrant more focused evaluation

Ackowledgements

• Team:
  • BWH/HMS/HSPH: Yi Zheng, Jaime Hart, Francine Laden, Peter James
  • UF: Jiang Bian, Jennifer Fishe, Jingchuan Guo, William Hogan, Elizabeth Shenkman
• Funding:
  • NIH/NIEHS R21ES032762
  • NIH/NHLBI K01HL153797, OTAHL161847
  • NIH/NCATS UL1TR001427, TL1TR001428, KL2TR001429 
  • PCORI CDRN-1501-26692, RI-CRN-2020-005
  • FDOH 4KB16

Thank you!

hui.hu@channing.harvard.edu

hui-hu.com

github.com/benhhu