Dissertations and Theses

Date of Degree

5-1-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy (Ph.D.)

Department

Environmental, Occupational, and Geospatial Health Sciences

Advisor(s)

Ghada A. Soliman

Committee Members

Andrew Maroko

Demetra Tsapepas

Subject Categories

COVID-19 | Environmental Health | Environmental Public Health | Public Health | Pulmonology

Keywords

COVID-19, Interstitial Lung Disease (ILD), Spatial Analysis, Health Disparities, Mendelian Randomization, Lung Function, Geographic Information Science (GISc), Social Determinants of Health, Environmental Inequality, New York City

Abstract

Introduction:

This dissertation explores the interrelationship between Interstitial Lung Disease (ILD), COVID-19, and environmental and social determinants of health in New York City. ILD, a group of chronic lung conditions, heightens vulnerability to respiratory infections, including COVID-19. It represents a medically and socially vulnerable population already burdened by chronic, progressive lung disease and unequal access to care. The research focuses on spatial and population-level disparities during the pandemic's initial phase, considering factors like race, income, housing, and air quality. By examining geographic clustering and the burden of disease on socially and environmentally disadvantaged populations, the study highlights the compounding risks patients with ILD face amid a global health crisis. It emphasizes the urgent need for data-driven, equity-focused public health interventions that address structural inequality.

Objectives:

The primary objective of this dissertation is to investigate the spatial and individual-level factors influencing COVID-19 outcomes in populations with ILD. The research is structured around three aims: (1) mapping COVID-19 clustering and associating it with geographic and socioeconomic characteristics in New York City; (2) analyzing individual-level patients with ILD risk factors stratified by smoking status using multivariable logistic regression; and (3) assessing the causal relationship between ILD and COVID-19 through bidirectional Mendelian Randomization (MR). These aims collectively seek to uncover both environmental and genetic contributors to disease vulnerability, advancing understanding of how patients with ILD were disproportionately affected during the pandemic.

Methods:

This study used a combined geospatial analysis, statistical modeling, and genetic epidemiology approach. For Aim 1, spatial cluster analysis using Getis-Ord statistics identified COVID-19 Hot and Cold Spots across New York City modified ZIP Code Tabulation Areas (MODZCTAs), integrating air quality, housing, and socioeconomic status data. Aim 2 used electronic medical records from a multicenter New York City hospital to analyze patients with ILD data with regression models, stratifying by smoking status. Aim 3 applied bidirectional two-sample Mendelian Randomization using genetic variants to test potential causal relationships between ILD, lung function traits, and COVID-19 outcomes.

Results:

Aim 1 identified significant spatial clustering of COVID-19 cases across Modified ZIP Code Tabulation Areas (MODZCTAs) in New York City using Getis-Ord spatial statistics. Hot Spots were concentrated in neighborhoods with high population density, overcrowded housing, lower median income, and higher proportions of Black and Hispanic residents. Aim 2 showed that among nonsmoking patients with ILD, lower oxygen saturation—a proxy for disease severity—was significantly associated with older age, higher BMI, male sex, and being of Black or Hispanic ethnicity. Multivariable logistic regression models demonstrated that social determinants such as poverty and lack of employment were also predictors of lower oxygen saturation.

Aim 3 revealed no evidence for a causal link between genetically predicted ILD and COVID-19 infection or hospitalization outcomes. However, higher genetically predicted FEV1/FVC ratios were significantly associated with increased COVID-19 hospitalization risk (β = 0.207, p = 0.0017), and FVC showed a more modest association (β = 0.087, p = 0.048), suggesting that both obstructive and restrictive lung patterns may influence COVID-19 severity. Furthermore, reverse MR analyses supported clinical evidence of long-term pulmonary impairment post-COVID-19 infection, with COVID-19 hospitalization associated with increased risk of ILD (β = 0.216, p = 0.013) and infection or hospitalization associated with decreased lung function/FVC (β = –0.043, p = 0.002).

These results highlight the environmental, clinical, and genetic vulnerabilities contributing to COVID-19 severity in populations with ILD, reinforcing the need for targeted public health strategies.

Conclusions:

This comprehensive study enhances our knowledge of the relationships between ILD, COVID-19, and individual-level attributes. The identified geographic patterns, risk factors, and causal associations contribute valuable insights for public health strategies and resource allocation and guide future research endeavors in this critical area, calling for bold, equity-centered action to protect vulnerable populations.

Download

Available for download on Friday, May 19, 2028

Share

COinS