Dissertations and Theses

Date of Degree

9-1-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy (Ph.D.)

Department

Epidemiology and Biostatistics

Advisor(s)

Denis Nash

Committee Members

Kelly Gebo

Zach Shahn

McKaylee Robertson

Keywords

Long COVID, community-based cohort, symptom clusters, symptom natural history, vaccine

Abstract

BACKGROUND: Long COVID, or post-acute sequelae of SARS-CoV-2 infection (PASC), represents a growing public health concern with diverse symptom profiles, unclear risk factors, and evolving evidence on potential prevention strategies. Most studies of Long COVID have relied on electronic health records or convenience samples, which may underestimate mild or community-based cases and lack standardized follow-up, and often do not include pre-infection health information or complete infection histories. This dissertation leverages prospective data from a national community-based cohort to characterize the heterogeneity of Long COVID symptoms, assess symptom trajectories over time, and evaluate the effect of COVID-19 third dose vaccination on Long COVID risk.

METHODS: This dissertation includes three studies drawing on longitudinal data from the CHASING COVID Cohort Study, a prospective cohort initiated in March 2020. Participants completed repeated surveys and provided serology specimens across multiple time points. In the first study, unsupervised clustering methods were applied to identify distinct symptom subgroups among individuals with Long COVID, using longitudinal, self-reported symptom data from 3 to 12 months post-infection. In the second study, we used a target trial emulation (TTE) approach with inverse probability weighting to estimate the risk of developing specific symptom clusters—neurological, autonomic, and exercise intolerance—at 4–8 and 9–12 months post-infection, comparing symptoms between infected individuals and contemporaneously uninfected controls. In the third study, TTE with inverse probability weighting was used to estimate the effect of receiving at least a third COVID-19 dose—compared to receiving the primary series only—on the risk of developing Long COVID, accounting for baseline and time-varying confounders.

RESULTS: Clustering analyses identified three distinct longitudinal symptom burden trajectories—low, moderate, and high—capturing variability in number of symptoms and symptom prevalence. Overall Long COVID risk was higher among infected versus uninfected individuals at 4–8 months (aRD: 11.3% [95% CI: 9.2-13.5%]; aRR: 2.01 [95% CI: 1.81 – 2.20]) and 9–12 months (aRD: 6.7% [95% CI: 4.6-8.9%]; aRR: 1.54 [95% CI: 1.37-1.72]). Neurological, autonomic, and exercise intolerance symptoms were significantly elevated among infected versus contemporaneously uninfected individuals across both follow-up periods, with adjusted risk differences (aRDs) ranging from 3.2% to 7.2% and adjusted risk ratios (aRRs) from 1.48 to 2.10. Receipt of a third vaccine dose was not associated with a statistically significant reduction in Long COVID risk at either 6 or 12 months compared to the primary series alone (aRD: -0.1%; 95% CI: -0.5% – 0.4%; aRR: 0.93; 95% CI: 0.54 – 1.44 at 6 months; aRD: 0.4%; 95% CI: -0.5% – 1.4%; aRR: 1.09; 95% CI: 0.90 – 1.33 at 12 months), suggesting limited additional protection among community-dwelling adults.

DISCUSSION: This dissertation provides robust evidence that Long COVID is a heterogeneous condition with distinct symptom phenotypes and prolonged impact on the health of community-based populations. By applying causal inference methods and prospective symptom tracking, this work advances the epidemiologic understanding of Long COVID and highlights the need for targeted evidence-based care strategies to specific symptom domains—such as cognitive dysfunction, autonomic symptoms, and post-exertional malaise—which may have different underlying mechanisms and treatment needs. While updated COVID-19 vaccination remains essential for preventing severe COVID-19, its added value in mitigating long-term sequelae appears limited, emphasizing the need for complementary preventive and therapeutic strategies. Together, these findings inform clinical practice, public health planning, and future research on post-viral syndromes.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.