Author ORCID Identifier

Defense Date


Document Type


Degree Name

Doctor of Philosophy



First Advisor

Theresa Swift-Scanlan

Second Advisor

Jeanne Salyer

Third Advisor

Terry Jones

Fourth Advisor

Ronald K. Elswick

Fifth Advisor

Maria deValpine


Introduction: Sepsis is a serious and life-threatening syndrome affecting 1.7 million Americans annually and resulting in approximately 270,000 deaths. An “obesity paradox” where obese individuals have lower sepsis mortality than their non-obese counterparts has been described. The problem is that the longevity of the effect is unknown, and few studies have examined the obesity paradox after 1-2 months post-diagnosis.

Methods: This retrospective cohort study examined clinical, demographic, and biomarker variables thought to affect sepsis mortality at three-time points: 30 days, 180 days, and at one year post-sepsis diagnosis in order to shed light on specific factors that might define a “sepsis survivor” phenotype. A convenience sample of adults age 18 and older admitted to an academic medical center between the years of 2007 to 2018 with a diagnosis of sepsis was identified. Simple logistic regression was used to test for significance between age, sex, race, c-reactive protein, lactate, white blood cells, body mass index, and sepsis severity on mortality at each of the three previously described time points. Variables with statistical and clinical significance were entered into multivariate logistic regression models to explore the contributions of each variable and interactions between variables at 30 days, 180 days, and one year after sepsis diagnosis.

Results: We found for every 5 unit increase in BMI, the odds of mortality were 0.92 (95% CI: 0.85, 0.99) times lower at 30 days since sepsis diagnosis. However, at 180-day and one-year post sepsis diagnosis, as BMI increased, there was an increase in odds of death for each sepsis type.

Discussion: In this dataset, it appears that the "obesity paradox" exists up to 30 days, but the protective effect of obesity on sepsis outcomes may not extend beyond one month. Future studies that control for comorbidities and other potential covariates, and that can test for the contributions of novel biomarkers on sepsis outcomes are needed.


© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission