"Mediation and Moderation of Influenza Vaccination on the Perceived Severity of Acute Respiratory Illness", Department of Epidemiology, School of Public Health. 

Committee

• G.K. Galasubramani, (advisor) EPI
• Catherine L. Haggerty, EPI
• Marnie Bertolet, EPI
• Richard K. Zimmerman, Family Medicine
• Mary Patricia Nowalk, Family Medicine
• Brendan Flannery, CDC

Abstract

Background: Influenza causes substantial outpatient morbidity and healthcare utilization each year. Influenza outcomes are shaped by complex relationships among host characteristics, vaccination, infection, viral load, symptoms, and care-seeking behavior. Although annual influenza vaccination is the primary strategy for reducing influenza-associated illness and complications, vaccine protection is incomplete and varies across seasons, viral subtypes, and population groups. A large body of prior research has evaluated influenza vaccine effectiveness against laboratory-confirmed infection and medically attended illness. However, less is known about how vaccination relates to patient-reported illness severity among people seeking outpatient care for acute respiratory illness. In addition, the relationships among influenza vaccination, influenza infection, viral load, demographic characteristics, and perceived illness severity remain incompletely understood.

Methods: Outpatients with acute respiratory illness (ARI) were enrolled across nine influenza seasons, from 2011–2012 through 2019–2020, through the U.S. Influenza Vaccine Effectiveness Network. Respiratory specimens were collected for influenza testing by RT-PCR. Influenza viral load was assessed using cycle threshold (Ct) values, with lower Ct values indicating higher viral load. Perceived illness severity was assessed at enrollment using the EQ-5D-5L Visual Analog Scale. Influenza vaccination status was ascertained from verified records and participant report. Demographic characteristics and symptoms were collected at enrollment. Linear regression, logistic regression, and mediation analyses were used for statistical modeling.

Results: The first aim assessed whether demographic factors modified influenza vaccine effectiveness and the associations among vaccination, influenza infection, and perceived illness severity. Among 34,937 participants, 9,830 (28%) tested positive for influenza and 18,205 (52%) were vaccinated. Overall VE against medically attended laboratory-confirmed influenza was 33%. VE did not significantly vary by age group. However, among adults aged ≥65 years, VE was higher in males than females (44% vs 18%), and VE against influenza A was lower in White compared to non-White participants (18% vs 50%). Influenza vaccination was associated with a 0.94-unit reduction in severity, whereas influenza infection was associated with a 6.44-unit increase in severity. The association between influenza infection and perceived severity varied by age and race/ethnicity, with a smaller difference in severity between influenza positive and negative patients among adults ≥65 years than younger ages and a larger difference among non-White participants than white participants.

The second aim examined associations among influenza vaccination, viral load, symptoms, and perceived influenza severity among influenza-positive outpatients. Influenza vaccination was significantly associated with lower influenza A Ct values (β = -0.44, SE = 0.12; p < 0.001) (higher viral load). Higher influenza A Ct values (lower viral load) were significantly associated with lower perceived influenza severity (β = -0.49, SE = 0.06; p < 0.001), and reduced odds of fever, sore throat, nasal congestion, fatigue, wheeze, and shortness of breath. For influenza B, higher Ct values were associated with lower perceived influenza severity (β = -0.20, SE = 0.10; p = 0.035), and decreased odds of fever and nasal congestion. Vaccination was associated with lower perceived severity, however, its association with lower influenza A Ct values partially attenuated the association between vaccination and lower perceived severity.

The third aim used mediation analysis to evaluate pathways linking demographic factors, influenza vaccination, influenza infection, symptoms, and perceived severity of acute respiratory illness. Influenza vaccination was associated with an overall 1.47-unit reduction in ARI severity score (95% CI: -1.93, -1.00). Approximately 60% of the association was statistically mediated through prevention of influenza virus infection while the remaining 40% represented the direct effect of vaccination. Vaccine uptake, influenza infection risk, and perceived severity varied across demographic groups. Vaccination was also associated with lower perceived severity of ARI through reduced symptom burden.

Conclusion: Building on these findings, this dissertation demonstrates that the impact of influenza vaccination extends beyond the prevention of laboratory-confirmed influenza infection and includes patient-reported illness severity. Vaccine effectiveness and perceived illness severity varied across demographic groups, and influenza viral load was associated with clinical severity among infected participants. However, the observed association between vaccination and perceived illness severity was modest, and its clinical significance warrants further investigation. This dissertation is also limited by the observational nature of the surveillance data, potential residual confounding, and restricted generalizability to medically attended outpatient ARI populations at participating sites.

By integrating vaccine effectiveness, demographic heterogeneity, viral load, symptom burden, and patient-reported severity, this work contributes to a broader framework for evaluating influenza vaccine impact. These findings support continued annual influenza vaccination and highlight the importance of incorporating patient-centered and equity-focused outcomes into influenza vaccine research, surveillance, and public health communication.