Evaluation of determinants of the serological response to the quadrivalent split‐inactivated influenza vaccine

The seasonal influenza vaccine is only effective in half of the vaccinated population. To identify determinants of vaccine efficacy, we used data from > 1,300 vaccination events to predict the response to vaccination measured as seroconversion as well as hemagglutination inhibition (HAI) titer levels one year after. We evaluated the predictive capabilities of age, body mass index (BMI), sex, race, comorbidities, vaccination history, and baseline HAI titers, as well as vaccination month and vaccine dose in multiple linear regression models. The models predicted the categorical response for > 75% of the cases in all subsets with one exception. Prior vaccination, baseline titer level, and age were the major determinants of seroconversion, all of which had negative effects. Further, we identified a gender effect in older participants and an effect of vaccination month. BMI had a surprisingly small effect, likely due to its correlation with age. Comorbidities, vaccine dose, and race had negligible effects. Our models can generate a new seroconversion score that is corrected for the impact of these factors which can facilitate future biomarker identification. Synopsis Computational modeling quantifies the effects of confounding factors on the serological response to flu vaccination in large human cohorts and reveals a differential impact of prior vaccination status, recipient age, and the month of vaccination. Using an extensive cohort dataset, a computational framework was developed to predict the serological response to flu vaccination in three subpopulations of different ages. The accuracy for a categorical response is 74% or higher for all three subpopulations. Vaccination history, baseline titer levels, participant age, and vaccination month are the major predictors of seroconversion, while baseline is the predominant predictor of the long‐term response. Body mass index only has a marginal effect on seroconversion in adults younger than 65 years old, likely due to its substantial correlation with participant age. Graphical Abstract Computational modeling quantifies the effects of confounding factors on the serological response to flu vaccination in large human cohorts and reveals a differential impact of prior vaccination status, recipient age, and the month of vaccination.