Relative effectiveness of high-dose vs. standard-dose quadrivalent influenza vaccine according to time of day of vaccination: a post-hoc analysis of the DANFLU-1 randomised clinical trial

Abstract Background Circadian variations have been shown to affect antibody response following vaccination. We sought to evaluate whether time of day of vaccination (ToV) modified the relative effectiveness of high-dose (QIV-HD) vs. standard-dose (QIV-SD) quadrivalent influenza vaccination, and further whether ToV was independently associated with hospitalisations and mortality. Methods This is a post-hoc analysis of the DANFLU-1 trial which was a randomized trial of QIV-HD vs. QIV-SD in adults aged 65-79 years conducted during the 2021-2022 influenza season. Vaccination took place between 7:05 AM and 8:08 PM. For this study, we stratified participants into early and late vaccination groups according to ToV before or after 11 AM, respectively. Prespecified outcomes included hospitalisations for pneumonia or influenza, respiratory disease, cardiovascular disease, and cardio-respiratory disease as well as all-cause hospitalisations and mortality. Hospitalisations were analysed as recurrent events using negative binomial regression, while mortality was analysed using Cox proportional hazard models, both of which were also used to test for interaction. For multivariable analysis, models were adjusted for baseline risk factors including sex, age, and multiple lung diseases. ToV was furthermore assessed as a continuous variable with 2-hour increments. Results In the final study population, the early group consisted of 5,371 (43%) participants (mean age 71.6±3.9 years, male sex 53.6%, QIV-SD 50.1%), while the late group consisted of 7,106 (57%) (mean age 71.8±4.0 years, male sex 52.4%, QIV-HD 49.9%). During follow-up, we observed 43 hospitalisations for pneumonia or influenza, 76 respiratory hospitalisations, and 62 mortalities. In recurrent event analysis, QIV-HD was associated with lower incidence of hospitalisation for pneumonia or influenza compared with QIV-SD irrespective of whether administered early (IRR 0.17 (95% CI 0.03-0.86), p=0.032) or late (IRR 0.37 (95% CI 0.16-0.88), p=0.024) (p-value for interaction = 0.69) (Figure 1). Effect estimates were consistent for all outcomes excluding cardiovascular hospitalisations regardless of early or late vaccination, and no effect modification was found (Figure 1). Interestingly, late vaccination was associated with higher incidence rate of respiratory hospitalisation (IRR 3.00 (95% CI 1.13-7.98), p=0.024, Figure 2), irrespective of vaccine type. The same association was found for ToV as a continuous variable (IRR 1.28 per 2-hour increment (95% CI 1.02-1.60), p=0.035). All above associations remained significant in multivariable analysis. Conclusion QIV-HD was associated with lower incidence of hospitalisation for pneumonia or influenza regardless of time of day for vaccination. In addition, later vaccination was associated with higher incidence of respiratory hospitalisation independent of vaccine type and known risk factors. Due to low power, however, this finding should be regarded as exploratory.Figure 1:Forest plotFigure 2:Incidence rates