Statistical analysis.

D’Agostino-Pearson normality tests indicated that values on most days were not normally distributed. All clinical, viral, antibody, and cellular immune parameters were compared using repeated-measure analyses of variance (ANOVAs) with Tukey’s post tests for multiple comparisons between the three groups (control/Ab4, Ab4ΔORF2/Ab4, Ab4/Ab4). Clinical parameters and virus detection were analyzed until days 10 and 18 p.i., respectively (Fig. 1). Statistical analyses of the nasal mediators (Fig. 2), nasal antibodies (Fig. 3), and secreted cytokines from restimulated PBMC (Fig. 5A and ​andB)B) were limited to day –2 to day 31 p.i. Correlations of prechallenge infection values of EHV-1 gC-specific serum Ig with infection outcomes (fever, clinical disease, virus shedding, and viremia at the times these reached peak values in infected horses) were analyzed using Spearman rank correlations. P values of <0.05 were considered significant. The statistical analysis was performed, and graphs were created using GraphPad Prism 7 for Mac OS X, version 7e.