because the biochemical and immune measures data are typically variable, baseline difference were likely to occur. Baseline measures were also examined for equality of means, and in the cases where means differed, the variable was entered as a covariate and an analysis of covariance was performed