Summated training and match load predictors of salivary immunoglobulin-A, alpha-amylase, testosterone, cortisol and T:C profile changes in elite-level professional football players: A longitudinal analysis

We examined how summated training and match load measures relate to salivary immunological and hormonal profile changes in professional football players. Data were collected from 18 elite-level professional male football players from one English Championship team across a complete 40 wk competitive season. Daily training (micro-technology) and match (computerised tracking) measures of total, high-speed and high-metabolic load running distance and sprint, acceleration, deceleration and sRPE load were converted into exponentially weighted moving average "acute" (7d), "chronic" (28d) and acute:chronic composite load measures. Bi-weekly morning saliva samples were analysed for immunoglobulin-A, alpha-amylase, testosterone, cortisol and testosterone:cortisol. A two-stage data reduction technique using partial least squares modelling and a backward stepwise selection procedure determined the most parsimonious model for each salivary variable. Testosterone had non-linear relationships with chronic total (P = 0.015; Cohen's D: large), high-metabolic load (P = 0.001;small) and high-speed (P = 0.001;trivial) running distance and linear relationships with chronic sRPE (P = 0.002;moderate down arrow) and acute:chronic high-speed running distance (P = 0.001; trivial up arrow). Cortisol had a non-linear relationship with chronic high-speed running distance (P = 0.001;trivial). Testosterone:cortisol had non-linear relationships with chronic decelerations (P = 0.039;small) and chronic summated acceleration and deceleration load (P = 0.039;small). Non-linear relationships typically indicated optimal hormonal responses at squad mean loads. No load variables clearly related to salivary immunoglobulin-A or alpha-amylase changes. We conclude that chronic total and high-intensity load measures relate to hormonal changes and might be useful indicators of player readiness. Acute load variables were not related to immunological or hormonal changes and consequently, should not be used as surrogate measures of player readiness in isolation.