Individual patterns in blood-borne indicators of fatigue - trait or chance.

Blood-borne markers of fatigue such as Creatine Kinase (CK) and Urea (U) are widely used to fine-tune training recommendations. However, predictive accuracy is low. A possible explanation for this dissatisfactory characteristic is the propensity of athletes to react with different patterns of fatigue indicators (e.g. predominantly muscular (CK) or metabolic (U)). The aim of the present trial was to explore this hypothesis by using repetitive fatigue-recovery cycles. 22 elite junior swimmers and triathletes (18 ±3 years) were monitored for nine weeks throughout two training phases (low-intensity, high-volume (LIHV) and high-intensity, low-volume (HILV)). Blood samples were collected each Monday (recovered) and Friday (fatigued) morning. From measured values of CK, U, free-testosterone (FT), and cortisol (C) as determined in the rested and fatigued state, respectively, Monday-to-Friday differences (Δ) were calculated and classified by magnitude before calculation of ratios (ΔCK/ΔU and ΔFT/ΔC). Coefficient of variation (CV) was calculated as group-based estimates of reproducibility. Linear mixed modelling was used to differentiate inter- and intra-individual variability. Consistency of patterns was analysed by comparison to threshold values (<0.9 or >1.1 for all weeks). Reproducibility was very low for fatigue-induced changes (CV ≥100%) with inter-individual variation accounting for 45-60% of overall variability. Case-wise analysis indicated consistent ΔCK/ΔU patterns for seven individuals in LIHV and seven in HILV; five responded consistently throughout. For ΔFT/ΔC the number of consistent patterns was two in LIHV and three in HILV. These findings highlight the potential value of an individualised and multivariate approach in the assessment of fatigue.