Purpose: To evaluate the classification performance of a new method to individualize reference ranges for biomarkers of muscle recovery in elite handball. Methods: In a longitudinal observational study, creatine kinase (CK) and urea levels were assessed in 16 male elite handball players during the 2019-20 preseason. Blood samples were collected at prespecified time points when players were considered either recovered or nonrecovered. Initially, linear mixed-effects models were calculated, as outlined in the study design, to examine the effect of recovery status on CK and urea levels. Finally, a fixed-effect model was calculated for urea based on the law of parsimony. Individualized reference ranges were calculated using a recently published algorithm. To investigate whether the individualized approach resulted in a more accurate classification of recovery status (recovered or nonrecovered), it was compared with a group-based approach derived from the same data set, utilizing predefined error rates. Results: Linear mixed-effects models showed a large effect of recovery status on CK (P < .001, d = 3.49) and a small effect on urea (P = .018, d = 0.382). In contrast to CK (P = .017), urea had no significant interindividual variation. Hence, individualization was examined only for CK. The numerical decrease in both CK error rates by the individualized approach was significant for the test-pass error rate (P = .0196, phi = .19). Conclusions: Our findings underscore the critical role of CK for monitoring in team sports such as handball. The observed improvement in CK error rates suggests a promising opportunity to individualize biochemical monitoring, although further studies encompassing larger sample sizes are warranted.
© Copyright 2022 International Journal of Sports Physiology and Performance. All rights reserved.