Critical difference applied to exercise-induced salivary testosterone and cortisol using enzyme-linked immunosorbent assay (ELISA): distinguishing biological from statistical change.

Due to its noninvasive, convenient, and practical nature, salivary testosterone (sal-T) and cortisol (sal-C) are frequently used in a clinical and applied setting. However, few studies report biological and analytical error and even fewer report the 'critical difference' which is the change required before a true biological difference can be claimed. It was hypothesized that (a) exercise would result in a statistically significant change in sal-C and sal-T and (b) the exercise-induced change would be within the critical difference for both salivary hormones. In study 1, we calculated the critical difference of sal-T and sal-C of 18 healthy adult males aged 23.2 ± 3.0 years every 60 min in a seated position over a 12-h period (08:00-20:00 hours [study 1]). As proof-of-concept, sal-C and sal-T was also obtained pre and at 5 and 60 min post a maximal exercise protocols in a separate group of 17 healthy males (aged 20.1 ± 2.8 years [study 2]). The critical difference of sal-T calculated as 90 %. For sal-C, the critical difference was 148 % (study 1). Maximal exercise was associated with a statistically significant (p < 0.05) changes in sal-T and sal-C. However, these changes were all within the critical difference range. Results from this investigation indicate that a large magnitude of change for sal-C and sal-T is required before a biologically significant mean change can be claimed. Studies utilizing sal-T and sal-C should appreciate the critical difference of these measures and assess the biological significance of any statistical changes.