Assessment of exercise-induced dehydration in underhydrated athletes: Which method shows the most promise?

Background & aims: Athletes are commonly exposed to exercise-induced dehydration. However, the best method to detect dehydration under this circumstance is not clear. This study aimed to analyze pre- and post-dehydration measurements of biomarkers, including saliva osmolality (SOsm), urine osmolality (UOsm), urine-specific gravity (USG), urine color (Ucolor), serum osmolality (SeOsm), serum arginine vasopressin (AVP), serum sodium (Na+), and thirst sensation in underhydrated athletes, using the body mass loss (BML) as the reference method. Methods: In this clinical trial (NCT05380089), a total of 38 athletes (17 females) with a regular low water intake (<35 mL/kg/day) were submitted to exercise-induced dehydration with a heat index of 29.8 ± 3.1 °C and an individualized running intensity (80–90% of first ventilatory threshold). Results: ROC curve analysis revealed significant discriminative abilities of SOsm, with AUC values of 0.76 at 1.5% BML, 0.75 at 1.75% BML, and 0.87 at 2% BML, while Na+ and SeOsm showed the highest AUC of 0.87 and 0.91 at 2% BML, respectively. SOsm showed high sensitivity at 1.5% of BML, while SeOsm and Na+ demonstrated high sensitivity at 2% of BML. Conclusion: This study highlights SOsm as a potential indicator of hydration status across different levels of BML. Additionally, Na+ and SeOsm emerged as accurate dehydration predictors at 1.75% and 2% of BML. Notably, the accuracy of urinary indices and thirst sensation for detecting hydration may be limited.