Assessment of the MX3 hydration testing system in adults at rest and following exercise in the heat.

The MX3 hydration testing system demonstrated low reliability for measuring pre-exercise saliva osmolality.

• Pre-exercise mean difference (MD) was 4 mOsm/kg with 95% limits of agreement (LoA) of -46 to 54 mOsm/kg
• Absolute difference (ABD) was 23 mOsm/kg and coefficient of variation (CV) was 31%
• Standard error of measurement (SEM) was 18 mOsm/kg
• Intraclass correlation coefficient (ICC) was 0.198, indicating poor reliability
• Twenty adults completed two matched exercise trials in warm-humid conditions

The MX3 hydration testing system demonstrated low reliability for measuring post-exercise saliva osmolality.

• Post-exercise MD was 5 mOsm/kg with 95% LoA of -78 to 68 mOsm/kg
• ABD was 28 mOsm/kg and CV was 29%
• SEM was 26 mOsm/kg
• ICC was 0.391, indicating poor-to-weak reliability

Pre-exercise saliva osmolality measured by MX3 showed poor diagnostic accuracy for identifying hypohydration relative to urine specific gravity.

• Sensitivity was 0.09, indicating very low ability to correctly identify hypohydrated individuals
• Specificity was 0.82 relative to USG >1.020 as the criterion measure
• Positive likelihood ratio (LR+) was 0.51 and negative likelihood ratio (LR-) was 1.11
• LR values near 1.0 and below indicate poor discriminative ability

Post-exercise saliva osmolality measured by MX3 showed weak diagnostic accuracy for identifying hypohydration relative to body mass loss.

• Sensitivity was 0.57 and specificity was 0.63 relative to >2% body mass loss (BML) as the criterion measure
• Positive likelihood ratio (LR+) was 1.52 and negative likelihood ratio (LR-) was 0.69
• These likelihood ratio values indicate weak predictive value for identifying post-exercise hypohydration

The study used a two-trial repeated-measures design to assess both reliability and diagnostic accuracy of the MX3 system.

• Twenty adults completed two matched exercise trials in warm-humid conditions
• Pre- and post-exercise saliva osmolality (Sosm) were assessed using MX3 on both occasions
• Reliability was assessed using MD, ABD, CV, SEM, ICC, and 95% LoA
• Diagnostic accuracy was assessed using crosstab analysis relative to urine specific gravity (USG) and body mass loss (BML)