The accuracy of a 4-item hydration self-assessment model to classify urine concentration using different cut-offs

Hydration is important for health and performance, but accurate self-assessment methods for active populations are limited. The study objective was to determine a model to classify by self-assessment a low vs. high 24-hour urine concentration, and to determine the accuracy of morning vs. afternoon assessments. The study describes the development and validation of a self-assessment model through exploratory modeling, using stepwise logistic regression. The aim was to determine which hydration markers predict urine specific gravity (using two different USG cut-offs: low ≤ 1.012, and high ≥ 1.020), in two intentionally overlapping 24-hour urine collections leading up to a morning and afternoon hydration self-assessment, and to assess its diagnostic accuracy based on the area under the curve (AUC).

A total of of n = 62 wildland firefighters (WLFFs), and n = 23 recreational athletes (12% female, median age 25 years, with interquartile range of 24-32) were included. The median USG values of the 24-hour urine collections leading up to the morning and afternoon assessments, were 1.011 and 1.012, respectively. The AUC for the final model, which included self-reported fluid intake; urine frequency, volume, and color, to classify low vs. high USG, for low cut-off (≤ 1.012) was fair in the morning (0.72), and good in the afternoon (0.85), and for high cut-off (≥ 1.020) good for morning (0.81) and afternoon (0.86), respectively.

This 4-item hydration self-assessment model classifies a low vs. high USG acceptable, with slightly higher accuracy in the afternoon. This insight may inform targeted hydration monitoring strategies in occupational settings, for example among WLFFs.

The online version contains supplementary material available at 10.1186/s44410-026-00024-y.