Abstract
BACKGROUND/OBJECTIVES: In sedentary adults, hydration is mostly influenced by total fluid intake and not by sweat losses; moreover, low daily fluid intake is associated with adverse health outcomes. This study aimed to model the relation between total fluid intake and urinary hydration biomarkers. SUBJECTS/METHODS: During 4 consecutive weekdays, 82 adults (age, 31.6±4.3 years; body mass index, 23.2±2.7 kg/m(2); 52% female) recorded food and fluid consumed, collected one first morning urine (FMU) void and three 24-h (24hU) samples. The strength of linear association between urinary hydration biomarkers and fluid intake volume was evaluated using simple linear regression and Pearson's correlation. Multivariate partial least squares (PLS) modeled the association between fluid intake and 24hU hydration biomarkers. RESULTS: Strong associations (|r| ≥ 0.6; P<0.001) were found between total fluid intake volume and 24hU osmolality, color, specific gravity (USG), volume and solute concentrations. Many 24hU biomarkers were collinear (osmolality versus color: r=0.49-0.76; USG versus color: r=0.46-0.78; osmolality versus USG: 0.86-0.97; P<0.001). Measures in FMU were not strongly correlated to intake. Multivariate PLS and simple linear regression using urine volume explained >50% of the variance in fluid intake volume (r(2)=0.59 and 0.52, respectively); however the error in both models was high and the limits of agreement very large. CONCLUSIONS: Hydration biomarkers in 24hU are strongly correlated with daily total fluid intake volume in sedentary adults in free-living conditions; however, the margin of error in the present models limits the applicability of estimating fluid intake from urinary biomarkers.