(1)H nuclear magnetic resonance (NMR) as a tool to measure dehydration in mice

Dehydration is a prevalent pathology, where loss of bodily water can result in variable symptoms. Symptoms can range from simple thirst to dire scenarios involving loss of consciousness. Clinical methods exist that assess dehydration from qualitative weight changes to more quantitative osmolality measurements. These methods are imprecise, invasive, and/or easily confounded, despite being practiced clinically. We investigate a non-invasive, non-imaging (1)H NMR method of assessing dehydration that attempts to address issues with existing clinical methods. Dehydration was achieved by exposing mice (n = 16) to a thermally elevated environment (37 °C) for up to 7.5 h (0.11-13% weight loss). Whole body NMR measurements were made using a Bruker LF50 BCA-Analyzer before and after dehydration. Physical lean tissue, adipose, and free water compartment approximations had NMR values extracted from relaxation data through a multi-exponential fitting method. Changes in before/after NMR values were compared with clinically practiced metrics of weight loss (percent dehydration) as well as blood and urine osmolality. A linear correlation between tissue relaxometry and both animal percent dehydration and urine osmolality was observed in lean tissue, but not adipose or free fluids. Calculated R(2) values for percent dehydration were 0.8619 (lean, P < 0.0001), 0.5609 (adipose, P = 0.0008), and 0.0644 (free fluids, P = 0.3445). R(2) values for urine osmolality were 0.7760 (lean, P < 0.0001), 0.5005 (adipose, P = 0.0022), and 0.0568 (free fluids, P = 0.3739). These results suggest that non-imaging (1)H NMR methods are capable of non-invasively assessing dehydration in live animals.