| description abstract | Dehydration is a common problem in healthy individuals as well as the elderly and chronically ill. People are often poorly attuned to hydration, and despite widespread awareness of the problem, fatal and near-fatal episodes occur frequently. Typical indicators of hydration status include changes in body weight, urine specific gravity, blood plasma levels, and bioelectrical impedance. Challenges to estimating hydration status from these indicators include the invasive nature of some methods as well as the cost and time required. We have developed a noninvasive device for monitoring hydration status. Our design is inspired by the traditional clinical protocol that approximates fluid loss on the order of 1-2% dehydration by assessing radial pulse before and after a supine to standing transition. The prototype comprises an inertial measurement unit (Xsens MTi) and a wearable heart rate monitor (Polar S810i). In order to compare heart rate behavior under normal and low hydration levels, fluid loss equivalent to 1-4% of the baseline body weight was induced by exercise in three healthy subjects during two data collection sessions. In the first (control) session, subjects replaced fluids every 15 minutes during exercise to maintain their body weight within 0.2% of their baseline value. Fluids were not replaced during the second (test) session, and subjects lost an average of 1.2% of their body weight. Heart rate and body position measurements were recorded before and after exercise while subjects performed repeated supine-to-standing movements and knee-to-chest stretching exercises (supine position only). All post-processing was performed using MATLAB (The MathWorks). Average heart rate was calculated over a 10 second period. Pilot data demonstrates the device's ability to detect hydration changes on the order of 1% in one-third the time required by the traditional clinical protocol (30 seconds compared to 90 seconds). The average rise time from baseline to maximum heart rate and the maximum heart rate following supine-to-standing transitions were significantly longer and greater, respectively, in the dehydrated subjects. Although not statistically significant, the average heart rate during knee-to-chest stretching exercises was elevated in the dehydrated state. | |
