Search for a command to run...
Hydration is vital for the regulation of temperature and metabolism in the human body. Variation in hydration has a clear association with morbidity and mortality [1, 2]. Other studies have shown that mild dehydration has a significant influence on functional status, including increased fatigue and lower cognitive function [1, 2]. Dehydration can also lead to hospitalizations and put a burden on hospital systems. Measurement of hydration is typically inferential. Body water exists in muscle, ligament, fat tissues, intravascular and extravascular spaces, and both inside and outside of cells. Multiple approaches are used clinically to measure hydration including serum osmolality, measurement of body weight and body weight changes, fluid monitoring (e.g., urine output, oral and intravenous input), radiographic measurement, laboratory chemical assessment of fluids composition and as in our study, via bioimpedance [2]. Cognitive status encompasses various aspects of brain functioning [3]. When 2% or more of body weight is lost due to water restriction, heat, or physical exertion, a decline in cognitive performance can occur (short-term memory, attention, and visual-motor tracking) [4, 5]. However, inconsistent conclusions have been made about the extent of the association between cognitive performance and hydration status. Studies have shown the need for more examination of this phenomenon with larger and more diverse samples as well as differing causes for water loss [5]. This study serves as a foundation for future research. The primary aim of this project was to determine the cognitive status relationship with dehydration. Our secondary aim was to determine the interest of participants in wearing a device that detects hydration levels. TritonX has developed a wearable wrist-worn device, illustrated in Figure 1, using bioimpedance spectroscopy (BIS) technology to provide real-time measurements of hydration, using segmental, upper body measurement (Table A1). The device is designed to be worn on a patient's wrist consisting of a processor, BIS sensor, eight metal electrodes, battery, Bluetooth low-energy, vibration/buzzer, and LED indicators (see Figure 1 and Figure A1). Bluetooth connectivity and an internet-connected smartphone were used for cloud-based data transmission. The figure below illustrates the precise placement of electrodes. BIS is a method that assesses fluid volume status where there are exciting and sensing electrodes where the alternating current is emitted and the output is received. As the current passes through body tissue, the measurement of electrical impedance is assessed and fluid volume status is measured [1]. BIS interference is caused by prolonged occlusion, skin thickness, differences in circulation, fever, surface effects, and salts. A convenience sample of adults, 18 years and older (n = 34) with no prior history of fluid retention issues, pregnancy, body weight not > 158.8 kg (350 lbs), no current abnormal bodily fluid retention, no neuromotor disorders, no current congestive heart failure or heart attack/myocardial infarction in the past 12 months, no implanted electronic devices, or neurological disorders. Mental health disorders were not screened. Demographics, hydration concern, and interest were documented in a survey following measurement. The average from three back-to-back device measurements was documented and participants were instructed to remain still and fasting between readings. Following the measurement, cognitive ability was measured using the Stroop exam, which assesses cognitive interference, known as the Stroop Effect, where the color of a word mismatches its meaning (e.g., “green” printed in blue ink) [6-8] Our primary outcome for analysis was the ECW fraction, ECW/BW, where ECW = extracellular water, and BW = body weight. The ECW fraction is used as a hydration assessment, representing overhydration within the body if elevated and recently used as a diagnostic variable of body water deficit [9, 10]. The program SPSS Version 29 was used to estimate the bivariate (Pearson's r) correlation using bootstrapped 95% confidence intervals, p ≤ 0.05 was considered statistically significant, and generated other descriptive statistics [11]. The study protocol and instruments were reviewed and approved by the MetroHealth Institutional Review Board. Prior to the study, all participants completed informed consent. The study sample included 34 participants: 12 males and 22 females, 14 Whites, 6 African Americans, 11 Asians, and 3 Mixed Race (see Table S1). The median age was 20 (range 18–53, first quartile of 20, third quartile of 21). The median BMI was 23.5 (range 18–47). In total, 32 participants were able to complete all stages of the study. One participant did not complete the cognitive exam and one did not complete hydration measurements. These two participants were not included in the cognitive or hydration analysis. Outlier tests indicated an additional three participants whose values on the cognitive test and/or the hydration device were outside of plausible/acceptable ranges, suggesting a failure of the test procedure. Thus, the final sample for analysis included 29 participants. The calibration time was not recorded, but it typically took 2–4 min. Participants were instructed to avoid sudden or erratic movements during this period (Figures A2-A4). Further details of the device calibration are provided in Appendix A. A statistically significant moderate correlation between cognitive ability and hydration level of −0.431 (p = 0.02; p < 0.05) was found (see Table S2). In total, 64.7% of our sample had an interest in monitoring their hydration status (very or somewhat) (see Figure S1). More than 47.1% of our sample reported dehydration concerns (very or somewhat) (see Figure S1). In total, 70.5% of our participants felt the device would help them live healthy (very or somewhat) (see Figure S2). An R2 value of 0.186 indicates that 18.6% of the variance in the dependent variable is explained by the independent variable(s) (see Figure 2). While this pilot study demonstrates promising results of the DRINK band, there are several limitations. First, our study uses a small convenience sample from a single region and our age distribution is unbalanced. The association between ECW/BW and Stroop Exam performance, while significant in bivariate correlation using bootstrap correction, is preliminary and future work is needed to confirm this result. The DRINK band utilizes a segmental approach (upper body only) to measurement of hydration and caution is required when interpreting results, especially in patients with fluid retention and heart/lung diseases. This approach might not be as effective in a sample of older participants with more variation in body mass index. However, ECW fraction in our study (hydration status) is sufficiently similar to values reported in other studies [9]. Our study also did not include potential covariates, such as exercise, nutrition, environmental factors, or alcohol/substance abuse. Persons who exercise more likely have lean muscle mass, and muscle tissue generally contains more water than fat tissue (see Tables S3 and S4). There is some potential that exercise and unreported mental health/neurological disorders, which influences cognitive performance has an unobserved confounding influence. Other unobserved covariates could also have important relationships with the key variables in our pilot study. A randomized control trial with robust measurement of covariates and hypothesis testing of water restriction, physical activity and heat exposure (in combination with the wrist-worn hydration sensor and cognitive measurements) would clarify the preliminary finding in our pilot work. Our study identified a moderate correlation between longer Stroop exam performance and lower ECW fraction. The lower the ECW/BW, the more time participants required. The majority of participants showed interest in future use and reported a positive experience. Use of a wrist-worn hydration monitoring device has potential for health measurement and across a variety of settings. The device did encounter issues with calibration time for participants to reach its working levels. As a pilot study, testing on a larger sample of community members with a wider age range would be useful in the future to determine other usage concerns. Isaac Opoku: writing – original draft, investigation, funding acquisition, conceptualization, methodology, data curation. Eamon Johnson: supervision, investigation. Shravan Khare: investigation, software, writing – review and editing. Adam T. Perzynski: methodology, conceptualization, investigation, writing – review and editing. Mary Joan Roach: conceptualization, writing – review and editing, methodology, supervision, formal analysis. This study was supported by the Support of Undergraduate Research and Creative Endeavors – Arts, Humanities, and Social Sciences Grant (SOURCE AHSS) at Case Western Reserve University and the Inamori International Center for Ethics and Excellence. Mary Joan Roach, Adam T. Perzynski, Shravan Khare, and Eamon Johnson were supported by an NIH/NIA Grant Number AG080886. The sources had no role in study design, collection, analysis, interpretation of data, writing of the report, and the decision to submit the report for publication. All ethical guidelines were followed throughout the research process. The study protocol and instruments were reviewed and approved by the MetroHealth Institutional Review Board. Prior to the study, all participants completed informed consent. Dr. Perzynski reports conflicts of interest outside the submitted work. He is cofounder of Global Health Metrics LLC and reports book royalty income from Taylor Francis and Springer Nature. Mr. Johnson is also a cofounder of Global Health Metrics. The lead author Isaac Opoku affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained. D.R.I.N.K. (Device That Recognizes the Need to Intake Water) Band Conception and Device Goals TritonX created a wearable sensor (Figure 1) based on the BIS technique to accurately measure dynamic fluid changes in the human body. Initial validation was conducted in healthy adults and other ongoing work has been funded by the National Institute on Aging (1R42AG080886-01) and Case Western Reserve University. The device measures changes in the total body water (TBW), intracellular water (ICW), and extracellular water (ECW) in real time and can alert care personnel if there are clinically meaningful changes in fluid volume. Device Overview The D.R.I.N.K. device is required to be worn on one of the wrists of a patient, and it consists of a processor, sensor, electrodes, battery, Bluetooth module, vibration buzzer, and LED indicators. An example of the electrode placement is shown in Figure 1. Results of a reading are available to the user immediately on a connected smartphone app. The D.R.I.N.K. device is designed to detect hydration using a very small AC current through the human body, far below as allowed by the ISO60601 standard. Hand-to-Hand (Segmental Active or “Finger Sense”) Mode: In this active measurement method, the user/patient wears the D.R.I.N.K. device on one wrist and touches the electrodes # 3 and 4 from the middle and index finger of the other hand to measure hydration. It takes < 5 s to measure user hydration via this method. In this method, the band uses electrodes # 1, 2, 3, and 4 for the measurement. In the D.R.I.N.K. device, the following parameters are measured for the user: TBW (in liters), extracellular fluid (liters), intracellular fluid (liters), fat-free mass (kg), and fat mass (kg). Other indicators can also be produced, including ECW/ICW ratio, Rinf, Rzero, Re, Ri, Ptbf, and Error Sum value for each measurement. Rinf: Resistance at infinite frequency. Rzero: Resistance at zero frequency. Re (Extracellular Resistance): Resistance of the extracellular fluid compartment. Ri (Intracellular Resistance): Resistance of the intracellular fluid compartment. Ptbf (Phase Angle): This represents the phase angle between the current and voltage. TBW (Total Body Water): The total amount of water in the body, both intracellular and extracellular. ICW (Intracellular Water): The amount of water inside the cells. ECW (Extracellular Water): The amount of water outside the cells, in the interstitial spaces, and blood plasma. Fat % (Body Fat Percentage): The percentage of body weight that is composed of fat. Error Sum: This refers to the sum of errors in the measurements, indicating the accuracy or precision of the BIS data. The device calculates body mass index, or BMI, based on self-reported height and weight values upon sign in to the associated smartphone application. Age of the user, by date of birth, is also self-reported. A table representing hypothetical results from the device is generated below. Measurements Outside of Feasible Range Measurements are generated on device and post-processing can (optionally) occur on a connected smartphone application. Embedded in the associated smartphone application are discard features for discrepancies outside of acceptable ranges. These discrepancies, based on statistical modeling for participant characteristics, could only be due to faulty data collection (i.e., movement/damage to electrodes, unsatisfactory finger placement, and unknown influences) that generates results outside of any healthy or unhealthy participant. The most straightforward approach to error calculation and discard is to estimate the model residual variance (the sum of errors). When that sum is several multiples of the upper limit of normal, data are discarded. Calibration For Optimal Readings In calibration, the bioimpedance measurements are sent in real time from the device to the mobile phone. Readings on the smartphone at the outset of a measurement session can be highly variable. Reasons for this are not well understood, but, theoretically, include: movements made by the user, relaxation of the muscles involved in taking the measurement, accumulation of tiny amounts of perspiration between the skin and electrodes, contact resistance due to device/electrode repositioning, and other unknown influences. Users were instructed by study coordinators to hold still during calibration time and refrain from sudden movements. These findings were based on the previous application of BIS techniques/device and during its adaption into a wearable wrist-worn device. In BIS devices, critical parameters for fine-tuning accuracy include electrode materials (minimization of contact resistance), electrode spacing, electrode size, and algorithms and constants for converting raw bioimpedance measures, handling capacitive leakage (“hook effect”) and determining user-interpretable body composition values. In a series of completed rapid prototype studies, we have recognized and adapted to these challenges, leading to accuracy for a small, inexpensive, and wearable device. In addition, we have implemented a proprietary machine learning approach that has further improved accuracy across persons of different age, sex, and body type/volume. The D.R.I.N.K. device includes a novel feature, which we have called “finger sense.” In this approach, the device's intelligence examines selected bioimpedance readings in real-time and does not begin recording a measurement session until reaching a stable and optimal range. We recommend an ideal body position for taking readings, and in combination with the “finger sense” approach minimizes concerns about user error (e.g., excess motion, short circuiting, misplacement of fingers). The data that support the findings of this study are not publicly available currently. For further information or requests to access the data, please contact the corresponding author, Isaac Opoku. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.