1. Armstrong, Lawrence E. PhD


Human fluid-electrolyte regulation involves multiple neuroendocrine responses to the hourly loss and gain of body water. Such dynamic complexity complicates hydration assessment and minimizes the likelihood that any single biomarker will validly and precisely describe hydration status in all life situations. This article describes the biomarkers that are currently used during daily living to assess mild dehydration, plus recent advances in our understanding of invasive and noninvasive techniques. This article also suggests directions for future exploration of novel hydration indices, in the belief that superior biomarkers exist but have not been discovered.


Article Content

Among its many functions, water supports cellular metabolism, transports nutrients and waste, and serves as an integral component of thermoregulation. Body water exists as a dynamic and complex fluid matrix, containing interconnected compartments that are perturbed by numerous life stressors, fluid losses, and food intake.1 These characteristics of total body water make precise evaluation of hydration status difficult across a spectrum of life activities and minimize the likelihood that any single biomarker will validly and accurately describe hydration status in all life situations. Indeed, hydration assessment techniques are best viewed as singular, momentary measures of a constantly changing system.


Humans often experience mild dehydration during daily activities because of inadequate water intake.2,3 This propensity to dehydrate during daily activities occurs, in part, because humans do not sense thirst until they have lost 1% to 2% of body mass. Our understanding of such mild dehydration recently changed, because of research4 showing that a 1.6% body mass loss, considered inconsequential by most adults, impaired cognitive performance and mood in healthy young men (see below). Therefore, this brief article (a) describes recent advances in our understanding of invasive and noninvasive hydration assessment techniques and (b) presents recommendations for future investigations of novel hydration biomarkers, which may be used during daily activities to assess mild dehydration (ie, near the threshold of thirst sensation, <2% body mass loss).



For the average person, daily assessment of personal hydration status requires a method that involves little technical expertise and no sophisticated instrumentation. This method also should be accurate, safe, and inexpensive. A previous review of hydration assessment1 suggests that the following techniques meet these criteria: urine color, urine specific gravity, 24-hour urine volume, and body mass change. The reader should note that all urinary biomarkers of hydration state may be influenced by dietary contents, nutritional supplements, diseases, renal dysfunction, and bacterial contamination of unpreserved samples.


Measurement of body mass change, which is often used to represent water loss, is usually considered to be the most reliable short-term (ie, within a few hours) form of hydration assessment. However, Maughan et al5 described several sources of error that give rise to misleading results. For example, respiratory water loss, substrate oxidation, ingestion of solid food or water, and fluid excretion complicate the interpretation of measurements. Thus, carefully controlled measurements of body mass are necessary to reduce these potential sources of error. Furthermore, body mass changes alone cannot discern from which body fluid compartments water has been lost or gained.6


Plasma osmolality has been declared the best hydration marker by some authors,6,7 but not all evidence supports this contention. For example, one recent publication7 presented a figure showing that approximately one-fourth of test subjects (exact number not provided) exhibited increased plasma osmolality, 90 minutes after ingesting 500 mL of water. The authors did not comment on this unusual finding, although it is counterintuitive and not explained by known physiological principles or previously published data. Other shortcomings in the use of plasma osmolality as a hydration marker have been described,1 supporting the concept that no single marker is adequate in all situations.


In fact, the dynamic complexity of hypohydration demonstrates that no single measure (eg, plasma osmolality, urine specific gravity, body mass change) is superior or inferior to another.1 None should be used indiscriminately to make clinical decisions regarding hydration status, by using a single number.6 An average healthy adult, who desires to precisely evaluate hydration status during daily activities, should compare information from 2 or more hydration indices (ie, because single measurements may reflect recent fluid consumption or other acute perturbation of body water), more than once each day.1



In 2005 and 2006, ultrasound was evaluated as a means to assess hydration status. The authors reported8,9 that ultrasound velocity through soft tissue was linearly related to tissue water content. A prototype hydration monitor that measured ultrasound velocity was tested in animals, adult patients with lower limb edemas, and athletes during acute dehydration and rehydration. However, similar to bioimpedance spectroscopy,10 ultrasound assesses the physical properties of water and does not directly measure water volume or changes in the neuroendocrine regulation of total body water.


During this same period, studies evaluated body fluids other than blood and urine. For example, human salivary osmolality and flow rate were measured by Walsh and colleagues11 in 2004, during progressive dehydration. Osmolality increased while flow rate decreased, as dehydration progressed to a body mass loss of 3%. The authors concluded that salivary osmolality was as sensitive as urine osmolality, for tracking hydration changes during hypertonic hypovolemia. In 2010, salivary osmolality was included in an evaluation of 5 hydration biomarkers.12 The diagnostic accuracy of salivary osmolality was reported to be significantly better than chance, and its intraindividual coefficient of variation was 9.5%, considerably less than urine osmolality (28.3%) and urine color (30.9%). These findings suggest that further studies of salivary osmolality as a hydration marker are warranted.


Similarly, human intraocular pressure was measured during progressive dehydration. The results of a single study13 suggested that intraocular pressure was not associated with other markers of hydration status and that it rose in parallel with systemic blood pressure during exercise. Given that few studies exist in the scientific literature, further research confirmation is required to support the potential of intraocular pressure as a hydration index.


A 2010 investigation, conducted in our laboratory, utilized a different statistical approach to hydration assessment. This publication14 reported measurements of 9 hydration indices in 59 healthy young men, at 5 time points during a 12-day period. As shown in the Table, we utilized percentiles to define euhydration plus 6 other hydration categories, ranging from extremely hyperhydrated to extremely dehydrated. As such, these categories provide reference values that anyone can use to assess hydration state. To our knowledge, this was the first publication to delineate statistically derived categories of hydration state on the basis of measured values. At present, a companion paper involving women has been accepted for publication and is in press (L. E. Armstrong, et al. Hydration biomarkers and dietary fluid consumption of women; J Acad Nutr Diet. 2012; in press).

TABLE Categories of ... - Click to enlarge in new windowTABLE Categories of Hydration Status for 75.1-kg Males

Cheuvront and colleagues12 recently utilized 2 distinct statistical approaches to hydration assessment, evaluating body mass loss and widely used biomarkers in blood and urine. Their 2010 publication analyzed change values for body mass, urine color, and urine specific gravity, as well as the osmolalities of urine, saliva, and blood. For all of these variables, the authors computed a "decision level" by adding the reference change value (ie, the magnitude of dynamic change that makes a difference statistically significant) to the euhydration grand mean of all values. Each decision level defined the presence of dehydration (range, 1.8%-7.0% body mass loss) on the basis of variance, across 4 probability levels (80%, 85%, 90%, or 95%). Their 2011 paper utilized the same database6 but evaluated only plasma osmolality, urine specific gravity, and body mass. The authors calculated levels of confidence (eg, likely, 0.80%; more likely, 0.90%; very likely, 0.95%; virtually certain, 0.99%) that dehydration had occurred and formed a scale of clinical likelihood for monitoring dehydration. Both of these publications6,12 involve multiple a priori assumptions and utilize varied statistical approaches, which may or may not be appropriate for all environments, populations, timelines, and life activities.



The following 7 recommendations describe approaches that may lead to future discoveries of novel techniques to assess hydration state.


1. Evaluate numerous concurrent hydration indices, within the context of an experimental intervention or a group comparison. An example exists in the unpublished observations of French investigators (Perrier E, Vergne S, Klein A, Poupin M, Rondeau P, Le Bellego L, Armstrong LE, Lang F, Stookey J, Tack I, unpublished observations, 2011) .This team of scientists concurrently evaluated 77 variables. Furthermore, their investigation compared 2 groups of adults: high-volume drinkers and low-volume drinkers. This group comparison technique offers great promise for future research that evaluates subtle differences due to daily fluid intake, during typical daily activities. It also minimizes investigator intervention.


2. Describe hydration categories on the basis of statistical analyses, to make abstract terms concrete. For example, numerical values that describe the terms euhydration, extremely hyperhydrated, and extremely dehydrated provide guidance (Table) that can be applied during daily activities, exercise, or labor.


3. Conduct studies of renal function because the kidneys are critical in the regulation of total body water. Measure hormones (eg, arginine vasopressin, aldosterone) that regulate fluid and electrolyte balance.


4. Evaluate the relationship between thirst and body mass loss because few published studies exist. The scientific literature often states that thirst "lags behind dehydration," meaning that thirst is not sensed until a body mass loss of 1% to 2% occurs. However, thirst is a complex sensation that is affected by several factors, including oropharyngeal dryness, stomach distention, gastric emptying, plasma osmolality, and fluid composition. Furthermore, drinking behavior (ie, volume consumed per hour) is influenced by environmental temperature, internal body temperature, fluid palatability, and fluid temperature. Much remains to be discovered about thirst.


5. Investigate the validity and accuracy of mood ratings and cognitive performance as behavioral indices of hydration status. Recently published evidence from our laboratory4 demonstrates that mild dehydration (1.6% body mass loss) without hyperthermia induced significant adverse changes in cognitive performance (eg, visual vigilance, working memory) and mood (fatigue and tension/anxiety) of healthy young men. This suggests that, with further study, behavioral measures may someday be used to assess, herald, or confirm small losses of body water.


6. Expand investigations of dietary water consumption. This should include carefully monitored measurements of water, beverages, and water in solid foods of different groups in various countries. The scientific literature contains few studies upon which to base recommendations for fluid intake.


7. Determine if the hydration assessment techniques that are used at present for adult men and women also apply to children, senior citizens, pregnant women, and lactating women. Obvious physiological differences among these groups suggest that widely used hydration indices may be valid in some but not all individuals.



These recommendations assume that excellent, perhaps superior biomarkers exist but have not been discovered to date.




1. Armstrong LE. Assessing hydration status: the elusive gold standard. J Am Coll Nutr. 2007; 26: 575S-584S. [Context Link]


2. Adolph EF. Voluntary dehydration. In: Adolph EF, ed. Physiology of Man in the Desert. New York: Interscience Publishers; 1947: 254-270. [Context Link]


3. Greenleaf J. Problem: thirst, drinking behavior, and involuntary dehydration. Med Sci Sports Exerc. 1992; 24: 645-656. [Context Link]


4. Ganio MS, Armstrong LE, Casa DJ, et al.. Mild dehydration impairs cognitive performance and mood of men [published online ahead of print June 7, 2011]. Br J Nutr. 2011; 7: 1-9. [Context Link]


5. Maughan RH, Shirreffs SM, Leiper JB. Errors in the estimation of hydration status from changes in body mass. J Sports Sci. 2007; 25: 797-804. [Context Link]


6. Cheuvront SN, Fraser CG, Kenefick RW, Ely BR, Sawka MN. Reference change values for monitoring dehydration. Clin Chem Lab Med. 2011; 49: 1033-1037. [Context Link]


7. Sollanek KJ, Kenefick RW Cheuvront SN, Axtell RS. Potential impact of a 500-mL water bolus and body mass on plasma osmolality dilution. Eur J Appl Physiol. 2011; 111: 1999-2004. [Context Link]


8. Sarvazyan AP, Tatarinov A, Sarvazyan N. Ultrasonic assessment of tissue hydration status. Ultrasonics. 2005; 43: 661-671. [Context Link]


9. Topchyan A, Tatarinov A, Sarvazyan N, Sarvazyan AP. Ultrasound velocity in human muscle in vivo: perspective for edema studies. Ultrasonics. 2006; 44: 259-264. [Context Link]


10. Armstrong LE, Kenefick RW, Castellani JW, et al.. Bioimpedance spectroscopy technique: intra-, extracellular, and total body water. Med Sci Sports Exerc. 1997; 29: 1657-1663. [Context Link]


11. Walsh NP, Montague JC, Callow N, Rowlands AV. Saliva flow rate, total protein concentration and osmolality as potential markers of whole body hydration status during progressive acute dehydration in humans. Arch Oral Biol. 2004; 49: 149-154. [Context Link]


12. Cheuvront SN, Ely BR, Kenefick RW, Sawka MN. Biological variation and diagnostic accuracy of dehydration assessment markers. Am J Clin Nutr. 2010; 92: 565-573. [Context Link]


13. Bakke EF, Hisday J, Semb SO. Intraocular pressure increases in parallel with systemic blood pressure during isometric exercise. Invest Ophthalmol Vis Sci. 2009; 50: 760-764. [Context Link]


14. Armstrong LE, Pumerantz AC, Fiala KA, et al.. Human hydration indices: acute and longitudinal reference values. Int J Sport Nutr Exerc Metab. 2010; 20: 145-153. [Context Link]