Water
SUMMARY
Water is the largest single constituent of the human body and is essential for cellular homeostasis and life. Total water intake includes drinking water, water in beverages, and water that is part of food. Although a low intake of total water has been associated with some chronic diseases, this evidence is insufficient to establish water intake recommendations as a means to reduce the risk of chronic diseases. Instead, an Adequate Intake (AI) for total water is set to prevent deleterious, primarily acute, effects of dehydration, which include metabolic and functional abnormalities.
The primary indicator of hydration status is plasma or serum osmolality. Because normal hydration can be maintained over a wide range of water intakes, the AI for total water (from a combination of drinking water, beverages, and food) is set based on the median total water intake from U.S. survey data. The AI for total water intake for young men and women (ages 19 to 30 years) is 3.7 L and 2.7 L per day, respectively. Fluids (drinking water and beverages) provided 3.0 L (101 fluid oz; ≈ 13 cups) and 2.2 L (74 fluid oz; ≈ 9 cups) per day for 19- to 30-year-old men and women, respectively, representing approximately 81 percent of total water intake in the U.S. survey.
[Conversion factors: 1 L = 33.8 fluid oz; 1 L = 1.06 qt; 1 cup = 8 fluid oz.]
Water contained in food provided approximately 19 percent of total water intake. Canadian survey data indicated somewhat lower levels of total water intake. As with AIs for other nutrients, for a healthy person, daily consumption below the AI may not confer additional risk because a wide range of intakes is compatible with normal hydration. In this setting, the AI should not be interpreted as a specific requirement. Higher intakes of total water will be required for those who are physically active or who are exposed to hot environments.
Over the course of a few hours, body water deficits can occur due to reduced intake or increased water losses from physical activity and environmental (e.g., heat) exposure. However, on a day-to-day basis, fluid intake, driven by the combination of thirst and the consumption of beverages at meals, allows maintenance of hydration status and total body water at normal levels.
Because healthy individuals have considerable ability to excrete excess water and thereby maintain water balance, a Tolerable Upper Intake Level (UL) was not set for water. However, acute water toxicity has been reported due to rapid consumption of large quantities of fluids that greatly exceeded the kidney’s maximal excretion rate of approximately 0.7 to 1.0 L/hour.
BACKGROUND INFORMATION
Water, which is the solvent for biochemical reactions, has unique physical properties (e.g., high specific heat) to absorb metabolic heat within the body. Water is also essential for maintaining vascular volume and serves as the medium for transport within the body by supplying nutrients and removing waste. In addition, cell hydration has been has been suggested to be an important signal to regulate cell metabolism and gene expression (Haussinger et al., 1994). Daily water intake must be balanced with losses in order to maintain total body water. Body water deficits challenge the ability to maintain homeostasis during perturbations (e.g., sickness, physical exercise, and environmental exposure) and can affect function and health. In very unusual circumstances, excess consumption of hypotonic fluids and low sodium intake may lead to excess body water, resulting in hyponatremia and cellular edema.
Despite the importance of adequate water intake, there is confusion among the general public and health care providers on the amount of water that should be consumed (Valtin, 2002), in part because of misinterpretation of previous recommendations (NRC, 1989).
BODY WATER
Fat-Free Mass
Body water volume, as a percentage of fat-free mass, is highest in infants and declines in older children (Fomon, 1967; Van Loan and Boileau, 1996). High body water volume is particularly evident in newborns, whose body water content of fat-free mass may exceed 75 percent (Fomon, 1967). Infants also have a relatively higher water content in the extracellular compartment and a lower water content in the intracellular compartment compared with older children (Van Loan and Boileau, 1996). Figure 4-1 presents total body water as a percentage of fat-free mass and body mass in children through the teenage years. Total body water as percentage of fat-free mass decreases during childhood, albeit more slowly than in infancy.
For adults, fat-free mass is approximately 70 to 75 percent water, and adipose tissue is approximately 10 to 40 percent water. With increasing fatness, the water fraction of adipose tissue decreases (Martin et al., 1994).
FIGURE 4-1 Total body water as a fraction of body mass (FW) and as a fraction of fat-free mass (FWFFM). Reprinted with permission, from Van Loan and Boileau (1996). Copyright 1996 by CRC Press.
FIGURE 4-2 Hydration of fat-free mass in relation to age for 95 African-American (closed circles) and 204 white (open circles) men. Reprinted with permission, from Visser and Gallagher (1998). Copyright 1998 by John Libbey Eurotext.
FIGURE 4-3 Hydration of fat-free mass in relation to age for 99 African-American (closed circles) and 270 white (open circles) women. Reprinted with permission, from Visser and Gallagher (1998). Copyright 1998 by John Libbey Eurotext.
Figures 4-2 and 4-3 provide the percentage of water (tritiated water) in fat-free mass measured by dual energy X-ray absorptiometry (DEXA) in relation to age for men and women, respectively (Visser and Gallagher, 1998; Visser et al., 1997). Note that individual variation exists for the hydration of fat-free tissue and values remain relatively stable with increasing age. Neither ethnicity nor gender altered the hydration of fat-free mass. Similar values were reported for whites (men = 74 percent, women = 74 percent) and African Americans (men = 75 percent, women = 75 percent). Other investigators have supported the observation that age and gender do not markedly alter the hydration of fat-free mass in adults (Baumgartner et al., 1995; Goran et al., 1994; Mazariegos et al., 1994).
Total Body Water
Total body water (TBW), comprising extracellular fluid (ECF) and intracellular fluid (ICF), averages approximately 60 percent of body weight, with a range from approximately 45 to 75 percent (Altman, 1961). Variability in TBW is primarily due to differences in body composition. TBW is usually measured by volume distribution of an appropriate indicator (e.g., antipyrine, deuterium oxide, tritium oxide). Table 4-1 provides TBW values for different age and gender groups based upon indicator dilution methods (Altman, 1961). Women and older persons have reduced TBW primarily because of having lower fat-free mass and increased body fat.
TABLE 4-1 Total Body Water (TBW) as a Percentage of Total Body Weight in Various Age and Gender Groups
Lifestage
TBW as a Percentage of Body Weight, Mean (range)
0–6 mo
74 (64–84)
6 mo–1 yr
60 (57–64)
1–12 yr
60 (49–75)
Males, 12–18 yr
59 (52–66)
Females, 12–18 yr
56 (49–63)
Males, 19–50 yr
59 (43–73)
Females, 19–50 yr
50 (41–60)
Males, 51+ yr
56 (47–67)
Females, 51+ yr
47 (39–57)
SOURCE: Altman (1961).
Gender differences in TBW are not observed until after approximately 12 years of age (Novak, 1989), when boys start increasing their fat-free mass at a rate faster than girls do.
Athletes have relatively high TBW values by virtue of having a high fat-free mass, low body fat, and high skeletal muscle glycogen levels. High skeletal muscle glycogen levels increase the water content of fat-free tissue due to osmotic pressure exerted by glycogen granules within the muscle sarcoplasm (Neufer et al., 1991; Olsson and Saltin, 1970).
Distribution
Body water is distributed between the ICF and the ECF, which contain 65 and 35 percent of TBW, respectively. The ECF is further divided into the interstitial and plasma spaces. An average 70-kg man has approximately 42 L of total body water, 28 L of ICF, and 14 L of ECF, with the ECF comprising approximately 3 L of plasma and 11 L of interstitial fluid. These are not static volumes, but represent the net effects of dynamic fluid exchange with varying turnover rates between compartments (Guyton and Hall, 2000). Perturbations such as exercise, heat exposure, fever, diarrhea, trauma, and skin burns will greatly modify the net volumes and water turnover rates between these fluid compartments.
Exchange
Water exchange between the ICF and ECF depends on osmotic gradients. Water passes through membranes from regions of lower to higher solute concentration by osmosis, which attempts to equalize the concentration differences across the membrane. Cell membranes are freely permeable to water, but they are only selectively permeable to solutes. Water thus distributes across cell membranes to equalize the osmotic concentrations of extracellular and intracellular fluids. Although the two compartments contain different individual solute concentrations, the total equilibrium concentration of cations and anions is the same in each compartment as described by the Gibbs-Donnan equilibrium. In the ECF, the most abundant cation is sodium, while chloride and bicarbonate are the primary anions. These ions represent 90 to 95 percent of the osmotically active components of the ECF, and changes in their content alter the ECF volume. In the ICF, the most abundant cations are potassium and magnesium, while proteins are the primary anions. The marked differences in sodium and potassium concentrations between ICF and ECF are maintained by active transport-mediated ion pumps within cell membranes.
Water exchange between the intravascular and interstitial spaces occurs in the capillaries. Capillaries of different tissues have varied anatomic structures and therefore different permeability to water and solutes. The transcapillary forces that determine if net filtration (i.e., water leaving the vascular space) or net absorption (i.e., water entering the vascular space) will occur are hydrostatic and oncotic pressures. Oncotic pressure is the osmotic pressure attributed to serum protein concentration (e.g., serum albumin levels) differences across the capillary membrane. Generally, filtration occurs at the arterial end of the capillary, while absorption occurs at the venous end.
Incomplete fluid replacement resulting in decreased total body water affects each fluid space as a consequence of free fluid exchange (Costill and Fink, 1974; Durkot et al., 1986; Nose et al., 1983). The distribution of body water loss among the fluid spaces, as well as among different body organs during water deficit (dehydration or hypohydration), was determined in an animal model (Nose et al., 1983). The fluid deficit in rats thermally dehydrated by 10 percent of body weight was apportioned between the intracellular (41 percent) and extracellular (59 percent) spaces. Organ fluid loss was 40 percent coming from muscle, 30 percent from skin, 14 percent from viscera, and 14 percent from bone. Neither the brain nor liver lost significant water content. Various dehydration methods influence the partitioning of water loss from the fluid spaces (Mack and Nadel, 1996).
Determinants of Body Water Balance
Body water balance depends on the net difference between water gain and water loss. Water gain occurs from consumption (liquids and food) and production (metabolic water), while water losses occur from respiratory, skin, renal, and gastrointestinal tract losses. Water is normally consumed by mouth via liquid and food, and this mixture is digested and absorbed within the gastrointestinal tract. Therefore, water intake can be estimated from measured liquid volumes and tables of food composition. Water losses can be estimated from a variety of physiological and biophysical measurements and calculations (Adolph, 1933; Consolazio et al., 1963; Johnson, 1964). Depending upon a person’s age, health, diet, activity level, and environmental exposure, different physiological and biophysical methods can be used to quantify the water balance components.
TABLE 4-2 Estimation of Minimum Daily Water Losses and Production
Reference Source Loss (mL/d) Production (mL/d)
Hoyt and Honig, 1996 Respiratory loss −250 to −350
Adolph, 1947b Urinary loss −500 to −1,000
Newburgh et al., 1930 Fecal loss −100 to −200
Kuno, 1956 Insensible loss −450 to −1,900
Hoyt and Honig, 1996 Metabolic production +250 to +350
Total −1,300 to −3,450 +250 to +350
Net loss −1,050 to −3,100
Assuming conditions in which there is minimal water loss from sweating.
Table 4-2 displays estimated minimum losses and production of water (mL/day) in healthy sedentary adults, assuming conditions in which there is minimal water loss from thermoregulatory sweating. The following sections describe each source of water loss or production listed in this table.
Respiratory Water Loss
The amount of respiratory water loss, via evaporation within the lungs, is dependent on both the ventilatory volume and water vapor pressure gradient (Mitchell et al., 1972). Ventilatory volume is increased by physical activity, hypoxia, and hypercapnia, whereas the water vapor pressure is modified by the ambient temperature, humidity, and barometric pressure. Physical activity generally has a greater effect on respiratory water loss than do environmental factors. Daily respiratory water loss averages about 250 to 350 mL/day for sedentary persons, but can increase to 500 to 600 mL/day for active persons living in temperate2 climates at sea level (Hoyt and Honig, 1996). For these conditions, respiratory water loss (y = mL/day) can be predicted from metabolic rate (x = kcal/day) by the equation y = 0.107x + 92.2 (Hoyt and Honig, 1996). High altitude exposure (greater than 4,300 m, 448 mm Hg) can further increase respiratory water losses by approximately 200 mL/day (Hoyt and Honig, 1996).
In general, dry bulb temperatures of approximately 70°F, 80°F, and 90°F are used for temperate, warm, and hot conditions, respectively, in this report.
Ambient air temperature and humidity modify respiratory water losses. Breathing hot, dry air during intense physical exercise can increase respiratory water losses by 120 to 300 mL/day (Mitchell et al., 1972). Breathing cold, dry air during rest and stressful physical exercise (Table 4-3) can increase respiratory water losses by approximately 5 mL/hour and approximately 15 to 45 mL/hour, respectively (Freund and Young, 1996). Freund and Young (1996) have calculated that for a 24-hour military scenario (8 hours of rest, 12 hours of moderate activity, and 4 hours of moderate-heavy activity), the respiratory water losses increase by approximately 340 mL/day when breathing −20°C versus +25°C air.
Urinary and Gastrointestinal Water Loss
The kidneys are responsible for regulating the volume and composition of the ECF via a series of intricate neuroendocrine pathways (Andreoli et al., 2000). Renal fluid output can vary depending upon the specific macronutrient, salt, and water load. However, for persons consuming an average North American diet, some of these effects may not be discernable (Luft et al., 1983). Since there is a limit to how much the kidneys can concentrate urine, the minimal amount of water needed is determined by the quantity of end products that need to be excreted (e.g., creatinine, urea). On typical Western diets, an average of 650 mOsmol of electrolytes and other
TABLE 4-3 Influence of Breathing Cold Air and of Metabolic Rate on Respiratory Water Losses.
Temperature °F /°C; Relative Humidity (%); Water Vapor Pressure (mm Hg); Metabolic Rate (Watts); Respiratory Water Loss (mL/h)
77/25; 65;15; Rest (100); ≈ 10
32/0; 100; 5; Rest (100); ≈ 13
−4/−20;100; 1; Rest (100); ≈ 15
77/25; 65; 15; Light-moderate (300); ≈ 30
32/0; 100; 5; Light-moderate (300); ≈ 40
−4/−20; 100; 1; Light-moderate (300); ≈ 45
77/25; 65; 15; Moderate-heavy (600); ≈ 60
32/0; 100; 5; Moderate-heavy (600); ≈ 80
−4/−20; 100; 1; Moderate-heavy (600); ≈ 90
SOURCE: Reprinted with permission, from Freund and Young (1996). Copyright 1996 by CRC Press.
On typical Western diets, an average of 650 mOsmol of electrolytes and other solutes must be excreted per day to maintain electrolyte balance; thus, if the urine is maximally concentrated (Uosm approximately 1,200 mOsmol/kg water), the minimum urine output is approximately 500 mL/day. For dehydrated subjects living in hot weather, minimum daily urine outputs can be less than 500 mL/day (Adolph, 1947b).
Urine output generally averages 1 to 2 L/day but can reach 20 L/day in those consuming large quantities of fluid (West, 1990). Healthy older individuals, however, cannot concentrate urine as well as young individuals and thus have a higher minimum urine output. For example, older men and women (mean age 79 years) had lower maximal urine osmolalities of 808 and 843 mOsm/kg, respectively, compared with 1,089 mOsm/kg for young men (mean age 24 years). This corresponds to higher minimum urine outputs of 700 and 1,086 mL/day for the older men and women compared with 392 mL/day for the young men (Dontas et al., 1972).
Urine output varies inversely with body hydration status.
FIGURE 4-4 Relation of urine output to body hydration status. Reprinted with permission, from Lee (1964). Copyright 1964 Handbook of Physiology, Section 4, American Physiological Society.
Figure 4-4 depicts the hyperbolic relationship between urine output and body hydration status: one asymptote ascends steeply with hyperhydration, while the other descends gradually with dehydration (Lee, 1964). The apex of this hyperbolic relationship approximates a urine output of approximately 50 mL/hour. The extremes depicted in Figure 4-4 can be exceeded. For example, investigators have reported that urine output can transiently increase to approximately 600 to 1,000 mL/hour with water loading (Freund et al., 1995; Noakes et al., 2001; Speedy et al., 2001) and decrease to approximately 15 mL/hour with dehydration (Adolph, 1947b). Urine output can vary widely to maintain total body water; however, there are clearly limits to the amount of conservation and excretion.
Physical activity and climate also affect urine output. Exercise and heat strain will reduce urine output by 20 to 60 percent (Convertino, 1991; Mittleman, 1996; Zambraski, 1996), while cold and hypoxia will increase urine output (Freund and Young, 1996; Hoyt and Honig, 1996).
Gastrointestinal and thus fecal water loss in healthy adults is approximately 100 to 200 mL/day (Newburgh et al., 1930).
Insensible and Sweat Losses
Water loss through the skin occurs via insensible diffusion and secreted sweat. For the average adult, loss of water by insensible diffusion is approximately 450 mL/day (Kuno, 1956). During heat stress, eccrine sweat glands secrete sweat onto the skin surface, which cools the body when water evaporates from the sweat. In hot weather, sweat evaporation provides the primary avenue of heat loss to defend the body’s core temperature. When a gram of sweat water is vaporized at 30°C, 2.43 kJ (0.58 kcal) of heat becomes kinetic energy (latent heat of evaporation) (Wenger, 1972). For a given hot weather condition, the required sweating rate for evaporative cooling is dependent upon the physical activity level (metabolic rate).
The following calculations provide the minimal sweat produced by persons performing moderately heavy (metabolic rate ≈ 600 W) exercise in the heat (Sawka et al., 1996a). If the activity is 20 percent efficient, the remaining 80 percent of metabolic energy produced is converted to heat in the body so that 480 W (0.48 kJ/second, or 28.8 kJ/minute or 6.88 kcal/minute) need to be dissipated to avoid heat storage. The specific heat of body tissue (amount of energy required for 1 kg of tissue to increase temperature by 1°C) approximates 3.5 kJ (0.84 kcal)/kg/°C. For example, a 70-kg man has a heat capacity of 245 kJ (59 kcal)/°C, and a 50-kg woman has a heat capacity of 173
Page 84
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
kJ (41 kcal)/°C. If these persons performed exercise in a hot environment that enabled only evaporative heat loss and they did not sweat, their body temperatures would increase by approximately 1.0°C every 8.5 min for the man (245 kJ/°C ÷ 28.8 kJ/minute or 59 kcal/°C ÷ 6.88 kcal/minute) and every 6 minutes for the woman (173 kJ/°C ÷ 28.8 kJ/minute or 41 kcal/°C ÷ 6.88 kcal/minute). Since the latent heat of evaporation is 2.43 kJ/g (0.58 kcal/g), such persons would need to evaporate approximately 12 g of sweat per minute (28.8 kJ/minute ÷ 2.43 kJ/g or 6.88 kcal/minute ÷ 0.58 kcal/ g) or 0.72 L/hour. Because secreted sweat drips from the body and is not evaporated, higher sweat secretions are often needed to achieve these cooling demands. If a person is physically active and exposed to environmental heat stress, sweat losses to avoid heat storage can be substantial over a 24-hour period.
For persons living in hot climates, daily sweat losses often exceed several liters. As described above, daily sweat losses are determined by the evaporative heat loss requirements, which are influenced by the metabolic rate (above example) and environment. The environmental factors that modify sweat losses include clothing worn, ambient temperature, humidity, air motion, and solar load. Therefore, considerable variability will exist for daily sweat losses among different people. Figure 4-5 provides the distribution of daily sweat-
FIGURE 4-5 Distribution of daily sweating rates for active soldiers in desert and tropical climates. Percent incidence refers to the percentage of the subject population achieving the given daily sweat loss.
SOURCE: Molnar (1947). Reprinted with permission from the Papers of Edward Adolph collection at the Edward G. Miner Library, University of Rochester Medical Center.
Page 85
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
ing rates for soldiers living in desert and tropical climates (without air conditioning). The average daily sweat loss for 97 men in the desert was 4.9 L; for 26 men in the tropics, it was 2.3 L. The lower daily sweat losses in the tropics were probably due to lower ambient temperatures and lower solar load (both acting to lower the required evaporative cooling), as the precise activity levels of both groups were unknown.
SUMMARY
Water is the largest single constituent of the human body and is essential for cellular homeostasis and life. Total water intake includes drinking water, water in beverages, and water that is part of food. Although a low intake of total water has been associated with some chronic diseases, this evidence is insufficient to establish water intake recommendations as a means to reduce the risk of chronic diseases. Instead, an Adequate Intake (AI) for total water is set to prevent deleterious, primarily acute, effects of dehydration, which include metabolic and functional abnormalities.
The primary indicator of hydration status is plasma or serum osmolality. Because normal hydration can be maintained over a wide range of water intakes, the AI for total water (from a combination of drinking water, beverages, and food) is set based on the median total water intake from U.S. survey data. The AI for total water intake for young men and women (ages 19 to 30 years) is 3.7 L and 2.7 L per day, respectively. Fluids (drinking water and beverages) provided 3.0 L (101 fluid oz; ≈ 13 cups) and 2.2 L (74 fluid oz; ≈ 9 cups) per day for 19- to 30-year-old men and women, respectively, representing approximately 81 percent of total water intake in the U.S. survey.
[Conversion factors: 1 L = 33.8 fluid oz; 1 L = 1.06 qt; 1 cup = 8 fluid oz.]
Water contained in food provided approximately 19 percent of total water intake. Canadian survey data indicated somewhat lower levels of total water intake. As with AIs for other nutrients, for a healthy person, daily consumption below the AI may not confer additional risk because a wide range of intakes is compatible with normal hydration. In this setting, the AI should not be interpreted as a specific requirement. Higher intakes of total water will be required for those who are physically active or who are exposed to hot environments.
Over the course of a few hours, body water deficits can occur due to reduced intake or increased water losses from physical activity and environmental (e.g., heat) exposure. However, on a day-to-day basis, fluid intake, driven by the combination of thirst and the consumption of beverages at meals, allows maintenance of hydration status and total body water at normal levels.
Because healthy individuals have considerable ability to excrete excess water and thereby maintain water balance, a Tolerable Upper Intake Level (UL) was not set for water. However, acute water toxicity has been reported due to rapid consumption of large quantities of fluids that greatly exceeded the kidney’s maximal excretion rate of approximately 0.7 to 1.0 L/hour.
BACKGROUND INFORMATION
Water, which is the solvent for biochemical reactions, has unique physical properties (e.g., high specific heat) to absorb metabolic heat within the body. Water is also essential for maintaining vascular volume and serves as the medium for transport within the body by supplying nutrients and removing waste. In addition, cell hydration has been has been suggested to be an important signal to regulate cell metabolism and gene expression (Haussinger et al., 1994). Daily water intake must be balanced with losses in order to maintain total body water. Body water deficits challenge the ability to maintain homeostasis during perturbations (e.g., sickness, physical exercise, and environmental exposure) and can affect function and health. In very unusual circumstances, excess consumption of hypotonic fluids and low sodium intake may lead to excess body water, resulting in hyponatremia and cellular edema.
Despite the importance of adequate water intake, there is confusion among the general public and health care providers on the amount of water that should be consumed (Valtin, 2002), in part because of misinterpretation of previous recommendations (NRC, 1989).
BODY WATER
Fat-Free Mass
Body water volume, as a percentage of fat-free mass, is highest in infants and declines in older children (Fomon, 1967; Van Loan and Boileau, 1996). High body water volume is particularly evident in newborns, whose body water content of fat-free mass may exceed 75 percent (Fomon, 1967). Infants also have a relatively higher water content in the extracellular compartment and a lower water content in the intracellular compartment compared with older children (Van Loan and Boileau, 1996). Figure 4-1 presents total body water as a percentage of fat-free mass and body mass in children through the teenage years. Total body water as percentage of fat-free mass decreases during childhood, albeit more slowly than in infancy.
For adults, fat-free mass is approximately 70 to 75 percent water, and adipose tissue is approximately 10 to 40 percent water. With increasing fatness, the water fraction of adipose tissue decreases (Martin et al., 1994).
FIGURE 4-1 Total body water as a fraction of body mass (FW) and as a fraction of fat-free mass (FWFFM). Reprinted with permission, from Van Loan and Boileau (1996). Copyright 1996 by CRC Press.
FIGURE 4-2 Hydration of fat-free mass in relation to age for 95 African-American (closed circles) and 204 white (open circles) men. Reprinted with permission, from Visser and Gallagher (1998). Copyright 1998 by John Libbey Eurotext.
FIGURE 4-3 Hydration of fat-free mass in relation to age for 99 African-American (closed circles) and 270 white (open circles) women. Reprinted with permission, from Visser and Gallagher (1998). Copyright 1998 by John Libbey Eurotext.
Figures 4-2 and 4-3 provide the percentage of water (tritiated water) in fat-free mass measured by dual energy X-ray absorptiometry (DEXA) in relation to age for men and women, respectively (Visser and Gallagher, 1998; Visser et al., 1997). Note that individual variation exists for the hydration of fat-free tissue and values remain relatively stable with increasing age. Neither ethnicity nor gender altered the hydration of fat-free mass. Similar values were reported for whites (men = 74 percent, women = 74 percent) and African Americans (men = 75 percent, women = 75 percent). Other investigators have supported the observation that age and gender do not markedly alter the hydration of fat-free mass in adults (Baumgartner et al., 1995; Goran et al., 1994; Mazariegos et al., 1994).
Total Body Water
Total body water (TBW), comprising extracellular fluid (ECF) and intracellular fluid (ICF), averages approximately 60 percent of body weight, with a range from approximately 45 to 75 percent (Altman, 1961). Variability in TBW is primarily due to differences in body composition. TBW is usually measured by volume distribution of an appropriate indicator (e.g., antipyrine, deuterium oxide, tritium oxide). Table 4-1 provides TBW values for different age and gender groups based upon indicator dilution methods (Altman, 1961). Women and older persons have reduced TBW primarily because of having lower fat-free mass and increased body fat.
TABLE 4-1 Total Body Water (TBW) as a Percentage of Total Body Weight in Various Age and Gender Groups
Lifestage
TBW as a Percentage of Body Weight, Mean (range)
0–6 mo
74 (64–84)
6 mo–1 yr
60 (57–64)
1–12 yr
60 (49–75)
Males, 12–18 yr
59 (52–66)
Females, 12–18 yr
56 (49–63)
Males, 19–50 yr
59 (43–73)
Females, 19–50 yr
50 (41–60)
Males, 51+ yr
56 (47–67)
Females, 51+ yr
47 (39–57)
SOURCE: Altman (1961).
Gender differences in TBW are not observed until after approximately 12 years of age (Novak, 1989), when boys start increasing their fat-free mass at a rate faster than girls do.
Athletes have relatively high TBW values by virtue of having a high fat-free mass, low body fat, and high skeletal muscle glycogen levels. High skeletal muscle glycogen levels increase the water content of fat-free tissue due to osmotic pressure exerted by glycogen granules within the muscle sarcoplasm (Neufer et al., 1991; Olsson and Saltin, 1970).
Distribution
Body water is distributed between the ICF and the ECF, which contain 65 and 35 percent of TBW, respectively. The ECF is further divided into the interstitial and plasma spaces. An average 70-kg man has approximately 42 L of total body water, 28 L of ICF, and 14 L of ECF, with the ECF comprising approximately 3 L of plasma and 11 L of interstitial fluid. These are not static volumes, but represent the net effects of dynamic fluid exchange with varying turnover rates between compartments (Guyton and Hall, 2000). Perturbations such as exercise, heat exposure, fever, diarrhea, trauma, and skin burns will greatly modify the net volumes and water turnover rates between these fluid compartments.
Exchange
Water exchange between the ICF and ECF depends on osmotic gradients. Water passes through membranes from regions of lower to higher solute concentration by osmosis, which attempts to equalize the concentration differences across the membrane. Cell membranes are freely permeable to water, but they are only selectively permeable to solutes. Water thus distributes across cell membranes to equalize the osmotic concentrations of extracellular and intracellular fluids. Although the two compartments contain different individual solute concentrations, the total equilibrium concentration of cations and anions is the same in each compartment as described by the Gibbs-Donnan equilibrium. In the ECF, the most abundant cation is sodium, while chloride and bicarbonate are the primary anions. These ions represent 90 to 95 percent of the osmotically active components of the ECF, and changes in their content alter the ECF volume. In the ICF, the most abundant cations are potassium and magnesium, while proteins are the primary anions. The marked differences in sodium and potassium concentrations between ICF and ECF are maintained by active transport-mediated ion pumps within cell membranes.
Water exchange between the intravascular and interstitial spaces occurs in the capillaries. Capillaries of different tissues have varied anatomic structures and therefore different permeability to water and solutes. The transcapillary forces that determine if net filtration (i.e., water leaving the vascular space) or net absorption (i.e., water entering the vascular space) will occur are hydrostatic and oncotic pressures. Oncotic pressure is the osmotic pressure attributed to serum protein concentration (e.g., serum albumin levels) differences across the capillary membrane. Generally, filtration occurs at the arterial end of the capillary, while absorption occurs at the venous end.
Incomplete fluid replacement resulting in decreased total body water affects each fluid space as a consequence of free fluid exchange (Costill and Fink, 1974; Durkot et al., 1986; Nose et al., 1983). The distribution of body water loss among the fluid spaces, as well as among different body organs during water deficit (dehydration or hypohydration), was determined in an animal model (Nose et al., 1983). The fluid deficit in rats thermally dehydrated by 10 percent of body weight was apportioned between the intracellular (41 percent) and extracellular (59 percent) spaces. Organ fluid loss was 40 percent coming from muscle, 30 percent from skin, 14 percent from viscera, and 14 percent from bone. Neither the brain nor liver lost significant water content. Various dehydration methods influence the partitioning of water loss from the fluid spaces (Mack and Nadel, 1996).
Determinants of Body Water Balance
Body water balance depends on the net difference between water gain and water loss. Water gain occurs from consumption (liquids and food) and production (metabolic water), while water losses occur from respiratory, skin, renal, and gastrointestinal tract losses. Water is normally consumed by mouth via liquid and food, and this mixture is digested and absorbed within the gastrointestinal tract. Therefore, water intake can be estimated from measured liquid volumes and tables of food composition. Water losses can be estimated from a variety of physiological and biophysical measurements and calculations (Adolph, 1933; Consolazio et al., 1963; Johnson, 1964). Depending upon a person’s age, health, diet, activity level, and environmental exposure, different physiological and biophysical methods can be used to quantify the water balance components.
TABLE 4-2 Estimation of Minimum Daily Water Losses and Production
Reference Source Loss (mL/d) Production (mL/d)
Hoyt and Honig, 1996 Respiratory loss −250 to −350
Adolph, 1947b Urinary loss −500 to −1,000
Newburgh et al., 1930 Fecal loss −100 to −200
Kuno, 1956 Insensible loss −450 to −1,900
Hoyt and Honig, 1996 Metabolic production +250 to +350
Total −1,300 to −3,450 +250 to +350
Net loss −1,050 to −3,100
Assuming conditions in which there is minimal water loss from sweating.
Table 4-2 displays estimated minimum losses and production of water (mL/day) in healthy sedentary adults, assuming conditions in which there is minimal water loss from thermoregulatory sweating. The following sections describe each source of water loss or production listed in this table.
Respiratory Water Loss
The amount of respiratory water loss, via evaporation within the lungs, is dependent on both the ventilatory volume and water vapor pressure gradient (Mitchell et al., 1972). Ventilatory volume is increased by physical activity, hypoxia, and hypercapnia, whereas the water vapor pressure is modified by the ambient temperature, humidity, and barometric pressure. Physical activity generally has a greater effect on respiratory water loss than do environmental factors. Daily respiratory water loss averages about 250 to 350 mL/day for sedentary persons, but can increase to 500 to 600 mL/day for active persons living in temperate2 climates at sea level (Hoyt and Honig, 1996). For these conditions, respiratory water loss (y = mL/day) can be predicted from metabolic rate (x = kcal/day) by the equation y = 0.107x + 92.2 (Hoyt and Honig, 1996). High altitude exposure (greater than 4,300 m, 448 mm Hg) can further increase respiratory water losses by approximately 200 mL/day (Hoyt and Honig, 1996).
In general, dry bulb temperatures of approximately 70°F, 80°F, and 90°F are used for temperate, warm, and hot conditions, respectively, in this report.
Ambient air temperature and humidity modify respiratory water losses. Breathing hot, dry air during intense physical exercise can increase respiratory water losses by 120 to 300 mL/day (Mitchell et al., 1972). Breathing cold, dry air during rest and stressful physical exercise (Table 4-3) can increase respiratory water losses by approximately 5 mL/hour and approximately 15 to 45 mL/hour, respectively (Freund and Young, 1996). Freund and Young (1996) have calculated that for a 24-hour military scenario (8 hours of rest, 12 hours of moderate activity, and 4 hours of moderate-heavy activity), the respiratory water losses increase by approximately 340 mL/day when breathing −20°C versus +25°C air.
Urinary and Gastrointestinal Water Loss
The kidneys are responsible for regulating the volume and composition of the ECF via a series of intricate neuroendocrine pathways (Andreoli et al., 2000). Renal fluid output can vary depending upon the specific macronutrient, salt, and water load. However, for persons consuming an average North American diet, some of these effects may not be discernable (Luft et al., 1983). Since there is a limit to how much the kidneys can concentrate urine, the minimal amount of water needed is determined by the quantity of end products that need to be excreted (e.g., creatinine, urea). On typical Western diets, an average of 650 mOsmol of electrolytes and other
TABLE 4-3 Influence of Breathing Cold Air and of Metabolic Rate on Respiratory Water Losses.
Temperature °F /°C; Relative Humidity (%); Water Vapor Pressure (mm Hg); Metabolic Rate (Watts); Respiratory Water Loss (mL/h)
77/25; 65;15; Rest (100); ≈ 10
32/0; 100; 5; Rest (100); ≈ 13
−4/−20;100; 1; Rest (100); ≈ 15
77/25; 65; 15; Light-moderate (300); ≈ 30
32/0; 100; 5; Light-moderate (300); ≈ 40
−4/−20; 100; 1; Light-moderate (300); ≈ 45
77/25; 65; 15; Moderate-heavy (600); ≈ 60
32/0; 100; 5; Moderate-heavy (600); ≈ 80
−4/−20; 100; 1; Moderate-heavy (600); ≈ 90
SOURCE: Reprinted with permission, from Freund and Young (1996). Copyright 1996 by CRC Press.
On typical Western diets, an average of 650 mOsmol of electrolytes and other solutes must be excreted per day to maintain electrolyte balance; thus, if the urine is maximally concentrated (Uosm approximately 1,200 mOsmol/kg water), the minimum urine output is approximately 500 mL/day. For dehydrated subjects living in hot weather, minimum daily urine outputs can be less than 500 mL/day (Adolph, 1947b).
Urine output generally averages 1 to 2 L/day but can reach 20 L/day in those consuming large quantities of fluid (West, 1990). Healthy older individuals, however, cannot concentrate urine as well as young individuals and thus have a higher minimum urine output. For example, older men and women (mean age 79 years) had lower maximal urine osmolalities of 808 and 843 mOsm/kg, respectively, compared with 1,089 mOsm/kg for young men (mean age 24 years). This corresponds to higher minimum urine outputs of 700 and 1,086 mL/day for the older men and women compared with 392 mL/day for the young men (Dontas et al., 1972).
Urine output varies inversely with body hydration status.
FIGURE 4-4 Relation of urine output to body hydration status. Reprinted with permission, from Lee (1964). Copyright 1964 Handbook of Physiology, Section 4, American Physiological Society.
Figure 4-4 depicts the hyperbolic relationship between urine output and body hydration status: one asymptote ascends steeply with hyperhydration, while the other descends gradually with dehydration (Lee, 1964). The apex of this hyperbolic relationship approximates a urine output of approximately 50 mL/hour. The extremes depicted in Figure 4-4 can be exceeded. For example, investigators have reported that urine output can transiently increase to approximately 600 to 1,000 mL/hour with water loading (Freund et al., 1995; Noakes et al., 2001; Speedy et al., 2001) and decrease to approximately 15 mL/hour with dehydration (Adolph, 1947b). Urine output can vary widely to maintain total body water; however, there are clearly limits to the amount of conservation and excretion.
Physical activity and climate also affect urine output. Exercise and heat strain will reduce urine output by 20 to 60 percent (Convertino, 1991; Mittleman, 1996; Zambraski, 1996), while cold and hypoxia will increase urine output (Freund and Young, 1996; Hoyt and Honig, 1996).
Gastrointestinal and thus fecal water loss in healthy adults is approximately 100 to 200 mL/day (Newburgh et al., 1930).
Insensible and Sweat Losses
Water loss through the skin occurs via insensible diffusion and secreted sweat. For the average adult, loss of water by insensible diffusion is approximately 450 mL/day (Kuno, 1956). During heat stress, eccrine sweat glands secrete sweat onto the skin surface, which cools the body when water evaporates from the sweat. In hot weather, sweat evaporation provides the primary avenue of heat loss to defend the body’s core temperature. When a gram of sweat water is vaporized at 30°C, 2.43 kJ (0.58 kcal) of heat becomes kinetic energy (latent heat of evaporation) (Wenger, 1972). For a given hot weather condition, the required sweating rate for evaporative cooling is dependent upon the physical activity level (metabolic rate).
The following calculations provide the minimal sweat produced by persons performing moderately heavy (metabolic rate ≈ 600 W) exercise in the heat (Sawka et al., 1996a). If the activity is 20 percent efficient, the remaining 80 percent of metabolic energy produced is converted to heat in the body so that 480 W (0.48 kJ/second, or 28.8 kJ/minute or 6.88 kcal/minute) need to be dissipated to avoid heat storage. The specific heat of body tissue (amount of energy required for 1 kg of tissue to increase temperature by 1°C) approximates 3.5 kJ (0.84 kcal)/kg/°C. For example, a 70-kg man has a heat capacity of 245 kJ (59 kcal)/°C, and a 50-kg woman has a heat capacity of 173
Page 84
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
kJ (41 kcal)/°C. If these persons performed exercise in a hot environment that enabled only evaporative heat loss and they did not sweat, their body temperatures would increase by approximately 1.0°C every 8.5 min for the man (245 kJ/°C ÷ 28.8 kJ/minute or 59 kcal/°C ÷ 6.88 kcal/minute) and every 6 minutes for the woman (173 kJ/°C ÷ 28.8 kJ/minute or 41 kcal/°C ÷ 6.88 kcal/minute). Since the latent heat of evaporation is 2.43 kJ/g (0.58 kcal/g), such persons would need to evaporate approximately 12 g of sweat per minute (28.8 kJ/minute ÷ 2.43 kJ/g or 6.88 kcal/minute ÷ 0.58 kcal/ g) or 0.72 L/hour. Because secreted sweat drips from the body and is not evaporated, higher sweat secretions are often needed to achieve these cooling demands. If a person is physically active and exposed to environmental heat stress, sweat losses to avoid heat storage can be substantial over a 24-hour period.
For persons living in hot climates, daily sweat losses often exceed several liters. As described above, daily sweat losses are determined by the evaporative heat loss requirements, which are influenced by the metabolic rate (above example) and environment. The environmental factors that modify sweat losses include clothing worn, ambient temperature, humidity, air motion, and solar load. Therefore, considerable variability will exist for daily sweat losses among different people. Figure 4-5 provides the distribution of daily sweat-
FIGURE 4-5 Distribution of daily sweating rates for active soldiers in desert and tropical climates. Percent incidence refers to the percentage of the subject population achieving the given daily sweat loss.
SOURCE: Molnar (1947). Reprinted with permission from the Papers of Edward Adolph collection at the Edward G. Miner Library, University of Rochester Medical Center.
Page 85
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
ing rates for soldiers living in desert and tropical climates (without air conditioning). The average daily sweat loss for 97 men in the desert was 4.9 L; for 26 men in the tropics, it was 2.3 L. The lower daily sweat losses in the tropics were probably due to lower ambient temperatures and lower solar load (both acting to lower the required evaporative cooling), as the precise activity levels of both groups were unknown.
Metabolic Water Production
Metabolic water is formed by oxidation of hydrogen-containing substrates during metabolism or energy-yielding nutrients. Oxidation of carbohydrate, protein, and fat produces metabolic water of approximately 15, 10.5, and 11.1 g/100 kcal of metabolizable energy, respectively (Lloyd et al., 1978). Therefore, metabolic water production is proportional to the energy expenditure with a small adjustment for the substrate oxidized. Figure 4-6 shows the metabolic water production relative to daily energy expenditure for persons eating a mixed diet (Hoyt and Honig, 1996). If the regression line in Figure 4-6 is extrapolated to the daily energy expenditures of ≈ 2,500 kcal/day, the metabolic water production will approximate 250 mL/day. Therefore, a reasonable estimate of daily metabolic
FIGURE 4-6 Metabolic water production relative to daily energy expenditure. Reprinted with permission, from Hoyt and Honig (1996). Copyright 1996 by CRC Press.
Page 86
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
water production is an average of approximately 250 to 350 L/day for sedentary persons—but which can increase to 500 to 600 mL/ day for physically active persons (Hoyt and Honig, 1996). Hence, respiratory water losses are roughly equivalent to, or offset by, metabolic water production (Table 4-2; Hoyt and Honig, 1996). Metabolic water, a by-product of metabolizing energy-yielding nutrients from foods into carbon dioxide and energy, does not include the water present in a foodstuff itself. This is considered compositional water, or moisture. It is often determined analytically as the difference in weight of a food item before and after drying to a constant weight.
Consumption
Fluid is consumed in the form of food and beverages, and, regardless of form, is absorbed by the gastrointestinal tract and acts the same physiologically. In one survey of the adult U.S. population (1977–1978 Nationwide Food Consumption Survey), total water intake was approximately 28 percent from foods, 28 percent from drinking water, and 44 percent from other beverages (Ershow and Cantor, 1989). National survey data for adults (Appendix Tables D-1, D-3, and D-4) likewise suggest that approximately 20 percent of water comes from food, and the remaining 80 percent comes from fluids.
Drinking induced by water deprivation is homeostatic (Greenleaf and Morimoto, 1996). Other factors (e.g., social, psychological) that influence drinking behavior are nonregulatory (Rolls and Rolls, 1982). Over an extended period, fluid consumption will match body water needs (if adequate amounts are available). However, mismatches can occur over short periods (Johnson, 1964). The fluid intake for healthy adults can vary markedly depending on activity level, environmental exposure, diet, and social activities; nonetheless, for a given set of conditions, intake is reproducible within persons (Johnson, 1964). Therefore, it is reasonable to assume that for large population studies of apparently healthy individuals, the fluid volume consumed is equal to or greater than body water needs.
METHODS FOR ESTIMATING WATER REQUIREMENTS
Water Balance
Water balance is regulated within ± 0.2 percent of body weight over a 24-hour period for healthy adults at rest (Adolph, 1943).
Page 87
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
Adolph (1943) described the rates of water gain and water loss relative to different levels of water deficit and excess. Induced water deficits or water excesses resulted in compensatory changes in water gains and water losses until water balance was reestablished. Likewise, Newburgh and colleagues (1930) demonstrated the accuracy of water balance studies to be within 0.5 percent of the water volume. Therefore, ad libitum water balance studies can be used to estimate daily water requirements, provided the subjects have adequate time for rehydration and physiologic compensation (Adolph, 1943; Newburgh et al., 1930). In both these studies, total water intake was measured.
Table 4-4 presents water balance studies that have estimated daily total water requirements for infants and children. Note that daily total water requirements increase with age from early infancy (approximately 0.6 L) through childhood (approximately 1.7 L). Since infants have rapid growth, some investigators express the daily water needs relative to body mass.
The minimal daily water requirement depends upon the person’s diet, environment, and activity level. After reviewing early water balance studies, Adolph (1933) concluded that for most adult men,
TABLE 4-4 Estimation of Daily Water Requirements of Infants and Children from Water Balance Studies
Reference
Subjects (age)
Conditions
Total Volume Intake, L/d (mL/kg/d)
Total Water Intake, L/d (mL/kg/d)
Goellner et al., 1981
Normal activity
15 infants
10 studies, 0–1 mo
0.66 (184)
0.56 (156)a
9 studies, 1–2 mo
1.00 (199)
0.85 (170)
14 studies, 2–4 mo
0.94 (161)
0.79 (137)
18 studies, 4–6 mo
1.13 (162)
0.96 (138)
39 studies, 6–12 mo
1.31 (158)
1.11 (135)
24 studies, 12–18 mo
1.57 (146)
1.33 (124)
21 studies, 18–24 mo
1.55 (129)
1.32 (110)
15 studies, 24–32 mo
1.62 (117)
1.38 (99)
Ballauff et al., 1988
21 children, 6–11 yr
Normal activity
≈ 1.7 for boys
≈ 1.5 for girls
a Goellner et al. (1981) estimated that water accounted for 85 percent or more of the determined volume intake. Thus total water intake was calculated as 85 percent of total volume intake.
Page 88
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
the minimal, average, and liberal water requirements approximated 2.1, 3.4, and 5.0 L/day, respectively. In addition, Adolph (1933) concluded that a convenient “liberal standard” for total water intake is 1 mL/kcal expended. Subsequent studies by Johnson (1964) recommended minimum daily water requirements of no less then 0.91 L for survival conditions and 3.0 L for hot weather.
Table 4-5 presents water balance studies that have estimated daily total water requirements for adults. These requirements are above minimal levels because some physical activity (although usually nominal) was allowed and because individuals self-selected the volume of consumed fluids (i.e., ad libitum water consumption). For the prolonged bed-rest studies, greater emphasis was placed on data obtained during the initial week, if available. Water balance studies suggest that the required water intake to maintain water balance for resting adult men is approximately 2.5 L/day (Adolph, 1933; Newburgh et al., 1930). If modest physical activity is performed, the
TABLE 4-5 Estimation of Daily Water Requirements of Adults from Water Balance Studies
Reference
Subjects
Conditions
Total Water Intake (L/d)
Women
Yokozawa et al., 1993
3 women
Temperate, bed-rest
≈ 1.6
Men
Newburgh et al., 1930
Repeated studies of men
Temperate, rest, variety of diets
≈ 2.6
Welch et al., 1958
53 men
Active, ambient temperature range of −30°C to +30°C
≈ 3.0 at −20°C to +20°C
≈ 6.0 at +30°C
Consolazio et al., 1967
6 men
Temperate, rest, starvation study
≈ 2.5 (1st 4 d; ~ 3.4 if corrected for negative balance)
Consolazio et al., 1968
24 men
Temperate, rest, sea level controls
≈ 2.5
Greenleaf et al., 1977
7 men
Temperate, bed-rest with 1 h of exercise/d
≈ 3.2
Gunga et al., 1993
6 men
Temperate, hyperbaric (1.5 atmospheres absolute), sedentary
≈ 3.2
Page 89
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
water intake requirements increase to approximately 3.2 L/day (Greenleaf et al., 1977; Gunga et al., 1993). Cold exposure did not alter intake, but heat stress increased total daily water intake (Welch et al., 1958).
Limited data were available for women. Women are physically smaller, thus they probably have lower water requirements due to lower metabolic expenditures. A study of three Japanese women (likely smaller than average U.S. adult women) indicated a water intake requirement of approximately 1.6 L/day (Yokozawa et al., 1993).
Water Turnover
Water turnover studies have been conducted to evaluate water needs and assume a balance between influx and efflux (Nagy and Costa, 1980). Rates of body water turnover can be determined by administering a drink with deuterium (D2O) or tritium (3H2O) labeled water and then following the decline (or disappearance) in hydrogen isotope activity over time. The isotope activity declines because of loss of the labeled water via excretion, evaporation, and dilution from intake of unlabeled water. If proper procedures are employed, these measurements will yield values within 10 percent or less of actual water flux (Nagy and Costa, 1980).
Figure 4-7 provides data on the daily water turnover for infants and children (Fusch et al., 1993). Water turnover (when expressed
FIGURE 4-7 Daily water turnover per kg of body weight in infants and children. Reprinted with permission, from Fusch et al. (1993). Copyright 1993 by Springer-Verlag.
Page 90
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
per kg of body weight) is highest during the first weeks of life and declines by some 40 percent during infancy. It declines further, but at a slower pace during childhood and adolescence. In a German study, mean water turnover at ages 1 to 3 months was 160 mL/kg/ day, compared with 97 mL/kg/day at ages 10 to 12 months, and 40 mL/kg/day at ages 13 to 15 years (Fusch et al., 1993). Daily fluid intake in bottle-fed infants was compared over a 15-day study period using two methods to determine intake (Vio et al., 1986). Water turnover as measured by deuterium tracer was compared with directly measured fluid intake. Daily fluid intakes of 0.71 L/day (153 mL/kg/day) and 0.70 L/day (151 mL/kg/day) were reported for the direct and water turnover methodology (r = 0.98), respectively. Other studies have found close agreement (Butte et al., 1988) or slightly higher (Butte et al., 1991) values for water turnover versus direct measurement of daily fluid intake in infants.
Table 4-6 provides studies examining daily water turnover for adults in a variety of conditions. These values are generally higher than in water balance studies because subjects are often more active and exposed to outside environments. Daily water turnover rates were approximately 3.2 L and 4.5 L for sedentary and active men, respectively. Several studies found daily water turnover rates greater than 5 L; presumably these were more active persons who may have encountered heat stress. Women generally had approximately 0.5 to 1.0 L/day lower daily water turnover rates than their male counterparts.
Water turnover was measured in 458 noninstitutionalized adults (ranging from 40 to 79 years of age) who lived in temperate climates (Raman et al., 2004). Daily turnover averaged 3.6 and 3.0 L in men and women, respectively. The water turnover values were corrected for metabolic water and water absorption from humidity to provide preformed water values. The preformed water values averaged 3.0 L/day (range 1.4 to 7.7 L/day) for men and 2.5 L/day (range 1.2 to 4.6 L/day) for women. The lower values in women were not accounted for by differences in body size.
METHODS FOR ESTIMATING HYDRATION STATUS
Total Body Water Changes
Total body water (TBW) is accurately determined by dilution of a variety of indicators. Repeated measurements are required to assess total body water changes. The technical requirements and cost for
Page 91
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
TABLE 4-6 Summary of Daily Water Turnover Studies on Adults
Reference
Subjects
Conditions
Water Turnover (L/d)
Schloerb et al., 1950
17 men
11 women
Not reported
3.4 men
2.3 women
Fusch et al., 1996
11 men, 2 women
Before and after high-altitude trek of 4,900 to 7,600 m
3.3 before (combined)
5.5 after (combined)
Leiper et al., 1996
6 men (sedentary)
6 men (active)
Temperate
3.3 (sedentary < 60 min exercise/d)
4.7 (active)
Lane et al., 1997
13 male astronauts
Ground-based period
3.8
Blanc et al., 1998
8 men
Sedentary
Head-down bed-rest
3.5
3.2
Fusch et al., 1998
11 men
4 women
Temperate
5.7 (combined)
Leiper et al., 2001
6 men (sedentary)
6 men (active)
Temperate
2.3 (sedentary)
3.5 (active)
Ruby et al., 2002
8 men
9 women
Arduous wildfire suppression activity
7.3 men
6.7 women
Raman et al., 2004
66 men (40–49 yr)
58 men (50–59 yr)
56 men (60–69 yr)
49 women (40–49 yr)
48 women (50–59 yr)
Temperate
36 women (60–69 yr)
3.8 (free living)
3.6
3.6
3.3
3.0
2.9
repeated measurements with dilution methods make them impractical for routine assessment of TBW changes. Bioelectric impedance analysis (BIA) has recently gained attention because it is simple to use and allows rapid, inexpensive, and noninvasive estimates of TBW. Absolute values derived from this technique correlate well with TBW values obtained by isotope dilution (Kushner and Schoeller, 1986; Kushner et al., 1992; Van Loan et al., 1995). These valida-
Page 92
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
tion studies were performed on euhydrated subjects under standardized clinical conditions (e.g., controlled diet, body posture, skin temperature, inactivity).
Studies have indicated that BIA may not have sufficient accuracy to validly detect moderate dehydration (approximately 7 percent TBW) and loses resolution with isotonic fluid loss (O’Brien et al., 1999). Because fluid, electrolyte, and plasma protein concentrations can have independent effects, BIA can provide misleading values regarding dehydration or hyperhydration status (Gudivaka et al., 1999; O’Brien et al., 2002). Fluid and electrolyte concentrations may have independent effects on the BIA signal, thus often providing grossly misleading values regarding dehydration status (O’Brien et al., 2002). The BIA with a 0/∞ − kHz parallel (Cole-Cole) multifrequency model may have promise to measure body hydration changes if corrections are made for changes in plasma protein concentration (Gudivaka et al., 1999). However, recently a multifrequency BIA with Cole-Cole analysis was reported not to be sensitive to hypertonic dehydration (Bartok et al., 2004).
Plasma and Serum Osmolality
Plasma osmolality provides a marker of dehydration levels. Osmolality is closely controlled by homeostatic systems and is the primary physiological signal used to regulate water balance (by hypothalamic and posterior pituitary arginine vasopressin secretion), resulting in changes in urine output and fluid consumption (Andreoli et al., 2000; Knepper et al., 2000). Plasma osmolality rarely varies beyond ± 2 percent and is controlled around a set-point of 280 to 290 mOsmol/kg; this set-point increases with aging and becomes more variable among people. Water deprivation (if it exceeds solute losses) increases the osmolality of plasma and of the ECF and thus fluids bathing the hypothalamus. This causes loss of ICF from osmoreceptor neurons, which then signals the release of arginine vasopressin from the hypothalamus and the posterior pituitary. Arginine vasopressin acts on the renal tubules to increase water reabsorption.
Arginine vasopressin release is proportional to increased plasma osmolality and decreased plasma volume. While body water loss will induce plasma volume reduction and increased plasma osmolality, the influence of body water loss on each depends upon the method of dehydration, physical fitness level, and heat acclimatization status (Sawka, 1988; Sawka and Coyle, 1999).
Many studies have measured plasma osmolality of euhydrated sub-
Page 93
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
TABLE 4-7 Plasma Osmolality for Euhydrated Subjects in Carefully Controlled Fluid Balance Studies
Reference
Subjects Mean age ± S.D.a
Plasma Osmolality (mOsmol/kg)
Sawka et al., 1983a
Men, 25 ± 4 yr
284
Sawka et al., 1983b
Men, 24 ± 3 yr
Women, 26 ± 3 yr
281
Sawka et al., 1984a
Men, 24 ± 3 yr
Women, 26 ± 3 yr
281
Fish et al., 1985
Men and women, 20–37 yr
Men and women, 62–88 yr
281
291
Sawka et al., 1988
Men, 33 ± 3 yr
283
Mack et al., 1994
Men, 18–28 yr
Men, 65–78 yr
281
287
Freund et al., 1995
Men, 24 ± 2 yr
287
Montain et al., 1995
Men, 24 ± 6 yr
281
Stachenfeld et al., 1996
Men and women, 24–33 yr
Men and women, 67–76 yr
282
286
Latzka et al., 1997
Men, 19–36 yr
282
Montain et al., 1997
Men, 24 ± 6 yr
281
Stachenfeld et al., 1997
Men and women, 20–28 yr
Men and women, 65–76 yr
285
288
Latzka et al., 1998
Men, 19–36 yr
283
O’Brien et al., 1998
Men, 24 ± 2 yr
280
Noakes et al., 2001
Men, 28–44 yr
279
Popowski et al., 2001
Men, 23 ± 3 yr
288
a S.D. ± stardard deviation.
jects in controlled fluid balance studies. Table 4-7 provides results from some of these studies. Note that plasma osmolality ranged from 279 to 291 mOsmol/kg and averaged approximately 284 mOsmol/kg, with slightly higher values for older populations. Elderly persons had approximately 3 to 6 mOsmol/kg higher plasma osmolality than the young adults studied (Mack et al., 1994; Stachenfeld et al., 1996, 1997).
Figure 4-8 provides a compilation of 19 studies (181 subjects) where plasma osmolality was measured at several hydration levels. TBW was either directly measured or calculated based upon body composition information. A strong negative relationship (p < 0.0001) (r = −0.76) was found between TBW changes and plasma osmolality changes. Similar relationships have been reported based on smaller sample sizes of individual data (Sawka et al., 2001; Senay and Christensen,
Page 94
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
FIGURE 4-8 Relationship of change in plasma osmolality to change in total body water from 19 studies representing 181 subjects (Armstrong et al., 1985, 1997; Cheung and McLellan, 1998; Gonzalez-Alonso et al., 1997; Maresh et al., 2001; Maughan et al., 1996; Miescher and Fortney, 1989; Montain and Coyle, 1992; Montain et al., 1995; Neufer et al., 1989a, 1991; Noakes et al., 2001; O’Brien et al., 1998; Sawka et al., 1983b, 1985, 1988, 1989a, 1989b, 1992). The data points represent mean data reported in these studies. y = 0.2943 − 1.2882x; p = < 0.0001.
1965). Clearly, plasma osmolality provides a good marker for dehydration status if water loss is greater than solute loss. When solute and water are lost proportionately, such as with diarrhea or vomiting, osmolality remains constant and vasopressin release is blunted. However, the resulting ECF loss will stimulate the renin-angiotensin-aldosterone system as a means to increase sodium and hence water retention (Share et al., 1972). This mechanism appears to be less robust in elderly individuals (Dontas et al., 1972).
Table 4-8 provides the serum osmolality for selected deciles of total water intake by gender in the Third National Health and Nutrition Examination Survey (NHANES III). A more complete presentation of NHANES III data can be found in Appendix Table G-1. Serum osmolality concentrations were essentially identical (maximum range 3 mOsmol/kg) for the lowest (1st), middle (5th), and highest (10th) deciles within each age group. These data indicate that persons in the lowest and highest deciles of total water intake were not systematically dehydrated or hyperhydrated. In agreement
Page 95
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
TABLE 4-8 Serum Osmolality Concentration for Selected Deciles of Daily Total Water Intake in Men and Women
Age
Decile of Total Water Intake
Men
Women
Total Water Intake, L/d (mean)
Mean Serum Osmolality (mOsmol/kg)
Mean Total Water Intake (L/d)
Mean Serum Osmolality (mOsmol/kg)
12–18 yr
1st
1.36
278
0.94
278
5th
2.79
279
2.20
276
10th
6.46
281
5.52
277
19–50 yr
1st
1.69
279
1.25
277
5th
3.31
280
2.61
277
10th
7.93
280
6.16
277
51–70 yr
1st
1.64
280
1.32
281
5th
3.17
283
2.68
281
10th
7.20
281
5.81
279
71+ yr
1st
1.44
283
1.19
282
5th
2.71
283
2.38
283
10th
5.45
281
4.85
282
SOURCE: Third National Health and Nutrition Examination Survey, Appendix Table G-1.
with Table 4-8, the oldest persons (greater than 70 years of age) had slightly higher serum osmolality levels. The serum osmolality concentrations observed in NHANES III (Table 4-8) were slightly lower for all age groups than the plasma osmolality levels from the balance studies previously described (Table 4-7). In general, serum and plasma osmolality values are usually nearly identical; however, several handling and analytical factors can cause small differences between them (Tietz, 1995).
Plasma Sodium Concentration
Sodium is the primary cation of the ECF. Any loss of water in greater proportion than electrolyte losses will increase sodium concentrations in ECF compartments. Figure 4-9 provides a compilation of four studies (32 subjects) where plasma sodium concentration was measured at several hydration levels. TBW was either directly measured or calculated based upon body composition information. A moderate negative relationship (r = −0.46) was obtained between the decrease in TBW and increase in plasma sodium levels (p = 0.14). If data are analyzed for only the studies that
Page 96
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
FIGURE 4-9 Relationship of change in plasma sodium to change in total body water from 4 studies representing 32 subjects (Fallowfield et al., 1996; Maughan et al., 1996; McConell et al., 1999; Montain et al., 1995). The data points represent mean data reported in these studies. y = 1.6927 − 0.4175x; p = 0.14.
presented both osmolality (Figure 4-9) and sodium data, then negative correlations of r = −0.82 and r = −0.28 were found between decreases in TBW and increases in osmolality and sodium levels, respectively. A negative relationship of r = −0.71 and r = −0.57 (based on 22 experiments) has been reported between decreases in TBW (as measured by body weight changes) and increases in plasma osmolality and plasma sodium levels, respectively (Senay and Christensen, 1965). Based on this data, plasma sodium changes are not as strongly related to changes in body hydration status as plasma osmolality changes.
Analysis of the data on plasma osmolality and sodium concentrations measured in nine heat acclimated subjects when euhydrated and after thermal dehydration by 3 and 5 percent of their weight indicated strong negative relationships between a decrease in total body water and (1) an increase in osmolality (r = −0.92), and (2) an increase in sodium (r = −0.90) (Montain et al., 1997). Further analysis indicated a relationship (r = 0.56) between the increases in sodium and in osmolality. Figure 4-10 depicts these data; note that the magnitude of increased plasma sodium concentration is markedly less than the increase in plasma osmolality. Therefore, the smaller increase in sodium concentration for a given water deficit may result in a smaller range for interstudy analyses and lead to
Page 97
Suggested Citation: "4 Water." Institute of Medicine. 2005. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. Washington, DC: The National Academies Press. doi: 10.17226/10925. ×
FIGURE 4-10 Relationship of change in plasma osmolality and plasma sodium concentration changes from thermal dehydration. Data from Montain et al. (1997). y = 0.2218x + 1.5461, p = 0.0002.
weaker relationships between change in plasma sodium and change in hydration status.
Plasma Volume Changes
Hyperhydration induces a modest increase in plasma volume (Freund et al., 1995; Latzka et al., 1997). Dehydration will decrease plasma volume, but the magnitude of reduction is variable. For example, heat acclimatized persons have a smaller plasma volume reduction for a given body water deficit than do unacclimatized persons (Sawka et al., 1988). By virtue of having a more dilute sweat, heat acclimatized persons have additional solutes remaining within the extracellular space to exert an osmotic pressure and redistribute fluid from the intracellular space. If an individual dehydrates from diuretic medication, a much greater ratio of plasma loss to total body water loss occurs compared with exercise-heat induced dehydration (O’Brien et al., 1998).
Figure 4-11 provides a compilation of 16 studies (146 subjects) where plasma volume was measured at several hydration levels. TBW was either directly measured or calculated based upon body composition information.
FIGURE 4-11 Relationship of change in plasma volume to change in total body water from 16 studies representing 146 subjects (Armstrong et al., 1985; Cheung and McLellan, 1998; Fallowfield et al., 1996; Gonzalez-Alonso et al., 1997; Kristal-Boneh et al., 1988; McConell et al., 1999; Miescher and Fortney, 1989; Montain and Coyle, 1992; Montain et al., 1995; O’Brien et al., 1998; Sawka et al., 1983b, 1985, 1988, 1989a, 1989b, 1992). The data points represent mean data reported in these studies. y = −1.0466 + 0.7270x, p = 0.0004.
A moderate correlation (r = 0.56) was observed between change in TBW and change in plasma volume. A strong relationship (r = 0.70) between plasma volume reduction and TBW reduction was seen in individual data on heat acclimatized subjects (Sawka et al., 2001). However, since subject status (e.g., heat acclimatization and perhaps physical fitness) and method of dehydration modifies the plasma volume reduction for a given dehydration level (Sawka, 1992), it is probably not a good index of hydration for all populations.
Blood Urea Nitrogen
Although blood urea nitrogen (BUN) is primarily considered an indicator of kidney function, it is also used as an indicator of dehydration in clinical settings. The pattern of high BUN (normal range 8 to 25 mg/dL) and otherwise normal renal function (e.g., normal creatinine or creatinine clearance) is considered an indicator of hypovolemia (a reduction in plasma or blood volume). However, BUN is also directly related to protein intake. Therefore, while BUN can be an indicator of hydration status, other biochemical values must be considered in order to assess hydration status versus kidney function.
An elevated BUN:creatinine ratio (greater than 25) was seen in 2 of 37 elderly, long-term care patients who experienced no febrile episodes and no documentation of impaired oral intake (Weinberg et al., 1994a). The BUN:creatinine ratio remained relatively constant over a 6-month period in stable male residents (Weinberg et al., 1994b). Still, although the BUN:creatinine ratio, like BUN itself, has been used to assess hydration status, lack of specificity hinders its use as a measure of hydration status.
Urine Indicators
Volume and Color
Urine volume is often used as an indicator of hydration status. If healthy individuals have urine outputs of approximately 100 mL/ hour, they are probably well hydrated (see Figure 4-4). Higher urine outputs (300 to 600 mL/hour) are probably indicative of fluid excess (Freund et al., 1995; Lee, 1964). If urine output falls to less than 30 mL/hour for extended periods with an average diet, the person is probably dehydrated (see Figure 4-4).
The color of urine darkens or lightens with low or high output levels (because the solute load is either concentrated or diluted, respectively). Thus urine color has been used as an indicator of hydration status (Wakefield et al., 2002). However, no precise relationship between urine color and hydration level exists. Furthermore, diet, medications, and vitamin use can affect urine color. Nonetheless, urine color can provide a good educational tool for dehydration or overhydration (Casa et al., 2000). A urine color chart for athletes to teach them about proper hydration is available (Casa et al., 2000). Although not nearly as precise as biochemical measures, urine color can give a crude indication of hydration status.
Urine Specific Gravity and Urine Osmolality
Because urine becomes more concentrated with dehydration, both urine specific gravity and urine osmolality have been used as indicators of hydration status. Urine specific gravity and urine osmolality increase with dehydration and are strongly correlated (r = 0.82−0.97) with each other (Armstrong et al., 1994; Popowski et al., 2001). It should be noted that the validity of the urine specific gravity and urine osmolality as indices in assessing hydration status is improved when the first morning urine, rather than a random collection, is used due to a more uniform volume and concentration (Sanford and Wells, 1962; Shirreffs and Maughan, 1998). Many studies have used these urine indices to access fluid balance and found poor (Armstrong et al., 1994; Francesconi et al., 1987; Hackney et al., 1995; O’Brien et al., 1996) or moderate (Adolph, 1947b; Shirreffs and Maughn, 1998) relationships with different indicators of dehydration status. For example, nonsignificant relationships between plasma osmolality with urine specific gravity (r = 0.46) and with urine osmolality (r = 0.43) were found in a well-controlled study of thermally dehydrated subjects (Popowski et al., 2001).
For “normally” hydrated (euhydrated) persons, urine specific gravity values range from 1.010 to 1.030 (Armstrong et al., 1994; Popowski et al., 2001; Sanford and Wells, 1962; Zambraski et al., 1974). It has generally been accepted that a urine specific gravity of less than or equal to 1.02 represents euhydration (Armstrong et al., 1994; Popowski et al., 2001), and a urine specific gravity greater than 1.03 represents dehydration (Armstrong et al., 1994; Francesconi et al., 1987; Popowski et al., 2001). Adolph (1947b) published individual data regarding urine specific gravity at different levels of water deficit (Figure 4-12). Urine specific gravity increases with water deficit; however, considerable individual variability exists. Although a urine specific gravity greater than 1.03 indicates probable dehydration, the magnitude of the water deficit cannot be determined.
Normal values for urine osmolality vary from 50 to 1,200 mOsmol/L (Tilkian et al., 1995). Therefore, in the setting of such variability, there may be no single threshold for urine osmolality and hydration status. However, individual increases in urine osmolality can provide an approximation of a person’s water deficit, assuming the solute load remains constant (Armstrong et al., 1994; Shirreffs and Maughan, 1998). In addition, urine osmolality is increased when osmotically active solutes are excreted, such as glucose in patients with uncontrolled diabetes mellitus (Tilkian et al., 1995). For these reasons (i.e., high variability and its dependence on solute excretion), urine osmolality is not considered a good indicator of hydration status.
Saliva Specific Gravity
Saliva specific gravity is slightly higher than water (Shannon and Segreto, 1968).
FIGURE 4-12 Individual urine specific gravity values at a range of water deficit levels.
SOURCE: Adolph (1947b). Reprinted with permission from the Papers of Edward Adolph collection at the Edward G. Miner Library, University of Rochester Medical Center.
Several studies have examined dehydration and salivary specific gravity. Salivary flow was shown to decrease after a water deficit exceeding 2 percent of body weight, but there was considerable variability in response (Adolph and Wills, 1947). Significant decreases in saliva flow rate were found during dehydration of 2 to 3 percent body weight using 24-hour water deprivation studies (Ship and Fischer, 1997, 1999). One study determined that salivary osmolality increases during exercise in the heat accompanied by modest (2.9 percent body weight loss) dehydration (Walsh et al., 2004).
Body Weight Changes
Body weight changes are frequently used to estimate sweating rates and therefore changes in total body water (e.g., Gosselin, 1947). This approach is usually used to estimate changes over a relatively short duration when food and fluid intakes and excretions are carefully controlled. The validity of this estimate depends upon body weight measurements not being confounded by other nonfluid factors that can influence body weight changes. If proper controls are made, body weight changes can provide a more sensitive estimate of total body water changes than repeat measurements by dilution methods (Gudivaka et al., 1999).
Potential confounding effects of urine loss, fluid intake, respiratory water loss, metabolic mass loss, water trapped perspiration in clothing on sweat loss, and therefore total body water change estimates for individuals performing exercise in hot and cool conditions have been examined (Cheuvront et al., 2002). Significant errors in estimating sweating rate are introduced unless nonperspiration fluid losses are factored into the body weight changes (Cheuvront et al., 2002). Likewise, carbohydrate loading in athletes will result in elevated baseline body weights that do not reflect euhydration, as the muscle glycogen will osmotically hold water. Overall, body weight changes provide an effective index of body water changes if other factors influencing body weight are carefully controlled.
Thirst
Thirst is “the desire to drink by both physiological and behavioral cues, resulting from deficit of water” (Greenleaf, 1992), through which people replenish their fluid losses during short-term periods (several hours) (Adolph and Wills, 1947; Eichna et al., 1945). Various scales have been developed over the years to quantify thirst by rating the sensation of, for example, dry mouth or dry throat. However, the most practical and commonly used approach in animal and human studies has been to document the volume of ad libitum (voluntary) drinking as a surrogate measurement of thirst. Despite ad libitum drinking, humans tend to under-replace their fluid needs over the short term (Johnson, 1964).
Triggering of thirst occurs through perceptual and physiological mechanisms (Fitzsimons, 1976; Greenleaf and Morimoto, 1996; Rolls and Rolls, 1982). For example, increases in plasma osmolality, plasma volume reduction, and several thirst sensations all made substantial contributions to predicting ad libitum fluid replacement following water deficits of 3, 5, and 7 percent of body weight loss (Engell et al., 1987).
Perceptual Factors
Voluntary drinking of a beverage is affected by its palatability, which is determined by its color, flavor, odor, and temperature (Boulze et al., 1983; Hubbard et al., 1984; Meyer et al., 1994; Szlyk et al., 1989; Wilk and Bar-Or, 1996; Zellner et al., 1991). These factors are greatly influenced by cultural preferences; therefore, broad generalizations are difficult. In a study on the effect of water temperature on voluntary drinking, dehydrated men drank the highest amounts when the water temperature was 15°C (59°F). Higher and lower temperatures resulted in a smaller drinking volume, even though the cooler drinks were rated more “pleasurable” (Boulze et al., 1983). In another study, water at 15°C (59°F) was consumed at greater volumes than water at 40°C (104°F) (Szlyk et al., 1989). When children were exposed to 3 hours of intermittent exercise at 35°C (95°F) and 45 to 50 percent relative humidity, their ad libitum consumption of flavored water was 45 percent greater than with unflavored water (Figure 4-13) (Wilk and Bar-Or, 1996). Likewise, adults who performed desert-simulated walks at 40°C (104°F) drank approximately 50 percent more flavored water than unflavored water (Hubbard et al., 1984).
The sweetness of a drink is a major factor in its palatability, but people differ in their preferred flavor. Flavor preference depends on various factors, including ethnic and cultural backgrounds. For example, in one study with Canadian children, most preferred grape to orange or apple flavors and drank more when presented with a
FIGURE 4-13 Cumulative voluntary drink intake of unflavored water (open circles), flavored water (black circles), and flavored sodium chloride (18 mmol/L) plus carbohydrate (6 percent) solution (triangles). Twelve 9- to 12-year-old boys cycled intermittently (black bars) at 35°C, 45 to 50% relative humidity. Reprinted with permission, from Wilk and Bar-Or (1996). Copyright 1996 by the American Physiological Society.
grape-flavored beverage (Meyer et al., 1994). In contrast, children in Puerto Rico had no preference for any single flavor (Rivera-Brown et al., 1999).
Physiological Triggers
Based on studies of various animal species, including humans, there seems to be three main physiological triggers for thirst: cerebral osmoreceptors, extra-cerebral osmoreceptors, and volume receptors (Fitzsimons, 1976; Greenleaf, 1992; Greenleaf and Morimoto, 1996). The osmoreceptors respond to cellular dehydration, which occurs when fluids leave the cells as a result of osmotic forces. The volume receptors respond to extracellular dehydration that results from loss of fluid from the vascular and interstitial spaces. While the osmoreceptors respond to small increases in osmolality, the volume receptors are activated by more drastic fluid losses. The osmoreceptors, therefore, are considered the first line of homeostatic defense against dehydration.
The location of these cells varies among species, but they are concentrated mostly in the hypothalamic area of the brain. Stimulation of the osmoreceptors activates drinking behavior and the release of arginine vasopressin hormone. The latter increases water permeability of the collecting tubules and thereby reduces free water loss and urine volume. There is evidence that either sodium chloride or an increase in osmolality (probably through separate cells) can activate the cerebral osmoreceptors, but it is assumed that the increase in osmotic forces is the more important stimulus (Greenleaf and Morimoto, 1996). The addition of 18 mmol/L of sodium chloride to flavored water triggered an increase of 31 percent in ad libitum drinking of children who exercised in the heat, compared with flavored water alone (Wilk and Bar-Or, 1996). Similar responses have been described for animals (Okuno et al., 1988) and adult humans (Nose et al., 1988).
Other osmoreceptors located in the oropharynx, gastrointestinal tract, and particularly the liver-portal system respond to drinking and modulate the thirst drive. Their existence has been postulated through experiments in which thirst and arginine vasopressin levels were modulated soon after drinking (or after injection of fluid to the liver portal system), before there were any changes in plasma osmolality or volume.
Thirst may be triggered by a decrease in blood volume, such as in hemorrhage or severe dehydration. This occurs through volume or stretch receptors that are sensitive to a drop in pressure at sites such
as the large systemic veins and the right atrium. These receptors, through the vagal system, stimulate thirst and drinking. Because of the compensatory activation of the renin-angiotensin-aldosterone system, preservation of body fluid is also achieved through a reduction in urinary output. Triggering of thirst through hypovolemia requires more than small changes in blood volume. The role of various thirst mechanisms with altered hydration status has been reviewed in detail elsewhere (Mack and Nadel, 1996; Stricker and Sved, 2000). However, in almost all situations where smaller volumes are lost over time (such as 2 to 3 L of sweat over 6 hours due to high temperatures or exercise), thirst mechanisms come into play over the ensuing 24 hours to trigger replacement of fluids lost; thus, in general, normal hydration is maintained by thirst mechanisms and normal drinking behavior. Such replacement is enhanced by consuming beverages at meals and in other social situations (Engell, 1995; Szlyk et al., 1990), which may be a necessary component to achieve adequate rehydration within a short period of time due to minor fluid deficits induced by exercise or heat strain.
Dehydration, Health, and Performance
Well-Being and Cognition
Dehydration can adversely influence cognitive function and motor control. Dehydration and poor mental function have been reported to be associated in physically ill older people (Seymour et al., 1980). Table 4-9 summarizes studies that examined the effects of dehydration on cognitive performance and motor function in healthy individuals.
Interpretation of these reports is difficult because the experimental designs often do not allow discrimination of confounding factors, such as effect of thermal (or exercise) stress and that of dehydration per se (Epstein et al., 1980; Hancock, 1981; Leibowitz et al., 1972; Sharma et al., 1983). For example, a degradation in mental alertness, associative learning, visual perception, and reasoning ability were noted when healthy men exercised while exposed to a high climatic heat stress (Sharma et al., 1983). Although the subjects drank water ad libitum, they may not have consumed enough fluids over the 4-hour session and thus became dehydrated due to the exercise and heat stress. However, the possible effect of dehydration on the above mental functions was not addressed. In another study, men and women exercised in the heat for 6 hours to elicit dehydration levels of 2.5 and 5 percent (Leibowitz et al., 1972).
TABLE 4-9 Cognitive and Motor Control Functions Reported to Be Affected by Dehydration
Function
Reference
Subjects
Conditions
Results
Perception of fatigue
Cian et al., 2000
8 men
2.8% dehydration by exercise or climatic heat
Increased rating of fatigue
Rating of mood
Cian et al., 2000
8 men
2.8% dehydration by exercise or climatic heat
No effect on mood
Target shooting
Epstein et al., 1980
9 men
2.5% dehydration by climatic heat
Reduced speed and accuracy and increase in physiologic strain
Perceived discrimination
Cian et al., 2000
8 men
2.8% dehydration by exercise or climatic heat
Discrimination impaired
Choice reaction time
Leibowitz et al., 1972
4 men, 4 women
6-h exercise in the heat, causing 2.5% or 5% dehydration visual
Faster response time to peripheral stimuli, no effect on response time to central visual stimuli
Cian et al., 2000
8 men
2.8% dehydration by exercise or climatic heat
No effect on response time
Visual-motor tracking
Gopinathan et al., 1988
11 men
1, 2, 3, or 4% dehydration, induced by exercise in the heat
Tracking impaired at 2% or more dehydration
Short-term memory
Cian et al., 2000
8 men
2.8% dehydration by exercise or climatic heat
Short-term memory impaired
Gopinathan et al., 1988
11 men
1, 2, 3, or 4% dehydration, induced by exercise in the heat
Short-term memory impaired at 2% or more dehydration
Function
Reference
Subjects
Conditions
Results
Long-term memory
Cian et al., 2000
8 men
2.8% dehydration by exercise or climatic heat
Impaired recall, especially following exercise
Attention
Gopinathan et al., 1988
11 men
1, 2, 3, or 4% dehydration, induced by exercise in the heat
Attention impaired at 2% or more dehydration
Arithmetic efficiency
Gopinathan et al., 1988
11 men
1, 2, 3, or 4% dehydration, induced by exercise in the heat
Arithmetic ability impaired at 2% or more dehydration
There was no difference in reaction time in response to central visual cues, but reaction time decreased when the visual cues were given at the periphery of the field of vision during the two dehydration conditions. Once again, interpretation of this finding is difficult because factors such as climatic heat stress, exercise-related fatigue, and boredom were not removed.
In a well-designed study, the arithmetic ability, short-term memory, and visual-motor tracking of 11 men who, on separate days, had water deficits of either 1, 2, 3, or 4 percent of body weight via thermal dehydration were assessed (Gopinathan et al., 1988). The subjects had ample rest in a temperate environment once they reached the target dehydration. This design allowed the researchers to observe the effects of dehydration per se, without fatigue or heat stress. This study revealed that a threshold level of 2 percent dehydration is required for deterioration of mental functions. A similar threshold was reported by other investigators (Sharma et al., 1986).
The adverse effects on mental function occurred irrespective of whether dehydration was achieved through exposure to the heat or as a result of exercise (Cian et al., 2001). A previous study by the same group suggested that exercise-induced dehydration was accompanied by a greater reduction in long-term memory (Cian et al., 2000), but the decrement in other functions was similar despite the mode of dehydration.
In conclusion, there is evidence to suggest that water deficits of 2 percent of body weight or more are accompanied by declining men-
tal function (Epstein et al., 1980). The mechanisms for this deficiency have not been elucidated.
Physical Work
TABLE 4-10 Dehydration Effects on Maximal Aerobic Power and Physical Work Capacity
Study
Subjects
Environmenta
Dehydration Process
Buskirk et al., 1958
13 men
83°C (115°F)
Heat
Saltin, 1964
10 men
36–38.5°C (68–70.5°F)
Heat and exercise
Craig and Cummings, 1966
9 men
46°C (78°F)
Heat and exercise
Herbert and Ribisl, 1972
8 men
N/A
Fluid restriction
Houston et al., 1981
4 men
N/A
Fluid restriction
Caldwell et al., 1984
16 men
N/A
Exercise
15 men
N/A
Diuretic
16 men
80°C (112°F), 50% RH
Sauna
Pichan et al., 1988
25 men
39°C (71°F), 60% RH
Fluid restriction and exercise in sauna
Webster et al., 1990
7 men
N/A
Exercise in rubberized sweat suit
Sawka et al., 1992
17 men
49°C (81°F), 20% RH
Fluid restriction and exercise
Burge et al., 1993
8 men
N/A
Exercise and fluid restriction
Walsh et al., 1994
6 men
30°C (62°F), 60% RH
Fluid restriction
Below et al., 1995
8 men
31°C (63°F), 54% RH
Fluid restriction
Fallowfield et al., 1996
4 men, 4 women
N/A
Fluid restriction
Montain et al., 1998b
5 men, 5 women
40°C (72°F), 20% RH
Exercise and hot room
a N/A = not available, RH = relative humidity.
b TM = treadmill, CY = cycle ergometer.
c NC = no change.
Body water deficits can adversely influence aerobic exercise tasks (Sawka, 1992; Sawka and Coyle, 1999). The critical water deficit and magnitude of performance decrement are related to the environmental temperature, exercise task, and probably the subject’s unique biological characteristics (physical fitness, acclimatization state, tolerance to dehydration). Table 4-10 presents a summary of investigations concerning the influence of dehydration on maximal aerobic power and physical work capacity (e.g., how much aerobic-type exercise could be completed under a given set of conditions) in adults.
% Δ Wt
Exercise Modeb
Baseline Maximum Power (L/min)
Δ Maximum Aerobic Powerc
Physical Work
−5
TM
↓ (−0.22 L/min)
—
−4
CY
3.96
NC
↓ (33%)
−2
TM
≈ 3.8
↓ (10%)
↓ (22%)
−4
TM
≈ 3.8
↓ (27%)
↓ (48%)
−5
CY
—
↓ (17%)
−8
TM
4.3
NC
—
−3
CY
3.61
NC
↓ (7 Watts)
−4
CY
4.15
↓ (8%)
↓ (21 Watts)
−5
CY
4.25
↓ (4%)
↓ (23 Watts)
−1
CY
—
↓ (6%)
−2
CY
↓ (8%)
−3
CY
↓ (20%)
−5
TM
3.76
↓ (7%)
↓ (12%)
−8
TM
—
↓ (54%)
−5
Rowing
4.65
NC
↓ (5%)
−1.8
CY
2.9
NC
↓ (34%)
−2
CY
↓ (6.5%)
—
−2
TM
—
↓ (25%)
−4
Leg kick
—
↓ (15%) endurance
In a temperate climate, body water deficits of less than 3 percent of body weight did not reduce maximal aerobic power; however, in hot climates, water deficits of 2 percent resulted in large reductions. Physical work capacity was reduced by dehydration in almost all examined conditions, with a greater effect when heat stress was also present. The influence of factors such as a person’s initial maximal aerobic power, training status, and heat acclimatization status on the magnitude of aerobic performance decrements from body water deficits has not been delineated. In a study of dehydration in children at 1 and 2 percent of body weight loss, a greater increase in core body temperature than would have been expected to be observed in adults exercising in hot weather was noted (Bar-Or et al., 1980). Therefore, children may have greater adverse performance effects from the same extent of dehydration during heat stress than do adults.
The effects of body water loss on endurance exercise performance in 13 endurance exercise studies have been reviewed (Cheuvront et al., 2003) (see Table 4-11). Based on these studies, dehydration appears to alter cardiovascular, thermoregulatory, central nervous system, and metabolic functions. One or more of these alterations will degrade endurance exercise performance when dehydration exceeds 2 percent of body weight. These performance decrements are accentuated by heat stress.
In summary, the literature indicates that dehydration can adversely influence aerobic and endurance-type exercise performance. The level of body water deficit needed to induce performance decrements probably approximates 2 percent body weight deficit; however, some individuals are probably more sensitive and others less sensitive to the amount of body water deficit on performance consequences. In addition, experimental evidence supports the concept that greater body water deficits result in a greater magnitude of performance decrements. Finally, it appears that heat stress increases these adverse performance consequences from body water deficits.
Body water deficits can adversely affect anaerobic exercise performance but do not appear to alter muscular strength. Table 4-12 lists a summary of investigations concerning the influence of dehydration on anaerobic exercise performance. Note that half of the studies reported reductions in anaerobic performance with considerable variability in the magnitude of performance reduction. Table 4-13 presents a summary of investigations examining the influence of dehydration on muscular strength. Most studies reported no effect of dehydration on muscular strength.
Thermoregulation (Fever and Hyperthermia of Exercise) and Heat Strain Tolerance
Fever is a regulated rise in body temperature and is a common response to inflammation, infection, and trauma (Blatteis, 1998; Leon, 2002). Dehydration will probably enhance the fever response and therefore has implications for management of clinical conditions. Rats dehydrated by a 24-hour water deprivation period exhibited a more severe fever than normally hydrated rats after being injected with bacterial endotoxin (Morimoto et al., 1986). Subsequent studies by other investigators have reproduced these findings in rats (Watanabe et al., 2000), as well as in rabbits (Richmond, 2001), and suggest the enhanced fever is due to angiotensin II secretion, which increases production of pyrogenic cytokines, such as interleukin-1.
However, studies in guinea pigs have reported that dehydration reduced the febrile response to bacterial endotoxin and suggest that the mechanism may be an antipyretic effect of central arginine vasopressin (Roth et al., 1992). Although there may be some species differences, it seems reasonable to conclude that dehydration may induce higher fevers. In support of this belief, febrile episodes have been found to be frequently associated with dehydration in nursing home residents (Weinberg et al., 1994a).
Dehydration and Heat Strain Tolerance
During exercise, unlike with a fever, an increase in body temperature does not represent a set-point change and is proportional to the metabolic rate (Sawka et al., 1996a). Dehydration increases core temperature responses during exercise in temperate and hot climates (Sawka and Coyle, 1999). A deficit of only 1 percent of body weight has been reported to elevate core temperature during exercise (Ekblom et al., 1970). Figure 4-14 summarizes results from studies that examined multiple dehydration levels within the same subjects during exercise. As the magnitude of water deficit increased, there was a concomitant graded elevation of core temperature. The magnitude of core temperature elevation ranged from 0.1°C to 0.23°C for every percent body weight lost (Brown, 1947a; Gisolfi and Copping, 1974; Greenleaf and Castle, 1971; Montain et al., 1998a; Sawka et al., 1985; Strydom and Holdsworth, 1968). The core temperature elevation from dehydration may be greater during exercise in hot compared with temperate climates. Dehydration not only elevates core temperature, but it negates many thermal
TABLE 4-11 Dehydration Effects on Endurance Exercise Performance
Reference
Sample Sizea
Exerciseb
Pitts et al., 1944
5 men
Walk 3.5 mph, 2.5% grade for 5 h
Brown, 1947a
13 men, NF
9 men, AL
21-mi desert hike
Ladell, 1955
4 men
Bench step to exhaustion
Maughan et al., 1989
6 men
CE 70% VO2max to exhaustion
Barr et al., 1991
5 men
3 women
CE 55% VO2max for 6 h (intermittent)
Walsh et al., 1994
6 men
CE 70% VO2max for 60 min, then 90% VO2max to exhaustion
Below et al., 1995
8 men
CE 50% VO2max for 50 min, then PR
Robinson et al., 1995
8 men
CE PR (total work in 60 min)
Fallowfield et al., 1996
4 men
4 women
TM run at 70% VO2max to exhaustion
McConell et al., 1997
7 men
CE 69% VO2max for 120 min, then 90% VO2max to exhaustion
Mudambo et al., 1997a
18 men, NF
6 men, SF
Walk/run/obstacle course (3 h)
McConell et al., 1999
8 men
CE 80% VO2max for 45 min, then 15 min PR
Bachle et al., 2001
4 men
7 women
CE 60 min PR
a NF = no fluid, AL = ad libitum, SF = some fluid (> NF, < F), F = fluid ≥ sweat losses.
b CE = cycle ergometer, PR = performance ride or run, TM = treadmill, VO2max = maximal oxygen uptake.
c RH = relative humidity.
d RPE = rating of perceived exertion, TTE = time to exhaustion.
SOURCE: Cheuvront et al. (2003). Reprinted with permission, from Cheuvront et al. (2003). Copyright 2003 by Current Science, Inc., Philadelphia, PA.
Environmentc
Drink Conditions
Dehydration (% body weight)
Performance Resultsd
35°C, 83% RH
NF, AL, F
No data
NF = ↓ (~60%) in walk duration; ↑ RPE vs. AL and F
31–39°C
NF, AL
NF = 6.3
AL = 4.5
NF = 7 of 13 failed to complete hike (54%)
AL = 3 of 9 failed to complete hike (33%)
38°C, 78% RH
38°C, 30% RH
NF, F
No data
NF = ↓ (25%) in work tolerance time vs. F
NF = ↓ (~20%) in walk duration; ↑ RPE vs. AL and F
Laboratory
NF, SF
NF = 1.8
SF = 2.0
No differences in TTE between NF and SF
30°C, 50% RH
NF, SF
NF = 6.4
F = 1.2
NF = ↓ (25%) in TTE and ↑ RPE vs. SF
30°C, 60% RH
NF, F
NF = 1.8
F = 0.0
NF = ↓ (31%) in TTE and ↑ in RPE vs. F
31°C, 54% RH
NF, F
NF = 2.0
F = 0.5
NF = ↓ (7%) in performance vs. F
20°C, 60% RH
NF, F
NF = 2.3
F = 0.9
NF = ↑ (1.7%) in PR vs. F
20°C
NF, SF
NF = 2.0
SF = 2.7
NF = ↓ (25%) in TTE vs. SF
21°C, 43% RH
NF, SF, F
NF = 3.2
SF = 1.8
F = 0.1
NF = ↓ (48%) in PR vs. F only
39°C, 28% RH
NF, SF
NF = 7
SF = 2.8
NF = 6/18 subjects failed to complete 3-h exercise bout vs. SF
NF = ↑ in RPE vs. SF
21°C, 41% RH
NF, SF, F
NF = 1.9
SF = 1.0
F = 0.0
No differences in PR among trials
21°C, 72% RH
NF, F
NF = 1.0
F = ↑ 0.5
TABLE 4-12 Dehydration Effects on Anaerobic Performance
Reference
Subjects
Dehydration Processa
% Δ Wt
Anaerobic Method
Resultb
Jacobs, 1980
11 men
Heat
−5
Wingate Anaerobic Test
NC
Houston et al., 1981
4 men
Fluid restriction
−8
Supramaximal run
NC
Nielsen et al., 1981
6 men
Diuretic
−3
Supramaximal cycle
↓ (18%) anaerobic capacity
6 men
Sauna
−3
Supramaximal cycle
↓ (35%) anaerobic capacity
5 men
Exercise
−3
Supramaximal cycle
↓ (44%) anaerobic capacity
Webster et al., 1990
7 men
Exercise in rubberized sweat suit
−5
Wingate Anaerobic Test
↓ (21%) anaerobic power
↓ (10%) anaerobic capacity
Fritzsche et al., 2000
8 men
Heat, 35°C, 25% RH
−4
Inertial load, cycling
↓ (4%)
a RH = relative humidity.
b NC = no change.
No differences in PR or RPE among trials advantages conferred by high aerobic fitness and heat acclimatization (Buskirk et al., 1958; Sawka et al., 1983b). Women and men who are of comparable physical fitness and heat acclimatization status appear to respond similarly to dehydration and exercise-heat stress (Sawka et al., 1983b).
The elevated core temperature responses to dehydration result from a decrease in heat loss (Sawka and Coyle, 1999). The relative contributions of evaporative and dry heat loss during exercise depend upon the specific environmental conditions, but both avenues of heat loss are adversely affected by dehydration. Local sweating (Fortney et al., 1981, Montain et al., 1995) and skin blood flow (Fortney et al., 1984; Kenney et al., 1990) responses are both reduced for a given core temperature when a person is dehydrated. Whole-body sweating is usually either reduced or unchanged during exercise at a given metabolic rate in the heat (Sawka and Coyle, 1999). However, even when dehydration is associated with no change in whole-body sweating rate, core temperature is usually elevated; therefore, the whole-body sweating rate for a given core temperature is lower when a person is dehydrated (Sawka et al., 1984b).
Both the singular and combined effects of plasma hyperosmolality and hypovolemia have been demonstrated as mediating the reduced heat loss response during exercise-heat stress (Sawka, 1992).
Dehydration reduces a person’s ability to tolerate exercise-heat stress. In experiments in the desert during 1942 and 1943, male soldiers serving as subjects attempted endurance (2 to 23 h) walks and were either allowed to drink water ad libitum or had to refrain from drinking (Brown, 1947c). One out of 59 (2 percent) subjects suffered exhaustion from heat strain during a desert walk when they were allowed to drink, whereas 11 of 70 (16 percent) subjects suffered exhaustion when they did not drink. In another study, “hyperacclimatized” subjects attempted a 140-minute walk in a hot environment while ingesting different combinations of salt and water (Ladell, 1955). Exhaustion from heat strain occurred in 9 of 12 (75 percent) subjects when receiving neither water or salt, and 3 of 41 (7 percent) subjects when receiving only water. More recently, normal subjects acclimated to heat attempted 140-minute treadmill walks in a hot-dry environment when euhydrated and when dehydrated by 3, 5, and 7 percent of body weight (Sawka et al., 1985). All eight subjects completed the euhydration and 3 percent dehydration experiments, while seven subjects completed the 5 percent dehydration experiments. During the 7 percent dehydration experiments, six subjects discontinued after completing an average of 64 minutes.
To address whether dehydration alters physiologic tolerance to heat strain, subjects walked to voluntary exhaustion when either euhydrated or dehydrated (8 percent of total body water) during uncompensable heat stress (Sawka et al., 1992). Dehydration reduced tolerance time from 121 to 55 min, but more importantly, dehydration reduced the core temperature that a person could tolerate. Heat exhaustion occurred at a core temperature about 0.4°C lower when dehydrated than when euhydrated.
Hyperhydration and Heat Strain
Because water deficits impair thermoregulation (e.g., body temperature increases), a logical question is whether greater-than-normal body water (hyperhydration) could improve a person’s ability to thermoregulate during exercise in the heat. Many studies have examined hyperhydration effects on thermoregulation in the heat. Some investigators report lower core temperatures during exercise after hyperhydration (Gisolfi and Copping, 1974; Grucza et al., 1987; Moroff and Bass, 1965; Nielsen, 1974; Nielsen et al., 1971), while other studies do not (Blyth and Burt, 1961; Candas et al., 1988; Greenleaf and Castle, 1971; Latzka et al., 1997, 1998; Montner et al., 1996; Nadel et al., 1980).
TABLE 4-13 Dehydration Effects on Muscular Strength and Endurance
Reference
Subjects
Dehydration Process
Tuttle, 1943
13
Exercise and heat
Ahlman and Karvonen, 1961
32 men
Sauna or exercise
Saltin, 1964
10 men
Heat and exercise
Greenleaf et al., 1966
9 men
Fluid restriction
Bosco et al., 1968
9 men
Fluid restriction
Singer and Weiss, 1968
10
Fluid restriction
Bosco et al., 1974
21 men
Fluid restriction
Torranin et al., 1979
20 men
Sauna
Bijlani and Sharma, 1980
14 men
Hot room
Houston et al., 1981
4 men
Fluid restriction
Mnatzakanian and Vaccaro, 1982
Not reported
Not reported
Serfass et al., 1984
11
Fluid restriction
Webster et al., 1990
7 men
Exercise in rubberized sweat suit
Greiwe et al., 1998
7 men
Sauna
Montain et al., 1998b
5 men
5 women
Exercise and hot room
a NC = no change.
Some investigators report higher sweating rates with hyperhydration (Lyons et al., 1990; Moroff and Bass, 1965), while other studies do not (Blyth and Burt, 1961; Candas et al., 1988; Greenleaf and Castle, 1971; Latzka et al., 1997, 1998; Montner et al., 1996).
However, most of these studies have serious design problems, such as control conditions representing dehydration but not euhydration (Candas et al., 1988; Moroff and Bass, 1965), control conditions not adequately described (Grucza et al., 1987; Nielsen, 1974; Nielsen et al., 1971), and cool fluid ingestion that might have caused reduced core temperature (Gisolfi and Copping, 1974; Moroff and Bass, 1965). No studies were found that examined the influence of gender on thermoregulatory responses to hyperhydration.
Δ Wt
Strength Method
Resulta
−5%
Isometric
NC in strength
−2 kg
Isokinetic
NC in strength
−4%
Isometric
NC in strength
−7%
Isometic
NC in strength with up to 4% dehydration
−3%
Isometric
↓ (11%) in strength
−7%
Isometric
NC in strength
−6%
Isometric
↓ (10%) in strength
↓ (9%) in endurance
−4%
Isometric
↓ (31%) in endurance
Isotonic
↓ (29%) in endurance
−3%
Isometric
↓ in endurance
−8%
Isokinetic
↓ (11%) in strength
−4%
Isokinetic
NC in strength
NC in endurance
−5%
Isometric
NC in strength
NC in endurance
−5%
Isokinetic
NC in leg strength
↓ (5%) in shoulder strength
↓ (4%) in chest strength
−4%
Isometric
NC in strength
NC in endurance
−4%
Isometric
NC in strength
Generally, the “best” designed studies did not report any thermoregulatory benefits from hyperhydration relative to euhydration (Greenleaf and Castle, 1971; Latzka et al., 1997, 1998; Nadel et al., 1980).
Hyperhydration and Performance
Several studies have examined whether hyperhydration improves exercise performance or heat tolerance. Blyth and Burt (1961) were the first to report the effects of hyperhydration on performance during exercise-heat stress. Their subjects ran to exhaustion in a hot climate when normally hydrated, as well as when hyperhydrated by drinking 2 L of fluid 30 minutes prior to exercise. When hyperhydrated, 13 of 18 subjects ran longer to exhaustion compared with their time to exhaustion when normally hydrated.
FIGURE 4-14 Relationship for elevation of core temperature (above that present with euhydration) at a given magnitude of water deficit during exercise conditions in different environments. VO2 is maximal oxygen uptake. Adapted with permission from Sawka (1992). Copyright 1992 by Lippincott, Williams and Wilkins.
The average time to exhaustion when hyperhydrated versus normally hydrated (17.3 versus 16.9 minutes) did not, however, reach statistical significance. In another study, subjects exercised to exhaustion during uncompensable exercise-heat stress when initially euhydrated (control) or hyperhydrated (increased total body water by approximately 1.5 L) (Latzka et al., 1998). Water hyperhydration did not extend endurance time beyond that seen in the control (euhydrated) condition in this study.
Dehydration and Cardiovascular Function
Dehydration increases resting heart rate when standing or lying down in temperate conditions (Rothstein and Towbin, 1947). In addition, dehydration makes it more difficult to maintain blood pressure during exposure to various perturbations. Dehydration induces fainting in individuals susceptible to postural fainting when tilted with feet down (Harrison et al., 1986; Rothstein and Towbin, 1947). Figure 4-15 presents data on a subject who was tilted with feet held downward for 10 min or until becoming unconscious (Rothstein and Towbin, 1947).
FIGURE 4-15 Relationship between body water deficit, heart rate (solid line), and fainting time (broken line) for a passively tilted subject.
SOURCE: Rothstein and Towbin (1947). Reprinted with permission from the Papers of Edward Adolph collection at the Edward G. Miner Library, University of Rochester Medical Center.
With increased levels of dehydration, the pulse rate increment increased and the time to faint decreased. Mild dehydration was recently shown to blunt baroreceptor control during an orthostatic tolerance test (Charkoudian et al., 2003), which may be an explanation for orthostatic intolerance (e.g., fainting upon standing) when an individual is dehydrated (≈ 1.6 percent of body weight). In addition, drinking water (0.5 L versus 0.05 L) markedly improved orthostatic tolerance in healthy men and women (Schroeder et al., 2002). The improved orthostatic tolerance could be mediated by plasma volume expansion or by the act of drinking resulting in increased sympathetic activation (Scott et al., 2001).
The effects of dehydration on cardiovascular responses to exercise have been investigated (Gonzalez-Alonso et al., 1997; Montain et al., 1998a; Nadel et al., 1980; Rothstein and Towbin, 1947; Sawka et al., 1979, 1985). Dehydration will increase heart rate in proportion to the magnitude of water deficit (Montain and Coyle, 1992; Montain et al., 1998a; Rothstein and Towbin, 1947; Sawka et al., 1985). Dehydration-mediated hypovolemia reduces central venous pressure (Morimoto, 1990) and cardiac filling (Coyle, 1998) and requires a compensatory increase in heart rate. During submaximal exercise with little heat strain, dehydration elicits an increase in heart rate and a decrease in stroke volume, and usually no change in cardiac output relative to euhydration levels. Heat stress and dehydration, however, have additive effects on increasing cardiovascular strain. During submaximal exercise with moderate (Nadel et al., 1980) or severe (Gonzalez-Alonso et al., 1997; Sawka et al., 1979) heat strain, dehydration (3 to 4 percent body weight) led to a decrease in cardiac output (compared with performing the exercise task when euhydrated) because the increase in heart rate was not of sufficient magnitude to compensate for the decline in stroke volume. The dehydration-mediated reduction in cardiac output (below euhydration levels) during heat stress was greater during high intensity (65 percent VO2max) than low intensity (25 percent VO2max) exercise (Montain et al., 1998a). In addition, severe water deficits (7 percent of body weight) in the absence of heat strain also reduced cardiac output during submaximal exercise (Sproles et al., 1976).
Death
For obvious reasons, experimental data are not available on the effects of dehydration with death as an outcome in humans. As discussed earlier, fever is a common response to inflammation, infection, and trauma and may be augmented by dehydration (Morimoto et al., 1986; Watanabe et al., 2000). Furthermore, dehydration increases cardiovascular strain. It is suggested that dehydration might contribute to the death of hospitalized patients who are ill (Weinberg et al., 1994a).
Humans can lose 10 percent of body weight as water and have little increased risk of death unless the dehydration is accompanied by other severe stressors (Adolph, 1947a). Reports from persons in survival situations indicate that those who dehydrated to greater than 10 percent of their body weight required medical assistance to recover (Adolph, 1947a).
Experimental studies regarding dehydration and death in animals have been performed (Adolph, 1947a; Keith, 1924; Wierzuchowski, 1936). When investigators infused sugar solutions to dehydrate dogs (Keith, 1924; Wierzuchowski, 1936), most could tolerate 7 to 10 percent dehydration; however, beyond this point body temperature rose rapidly and often led to death.
Adolph (1947a) reported on experiments in which dogs were slowly dehydrated by water deprivation in temperate conditions and were then exposed to heat stress. When the dogs were dehydrated by 10 to 14 percent of body weight and exposed to heat, their core temperature “explosively increased,” and they would only survive if removed from the heat stress or given water to drink (Adolph, 1947a). Deaths began as core temperatures approached 41.6°C (107°F) and would always occur when core temperatures reached 42.8°C (109°F). Lethal core temperatures were similar in the dehydrated and euhydrated dogs (Adolph, 1947a). Cats showed similar responses, but with water deficits of up to 20 percent body weight loss and core temperatures of up to 43°C (110°F) before dying.
There are many reports from civilian and military communities of persons being stranded in very hot conditions (such as desert conditions in the summer) for extended durations in which those who had water survived and those without water died. Dehydration is believed to contribute to life-threatening heat stroke. In view of physiological changes (e.g., elevated body temperatures and reduced tissue perfusion from inadequate cardiac output), this presumed association is reasonable (Bouchama and Knochel, 2002). Dehydration contributed significantly to an outbreak of serious heat illness of Massachusetts State Police recruits who had limited water availability during summer training sessions. Eleven of a class of 50 had serious rhabdomyolysis and/or heat injury and were hospitalized—two underwent kidney dialysis and one required a liver transplant and later died (Commonwealth of Massachusetts, 1988). In 1987, three collegiate wrestlers died of cardiorespiratory arrest while undergoing severe and rapid weight loss combined with stressful exercise in the heat (Remick et al., 1998). Dehydration was implicated in these three deaths; however, those athletes appeared to be employing exercise-heat dehydration procedures that were similar to those used by other interscholastic and collegiate wrestlers. Since these were the first deaths since record keeping was initiated in 1982, it is probable that some other unknown factor may have contributed. Thus dehydration is a serious health risk, particularly when associated with febrile illness or extreme heat and exercise.
Urinary Tract Infections
Dehydration may increase the risk of infections. Hydration monitoring was assessed to determine if it would encourage individuals to increase fluid intake and thus decrease their risk for urinary tract infections (Eckford et al., 1995). Twenty-eight premenopausal women who had at least two idiopathic urinary tract infections within 6 months of the study were taught to use a simple hand-held probe (a conductivity meter) to assess their urine osmolality (Eckford et al., 1995). Although this 4-month study was only completed by 17 of the 28 women, these women increased hydration and significantly decreased their incidence of urinary tract infection due to their greater consumption of fluids. In another study of over 300 subjects, increased fluid intake resulted in a lower rate of urinary tract infections (Pitt, 1989). While it cannot be assumed that urinary tract infections are the result of dehydration, adequate hydration may contribute to the prevention of such infections in humans (Hooton, 1995).
However, the utility of using the prevention of urinary tract infections as an indicator of adequacy is not adequately established on a quantitative basis to be used as the criterion on which to base recommended intakes of total water.
Dehydration and Chronic Diseases
Kidney Stones
Increased fluid intake has been found to be inversely associated with an increased risk of developing kidney stones (Curhan et al., 1997, 1998), and increased fluid consumption has long been suggested as means to prevent recurrence of kidney stones (nephrolithiasis). As a result of increased urine flow, the urinary concentrations of calcium, oxalate, phosphorus, and uric acid fall, thereby reducing the degree of saturation of their salts, which leads to the formation of kidney stones. Most of the available studies have been conducted on individuals who have already had stones, with the goal of preventing recurrence.
One of the first studies to evaluate the therapeutic effects of increased fluid intake was a retrospective case-series study (Hosking et al., 1983). One-hundred eight patients (83 men and 25 women) who had idiopathic calcium nephrolithiasis had been advised to increase their fluid intake to achieve a 24-hour urinary output greater than or equal to 2.5 L. Over an average follow-up period of 5 years, 58 percent of these patients had no evidence of stone growth or new stone formation (Hosking et al., 1983). In another case-series, 98 individuals (87 men, 11 women), all of whom were diagnosed as having been chronically dehydrated due to either defined history of exposure to heat due to climate or occupation or due to poor fluid intake, were asked to increase fluid intake to about 2.5 L/day (Embon et al., 1990). Resulting mean urinary volume increased from 1.7 to 2.5 L/day based on periodic random sampling during the follow-up period. After more than 4 years, the stone recurrence rate was approximately 7 percent (7/98), which was comparatively low (Embon et al., 1990). One randomized controlled trial with 5 years of follow-up tested the effects of increased water intake as a means of preventing recurrent kidney stones in 199 individuals (134 men and 65 women) with idiopathic calcium nephrolithiasis (Borghi et al., 1996). At baseline, estimated 24-hour urine volume was approximately 1 L. During the fifth year of follow-up, 24-hour urine volume remained unchanged in the control group but increased to 2.6 L in the active treatment group. Over the course of follow-up, recurrent stones occurred in 27 percent of control participants, but in just 12 percent of those in the active treatment group (p = 0.008) (Borghi et al., 1996).
More recent evidence suggests that increased fluid intake may prevent the initial occurrence of kidney stones; however, data are limited. Two prospective observational studies have assessed the relationship of fluid intake with incident kidney stones, while another study assessed the relationship of specific beverages. In a study of 45,619 male health professionals without kidney stones, the adjusted relative risk of developing a stone during 4 years of follow-up was 0.71 (95 percent confidence interval [CI]: 0.52 to 0.97) comparing the highest and lowest quintiles of fluid intake (> 2.5 versus < 1.3 L/ day) (Curhan et al., 1993). Similar findings were evident in a subsequent study of 91,731 female nurses without kidney stones; the adjusted relative risk of developing a stone during 12 years of follow-up was 0.61 (95 percent CI: 0.48 to 0.78) comparing the highest quintile of fluid intake (median intake of 4.7 L/day) to the lowest quintile of fluid intake (median intake of 1.9 L/day) (Curhan et al., 1997). Because the principal objective of both studies was to assess the relationship of dietary calcium intake with kidney stones, there were few analyses on the effects of fluid consumption. Subsequent reports on the beverages consumed (Curhan et al., 1996, 1998) provide data on intake of specific beverages, as well as intake of foods. A third prospective study (Hirvonen et al., 1999), which did not collect data on drinking water and total fluid intake, did report an inverse association of beer consumption with incident kidney stones.
Overall, available evidence, including the results of one clinical trial, strongly suggests that increased total water consumption can be effective therapy to prevent recurrent kidney stones. There is also some evidence from observational studies that increased fluid intake lowers the risk of incident kidney stones. However, this limited evidence is insufficient to set requirements for water intake as a means to prevent kidney stones.
Gallstones
Water ingestion has been shown to induce gallbladder emptying (Yamamura et al., 1988) via vagal stimulation (Svenberg et al., 1985). An association of gallstone formation (cholelithiasis) with low fluid consumption was suggested in a small group of patients (n = 30) with gallstones whose typical daily drinking water intake was estimated to be 0.4 to 0.7 L/day (Math et al., 1986). Subsequently, six individuals, one of whom had gallstones, were evaluated for the effect of water consumption on gallbladder emptying time. They consumed 0.5 L of water rapidly following an overnight fast; this resulted in gallbladder emptying within 10 to 20 minutes for those without gallstones, and 30 minutes for the patient with gallstones. It was concluded that a high daily water intake and consumption of water at regular intervals could assist with promotion of gallbladder emptying, and perhaps prevent gallstone formation (Math et al., 1986). While not tested, other beverages may have a similar effect.
Bladder, Colon, and Other Cancers
The relationship between colon cancer and total water intake has been evaluated, primarily in case-control studies. An early study that reported an inverse relationship between water consumption and colon cancer risk compared 238 men and 186 women with colon cancer to 224 men and 190 women who served as controls (Shannon et al., 1996). In the men studied, consumption of more than four glasses of water/day (~ 0.9 L) in addition to food versus one or fewer glasses/day was marginally associated with decreased colon cancer risk (odds ratio [OR] = 0.68; p = 0.16). In women, more than five glasses of water/day (~ 1.2 L) were associated with decreased colon cancer risk (OR = 0.55; p = 0.004). In another study, a fluid intake of greater than approximately 1.7 L/day was significantly associated with a decreased risk of colorectal adenoma (OR = 0.4; p < 0.01) (Lubin et al., 1997). Water intake levels of greater than six cups (1.4 L)/day have been reported to be protective for distal colon tumors (OR = 0.68) (Slattery et al., 1999).
Bladder cancer risk may also be reduced with increased fluid consumption. Although decreased bladder cancer risk with increased fluid intake has been reported, available studies did not all focus solely on fluid intake and bladder cancer risk (Bitterman et al., 1991; Braver et al., 1987; Pohlabeln et al., 1999; Wilkens et al., 1996). The strongest study to show a clear relationship between fluid intake and bladder cancer risk assessed the total daily fluid intake of 47,909 men (Michaud et al., 1999). Individuals who consumed greater than approximately 2.5 L/day of fluid were reported to have a 49 percent lower risk of bladder cancer than individuals who consumed less than approximately 1.3 L/day. It was also noted that the risk of bladder cancer was reduced by 7 percent for every addition of 240 mL (~1 cup) in daily fluid intake. However, several other studies have failed to demonstrate an overall association between fluid intake and bladder cancer risk (Bruemmer et al., 1997; Geoffroy-Perez and Cordier, 2001; Slattery et al., 1988).
Arrhythmias
One study has reported electrocardiogram (ECG) changes associated with varying levels of water deficit (Sawka et al., 1985). ECG abnormalities (arrhythmias and premature ventricular contractions) during exercise in the heat in healthy young adults who were dehydrated at 5 percent or greater of body weight loss were assessed (Sawka et al., 1985). All eight subjects completed 140 min of exercise without any ECG abnormalities when euhydrated or when dehydrated by 3 percent of body weight. Numerous premature ventricular contractions during exercise-heat trials at 5 and 7 percent dehydration were seen on the remaining subjects.
In another report, three collegiate wrestlers died of cardiorespiratory arrest while undergoing severe and rapid weight loss combined with stressful exercise in the heat (Remick et al., 1998). Because neither cardiorespiratory arrest nor heat injury/stroke had been previously reported with the rapid and severe dehydration procedures used in scholastic or collegiate wrestling, and because these deaths occurred over a short period of time, perhaps an unknown factor may have contributed. However, it is possible that the fluid-electrolyte imbalances resulting from marked dehydration, particularly if combined with stressful exercise, may contribute to ECG abnormalities in some individuals.
Ingestion of cold fluids has been thought to induce cardiac arrhythmias. However, the research in this area is equivocal. Electrocardiogram (ECG) changes after consumption of ice-cold beverages in healthy individuals without known cardiac or gastrointestinal problems were assessed (Pratte et al., 1973). In this controlled study, after ingestion of cold water there were significant changes in the ST segment. These segment changes were greater with larger volumes of cold water ingestion. Conversely, significant ECG changes (using a Holter monitor) were not seen in individuals who consumed iced fluids in another study (Haughey, 1990). Hence, available data on the effects of cold fluid ingestion as a risk factor for arrhythmia are sparse and inconsistent.
Blood Clots
Few studies have been conducted on the effects of fluid intake on factors that may increase blood clots. In one study, water and an electrolyte-carbohydrate beverage were compared to assess which would maintain hydration and decrease blood viscosity during a 9-hour plane flight (Hamada et al., 2002). Forty healthy men (mean age 23 years) were given approximately 1.3 L of either an electrolyte-carbohydrate beverage or water in five servings during the long flight. Compared with the water group, the men given the electrolyte-carbohydrate beverage gained more body weight, had lower urine output, and had improved net fluid balance. In addition, those who consumed the electrolyte-carbohydrate beverage had less viscous blood than those who drank water. Based on this one study, it appears that on long flights the concomitant consumption of fluid and solute may be more suitable to maintain hydration status and decrease blood viscosity than water alone; however, additional studies are needed to validate this effect.
Mitral Valve Prolapse
The effects of dehydration on mitral valve prolapse (MVP) have been evaluated in order to assess if dehydration could be used as a diagnostic tool for MVP (Lax et al., 1992). MVP, or symptoms associated with it, was induced by mild dehydration and, upon rehydration, the symptoms disappeared (Aufderheide et al., 1994; Lax et al., 1992). A lower atrial filling pressure and volume would result in a floppy valve balloon (prolapse more). It has been recommended that hydration status should be considered if a person with MVP is suspected of having atypical chest pain or palpitations (Aufderheide et al., 1994; Lax et al., 1992).
Osteoporosis
Longitudinal research on the effects of fluid intake on bone mineral density and osteoporosis has not been conducted. However, some short-term studies evaluating bone mineral density changes due to hydration status or the type of ingested fluids are available. The extent to which drinking various amounts of fluids between meals, and the meals themselves, affected body composition and bone mineral density were assessed in healthy individuals or individuals undergoing hemodialysis (Horber et al., 1992). No changes in bone mineral density as a result of the meals or hydration status were detected.
In a subsequent study, the calcium content of the water or beverage may have a greater impact on bone mineral density than the amount of fluids in terms of volume consumed (Costi et al., 1999).
https://www.nap.edu/read/10925/chapter/6#81
No comments:
Post a Comment