and to add to the confusion...
Evidence-Based Approach to Lingering Hydration Questions
Matthew S. Ganio, MS
Douglas J. Casa, PhD, ATC #8727;
Lawrence E. Armstrong, PhD
Carl M. Maresh, PhD
--------------------------------------------------------------------------------
Human Performance Laboratory, Department of Kinesiology, University of Connecticut, 2095 Hillside Road, U-1110, Storrs, CT 06269-1110, USA
--------------------------------------------------------------------------------
* Corresponding author.
--------------------------------------------------------------------------------
E-mail address:
douglas.casa@uconn.edu PII S0278-5919(06)00071-8
--------------------------------------------------------------------------------
Studies related to fundamental hydration issues have required clinicians to re-examine certain practices and concepts. The ingestion of substances such as creatine, caffeine, and glycerol has been questioned in regards to safety and hydration status. Reports of overdrinking (hyponatremia) also have brought into question the practices of drinking appropriate fluid amounts and the role that fluid-electrolyte balance has in the etiology of heat illnesses such as heat cramps. This article offers a fresh perspective on timely topics related to hydration, fluid balance, and exercise in the heat.
Core Temperature and Hydration
Proper hydration is important for optimal sport performance [1] and may play a role in the prevention of heat illnesses [2]. Dehydration increases cardiovascular strain and increases core temperature (Tc) to levels higher than in a state of euhydration [3]. These increases, independently [4] and in combination [3], [5], impair performance and put an individual at risk for heat illness [6]. Exercise in the heat in which dehydration occurs before [3] or during exercise [7] results in Tc that is directly correlated (r = 0.98) [7] with degree of dehydration (Fig. 1). The link between dehydration and hyperthermia has shown that independently and additively they result in cardiovascular instability that puts individuals at risk for heat exhaustion [3].
Fig. 1 The degree of dehydration that occurs during exercise is correlated with the increase in esophageal (top graph) and rectal (bottom graph) temperatures. Subjects cycled for 120 minutes in a 33ˇăC environment at approximately 65% VO2max while replacing 0% (No Fluid), 20% (Small Fluid), 48% (Moderate Fluid), or 81% (Large Fluid) of the fluid lost in sweat. Subjects lost 4.2%, 3.4%, 2.3%, and 1.1% body weight in the conditions. (From Montain SJ, Coyle EF. Influence of graded dehydration on hyperthermia and cardiovascular drift during exercise. J Appl Physiol 1992;73(4):1340¨C50; with permission.)
Despite laboratory evidence linking dehydration with increased Tc, some authors argue that this physiologic phenomenon does not occur in field settings [8], [9], [10]. This may be because field studies fail to control exercise intensity [8], [9], [10], [11]. Tc is driven by metabolic rate, and when the same subject is tested in a controlled laboratory environment, a higher metabolic rate produces a higher Tc[12]. Without controlling or measuring relative exercise intensity, a hydrated individual could exercise at a higher metabolic rate and drive his or her Tc to the same level as a dehydrated individual working at a lower intensity. Without a randomized crossover experimental design that controls exercise intensity, field studies cannot validly conclude that hydration is not linked to Tc.
Field studies disputing relationships between Tc and dehydration also cite that laboratory studies use environments that are too hot, and that the physiologic relationship does not exist in temperate environments (approximately 23ˇăC) often associated with field studies [8]. Laboratory studies have shown that the increase of Tc with dehydration is exacerbated in hot environments, but still observed in cold environments (8ˇăC) [13]. Dehydration impairs thermoregulation independent of ambient conditions, but the effect is seen especially at high ambient temperatures when the thermoregulatory system is more heavily stressed. Laboratory-based studies have clearly shown that when exercise intensity and hydration state are controlled, Tc increases at a faster rate when subjects are dehydrated [7].
Caffeine
Caffeine and its related compounds, theophylline and theobromine, have long been recognized as diuretic molecules [14], which encourage excretion of urine via increased blood flow to the kidneys [15]. The recommendation that caffeine be avoided by athletes because hydration status would be compromised [6] is based on several studies examining the acute effects of high levels (>300 mg) of caffeine [16]. More recent studies have tested the credibility of this recommendation by re-examining hydration status in varying settings after short-term caffeine intake and, for the first time, after long-term intake.
Using increased urine output as an indicator of diuresis and dehydration, early studies showed that the threshold for an increase of urine output was 250 to 300 mg of caffeine intake [17]. Urine output was greater for the first 3 hours after ingestion [17], but when urine was collected for 4 hours, the difference in urine output between caffeine and placebo was negated [18]. When double the caffeine was ingested (612 mg or 8.5 mg/kg), urine volume increased over the next 4 hours [19]. The molecular properties of caffeine do not refute the fact that it may act as an acute diuretic, but when observations span a short time (<24 hours), it is difficult to understand long-term changes in hydration [15].
When 24-hour urine volume is examined, the ingestion of caffeine at levels of 1.4 to 3.1 mg/kg does not increase urine output or change hydration status [20]. When large amounts of caffeine are ingested (8.2¨C10.2 mg/kg), the increases in urine excretion vary from person to person, but may be 41% greater than control levels [21]. It cannot be concluded from these studies that ˇ°caffeine causes dehydrationˇ± because acute increases in urine volume with large caffeine intake (>300 mg) may be offset later by decreased urine output and result in no change in long-term hydration status [16].
Acute ingestion of caffeine before exercise (1¨C2 hours) at levels up to 8.7 mg/kg does not alter urine output and fluid balance [19], [22], [23], [24] when subjects exercise at 60% to 85% VO2max for 0.5 to 3 hours [19], [22], [23], [24]. The possible mechanism for a lack of a diuretic effect with caffeine during exercise is most likely due to an increase in catecholamines and diminished renal blood flow [19]. There is little evidence to suggest that short-term use of caffeine alters hydration status at rest or during exercise.
Because most Americans consume caffeine on a regular basis [15], it is surprising that few studies have examined the effects of controlled caffeine intake over several days. In 2004, the authors' research team conducted a field study involving a crossover design in which subjects exercised for 2 hours, twice a day, for 3 consecutive days [25]. Subjects rehydrated ad libitum and consumed a volume equal to 7 cans daily of either caffeinated or decaffeinated soda. Throughout the 3 days, no differences of urine volume, body weight, plasma volume, and urine specific gravity were observed between the two conditions. The authors reported similar results in an investigation in which subjects consumed 3 mg caffeine/kg/d for 6 days; during the following 5 days, 20 subjects decreased their intake to 0 mg/kg/d, 20 maintained intake at 3 mg/kg/d, and 20 doubled their intake to 6 mg/kg/d [26]. Urine volume and other markers of hydration status showed that, regardless of caffeine ingestion, hydration status did not change throughout the 11 days (Fig. 2). Heat tolerance and thermoregulation examined on the 12th day during exercise in a hot environment did not differ between conditions [27].
Fig. 2 Controlled consumption of caffeine at a level of 3 mg/kg/d for 6 days and then decreased to 0 mg/kg/d (C0), maintained at 3 mg/kg/d (C3), or increased to 6 mg/kg/d (C6); none of these conditions altered hydration status. Urine osmolality (top graph) and volume (data not shown) during repeated 24-hour collection periods did not change over the course of the investigation. Acute urine (middle graph) and serum (bottom graph) osmolality also did not differ as a result of the level of caffeine consumption. (Data from Armstrong LE, Pumerantz AC, Roti MW, et al. Fluid, electrolyte, and renal indices of hydration during 11 days of controlled caffeine consumption. Int J Sport Nutr Exerc Metab 2005;15(3):252¨C65.)
Acute ingestion of moderate to low levels of caffeine (<300 mg) does not promote dehydration at rest or during exercise. Long-term ingestion of low to high levels of caffeine does not compromise hydration status and thermoregulation at rest and during exercise. Varying one's level of caffeine ingestion (either increasing or decreasing) also does not seem to change hydration status [15], [16]. There is no evidence to support caffeine restriction on the basis of impaired thermoregulation or changes of hydration status at levels less than 300¨C400 mg/d.
Hyponatremia
Hyponatremia has received attention in the media as a result of its occurrence in popular road running races [28]. Hyponatremia is a serious complication of low plasma sodium levels (<130 mEq/L) [29]. The exact cause is likely multi-faceted and circumstantial [30]. Hyponatremia has been observed in exercising individuals who became dehydrated [31], [32], maintained hydration [32], and became overhydrated [31], [32]. Asymptomatic hyponatremia is the most common type of hyponatremia [32] and is defined as a decrease in sodium level (<130 mEq/L) that occurs in the absence of life-threatening symptoms [33]. Asymptomatic hyponatremia per se is not harmful or detrimental to performance [34]. When plasma sodium decreases to less than 125 mEq/L, hyponatremic illness may occur. Hyponatremic illness is a medical emergency that is symptomatic and requires immediate medical treatment [32], [33], [35].
Overdrinking, identified as an increase in body mass, significantly increases one's risk for developing hyponatremia and should be avoided [32], [35], [36]. Some observational studies have found that increased dehydration results in higher sodium levels [31], [32], [37], but this does not mean that dehydration prevents hyponatremia. The increased risk of heat illnesses associated with dehydration does not warrant dehydration as a method for preventing hyponatremia. High sweat rates or sodium-concentrated sweat may lead to large losses of sodium and put one at risk for hyponatremia, especially in events lasting more than 3 hours [38]. It is recommended that one should ingest fluid at a rate that closely matches fluid loss (ie, ˇÜ2% body weight loss) [39].
Replacing large fluid losses with equal amounts of pure water may dilute the plasma sodium level [36], so it has been suggested that replacement of electrolytes can be achieved through sports drinks or salt tablets [30], [34]. Mathematical modeling has shown that in a variety of conditions the ingestion of sodium may attenuate the decline of serum sodium over time (Fig. 3) [40]. However, recent studies involving consumption of sodium through sports drinks and salt tablets have confirmed [30], [34], [41] and refuted [37], [42], [43] this relationship (Fig. 4). Some of these differences in results may lie in methodologic differences, [30] assumptions, and conflicting conclusions [44].
Fig. 3 Predicted effectiveness of a carbohydrate-electrolyte sports drink (CHO-E) containing 17 mEq/L of sodium and 5 mEq/L of potassium for attenuating the decline in plasma sodium concentration (mEq/L) expected for a 70-kg person drinking water at 800 mL/h when running 10 km/h in cool (18ˇăC; upper panel) and warm (28ˇăC; lower panel) environments. The solid shaded areas depict water loss that would be sufficient to diminish performance modestly and substantially. The hatched shaded area indicates the presence of hyponatremia. M indicates the finishing time for the marathon run. IT indicates the approximate finishing time for an ironman triathlon. For the sodium figures, the solid lines reflect the effect of drinking water only, and hatched lines illustrate the effect of consuming the same volume of a sports drink. The pair of lines of similar type represent the predicated outcomes when total body water accounts for 50% and 63% of body mass. BML, body mass loss. (From Montain SJ, Cheuvront SN, Sawka MN. Exercise associated hyponatraemia: quantitative analysis to understand the etiology. Br J Sports Med 2006;40(2):98¨C105; with permission.)
Fig. 4 Ingestion of a carbohydrate-electrolyte beverage (CE) slightly attenuated the decline of plasma sodium observed with ingestion of plain water (W) over 180 minutes of exercise at a moderate intensity in a hot environment (34ˇăC). (Adapted from Vrijens DM, Rehrer NJ. Sodium-free fluid ingestion decreases plasma sodium during exercise in the heat. J Appl Physiol 1999;86(6):1847¨C51; with permission.)
Understanding the etiology and cause of hyponatremia may help to understand its prevention better. It is well agreed that overconsumption of fluids is the primary, but not the only, cause [35], [40]. Whether replacement of sweat losses with equal volumes of sodium-containing beverages would prevent or attenuate hyponatremia is still debated [35]. More studies that look at varying environmental conditions, sweat rates, and body masses may help shed light on this complex picture. Some authorities have suggested that allowing dehydration would prevent hyponatremia because the contraction of extracellular fluid would increase sodium concentration. Until further studies are conducted, promoting dehydration (ie, >2% of pre-exercise weight) is not warranted and may put some individuals at greater risk for exertional heat illnesses and could compromise performance [2].
Creatine
Creatine is one of the most popular nutritional supplements on the market. Athletes of all levels and varieties of sports are using it in hopes of gaining a competitive edge. During creatine supplementation, 90% of the increase in body weight (0.7¨C2.0 kg) is accounted for by increases of total body water (TBW) [45]. The increase of TBW during the ˇ°loading phaseˇ± results from increases of intracellular water stores [46], but prolonged use of creatine results in TBW increases in all body fluid compartments [45]. Some authors speculate that creatine use while exercising in the heat impairs heat tolerance and may be a contributing factor for heatstroke [47], [48]. Those authors propose that creatine increases one's risk for heat injury because the increases of intracellular water stores deplete intravascular volume [49]. Before any published conclusive studies concerning creatine's effect on hydration status and use in the heat, the American College of Sports Medicine published a consensus statement stating that ˇ°high-dose creatine supplementation should be avoided during periods of increased thermal stress ˇ there are concerns about the possibility of altered fluid balance, and impaired sweating and thermoregulation ˇˇ± [48].
Paradoxically, studies using short-term and long-term creatine supplementation have shown that subjects exercising in the heat (30¨C37ˇăC) for 80 minutes have either no change or an advantageous lower heart rate and Tc[46], [50], [51], [52]. Work from our laboratory also has shown that creatine supplementation does not alter exercise heat tolerance, even when subjects begin exercise in a dehydrated state (Fig. 5) [51]. One study that found lower Tc with creatine use during exercise in heat suggests that the increases of TBW with supplementation may hyperhydrate the body and lower Tc[46]. Despite early concerns about creatine supplementation and exercise in the heat [48], more recent studies have shown conclusively that heat storage does not increase as a result of creatine use [46], [50], [51], [52]. There is no evidence to support restriction of creatine use during exercise in the heat.
Fig. 5 The use of creatine monohydrate (CrM) does not compromise exercise heat tolerance. After becoming dehydrated, rectal temperature and mean weighted skin temperature (MWST) had similar responses in CrM and placebo treatments when subjects exercised in the heat and recovered in a cool environment. (From Watson G, Casa D, Fiala KA, et al. Creatine use and exercise heat tolerance in dehydrated men. J Athl Train 2006;41(1):18¨C29; with permission.)
Exercise-associated Cramps
Although the exact mechanism is unknown, skeletal muscle cramps are associated with numerous congenital and acquired conditions, including hereditary disorders of carbohydrate and lipid metabolism, diseases of neuromuscular and endocrine origins, fluid and electrolyte deficits (ie, owing to diarrhea or vomiting), pharmacologic agents (ie, ¦Â-agonists, ethanol, diuretics), and toxins [53]. The medical treatments for these various forms of muscle cramps are as varied as their etiologies. McGee [54] specifically classified leg muscle cramps as contractures (ie, electrically silent cramps caused by myopathy or disease), tetany (ie, sensory plus motor unit hyperactivity), dystonia (ie, simultaneous contraction of agonist and antagonist muscles), or true cramps (ie, motor unit hyperactivity). The last category includes skeletal muscle cramps that are due to heat, fluid-electrolyte disturbances, hemodialysis, and medications.
The International Classification of Diseases [55] defines heat cramps, a form of motor unit hyperactivity, as painful involuntary contractions that are associated with large sweat (ie, water and sodium) losses. Heat cramps occur most often in active muscles (ie, thigh, calf, and abdominal) that have been challenged by a single prolonged event (ie, >2¨C4 hours) or during consecutive days of physical exertion. A high incidence of heat cramps occurs among tennis players [56], American football players [57], steel mill workers [58], and soldiers who deploy to hot environments [59], [60]. These activities result in a large sweat loss, consumption of hypotonic fluid or pure water, and a whole-body sodium and water imbalance [59], [61]. The distinctions between heat cramps and other forms of exercise-associated cramps are subtle [54], [59], [62], but sodium replacement usually resolves heat cramps effectively [56], [59], [61], [62], [63]; successful treatment via sodium administration confirms a preliminary diagnosis of heat cramps.
Bergeron [62] described a tennis player who was plagued by recurring heat cramps. This athlete secreted sweat at a rate of 2.5 L/h and had a sweat sodium (Na+) concentration of 83 mEq/L. This sweat Na+ concentration is high, in that most heat-acclimatized athletes exhibit 20 to 40 mEq Na+/L of sweat (ie, heat acclimatization reduces sweat Na+ concentration), but occurs naturally in a small percentage of humans. During 4 hours of tennis match play, this young athlete lost 10 L of sweat and a large quantity of electrolytes (ie, 830 mEq of Na+; 19,090 mg of Na+; 48.6 g of sodium chloride). Given that the average sodium chloride intake of adults in the United States is 8.7 g (3.4 g Na+) per day, it is not difficult to see how this athlete could experience a whole-body Na+ deficit. To offset his 4-hour sodium chloride loss in sweat, this athlete would require 1.6 L of normal saline, 7.8 to 9.8 cans of canned soup (85¨C107 mEq per can), 12.6 servings of tomato juice (66 mEq of Na+ per serving), or 39.5 to 127.7 L of a sport drink (6.5¨C21 mEq Na+/L). These options are unreasonable. A long history of heat cramps ended when this tennis player began consuming supplemental salt during meals. Other tennis players have been successfully treated using a similar course of action [63].
In 2004, the authors' research team evaluated a female varsity basketball player (body mass 78.5 kg, height 187 cm) who experienced exercise-induced cramps during the winter months in New England, with signs and symptoms identical to heat cramps. The authors measured her sweat rate as 1.16 L/h, her sweat sodium concentration (ie, via whole-body washdown) as 42 mEq/L, and her daily consumption of sodium. These values were normal and typical of winter sport athletes. Three days of observations indicated that her dietary intake of Na+ per day was similar to her daily sweat Na+ loss (ie, both 3200¨C3600 mg). Because she did not train or compete in a hot environment, the authors hesitated to diagnose her malady as heat cramps. When she began ingesting supplemental sodium (ie, by liberally salting each meal at midseason), however, the skeletal muscle cramps resolved permanently. This case suggests that a history of skeletal muscle cramps, with a large daily Na+ turnover owing to a high sweat rate, indicates the need for an evaluation of whole-body Na+ balance. It further suggests that heat cramps may have been named because they usually occur in hot environments, but they also may occur in mild environments when sweat Na+ concentration and sweat losses are large.
A study by Stofan and colleagues [57] examined the link between sweat sodium losses and heat cramps. Sweat rate, sodium content, and percent body weight loss were measured on a single day of a ˇ°two-a-dayˇ± practice in subjects who had a history (episode within the last year) of severe heat cramps. Although heat cramps were not observed, football players with a history of heat cramps had sweat sodium losses two times greater than matched controls. Although the exact etiology of heat cramps may be unknown, sodium deficits seem to contribute to their development. In most cases, restoration and compensation of sodium losses seems to prevent further heat cramps.
Fluid Needs and Hydration Plan
Water losses during exercise should be replaced at a rate equal to (not greater than) the sweat rate [39]. Loss of sweat during exercise needs to be replaced after exercise, but dehydration (ˇÝ2% body weight) during exercise can be detrimental to performance in part by increases in Tc. It is difficult to replace 100% of fluid loss during exercise, especially if it occurs in hot environments for long durations or if sweat loss is great [11], [39]. Authorities have suggested that a minimal amount of dehydration (<2% body weight) may be tolerated without compromising performance [64]. Regardless, knowledge of sweat rate is necessary to develop a hydration plan (Table 1) [65], but without this it has been recommended to ingest 200 to 300 mL every 10 to 20 minutes [6]. Thirst lags behind changes in hydration (termed voluntary dehydration) [66]. When individuals have high sweat rates, and large volumes of fluid cause gastrointestinal stress, it may be advantageous for them to train themselves to tolerate consumption of fluids at a rate similar to their sweat losses [67].
Table 1 -- Self-testing program for optimal hydration[#8727;]
1. Make sure you are properly hydrated before the workoutˇŞyour urine should be pale yellow
2. Do a warmup run until you begin to sweat, then stop. Urinate if necessary
3. Weigh yourself naked on a floor scale (accurate to 0.1 kg)
4. Run for 1 h at an intensity similar to your targeted race or training run
5. Drink a measured amount of a beverage during the run, if you are thirsty. It is important that you measure exactly how much fluid you consume during the run
6. Do not urinate until post-body weight is recorded
7. Weigh yourself naked again on the same scale you used in step 3
8. You may now urinate and drink fluids as needed. Calculate your fluid need using the following formula
A. Enter your body weight from step 3 in Kg (To convert from lb to kg, divide lb by 2.2) _____________
B. Enter your body weight from step 7 in Kg (To convert from lb to kg, divide lb by 2.2) ______________
C. Subtract B from A A-B = ______________
D. Convert your total in step C to g by multiplying by 1000 C ˇÁ 1000 = ______________
E. Enter the amount of fluid consumed during the run in mL (To convert from oz to mL, multiply oz by 30) ______________
F. Add E to D E + D = ______________
This final figure is the number of ml that you need to consume per hour to remain well hydrated. If you want to convert mL back to oz, divide by 30
Adapted from Casa D. Proper hydration for distance runningˇŞidentifying individual fluid needs. Track Coach 2004;167:5321¨C8; with permission.
#8727; This table may be used to calculate the amount of fluid needed during an exercise bout to remain hydrated.
In attempts to optimize endurance performance in the heat, glycerol has been used to increase TBW. It is an osmotically active molecule that acutely (<4 hours) increases TBW stores [68]. Although using glycerol plus water is an effective prehydration strategy, it does not increase sweat rate or reduce performance time or Tc in a race setting [69]. Using glycerol as a part of a rehydration strategy between exercise bouts increases exercise time to exhaustion in the heat (37ˇăC). The increase is likely due to increases of plasma volume, not because of cardiovascular effects, thermoregulatory effects, or differences in fluid-regulating hormones [70]. It is generally accepted that glycerol, although a hyperhydrating agent, is not an ergogenic aid in most situations [64]. Future research should examine the importance of timing in glycerol ingestion for performance benefits.
When multiple, dehydrating exercise sessions are occurring over a short time (ie, two workouts per day in football or track and field), athletes must rehydrate immediately and quickly between bouts. Intravenous rehydration has been used in the belief that direct administration of fluid into the central circulation optimally replaces lost fluid. Contrary to this belief, when hydrating with equal amounts of intravenous and oral fluid ingestion, intravenous is not superior to restore plasma volume after dehydration [71]. Oral rehydration results in better cardiovascular stability, lower Tc, rating of perceived exertion, thirst, and thermal sensation than intravenous rehydration. However, these changes do not translate into improved exercise time to exhaustion [71], [72]. Regardless, oral hydration is preferred (versus intravenous) for individuals who would be exercising subsequently in the heat [71], [72]. An exception occurs when large amounts of fluid must be replaced in a short time, and gastric emptying and intestinal absorption rates may limit the ingestion of fluids orally. In such cases, a combination of intravenous and oral rehydration may be warranted so that fluid requirements are met, and the oropharyngeal reflex is stimulated [73].
Athletes often supplement with glycerol or choose to use intravenous rehydration because of the difficulty of matching fluid intake with fluid losses during intense exercise in the heat. This makes theoretical sense given the possibility of large sweat rates (ie, >1.5 L/h) and the likelihood that fluid consumption could not match the sweat rate given gastric emptying and intestinal absorption rates, especially when the ingestion must occur when the exercise is intense. An individualized rehydration plan that considers sweat rate, the semantics of the actual competition parameters, and personal preferences and tolerance is recommended to ensure that rehydration is optimized in these circumstances [65]. When the individualized rehydration plan is practiced and rehearsed in practices and preliminary competitions, the need for glycerol and intravenous rehydration will likely be eliminated because of the benefits associated with the ˇ°rehydration training,ˇ± and ultimately the degree of dehydration would be minimized [65], [74].
Summary
Hydration status affects exercise performance in the heat and may influence the development of exertional heat illnesses. However, numerous factors that influence hydration state are not understood by the public. Field-based studies may lead athletes to believe that Tc is not influenced by hydration, but these studies contradict well-controlled laboratory experiments. For many years, recommendations have been published that active individuals should avoid caffeinated beverages with little supporting scientific evidence. Research from the authors' laboratory shows that long-term intake of moderate levels of caffeine does not compromise hydration status. Hyponatremia also has received a lot of attention, but until more is known about its etiology and prevention, it is recommended that athletes drink an amount of fluid to minimize dehydration (but not overdrink). The use of creatine as an ergogenic aid initially was overshadowed by questions regarding its safety during exercise in the heat. Research shows no reason for these concerns. Although the mechanism of heat cramps is still not fully understood, it seems that deficits in sodium from sweating and/or diet is a predisposing factor. The reader is encouraged to read thorough review articles on these topics [64], [75]. Ultimately, clinical practice should be dictated by evidence in the literature and not perpetuate unproven myths.
