Rehydration during Endurance Exercise: Challenges, Research, Options, Methods

2.1. Effects of Dehydration on Endurance Exercise Performance

The negative influences of dehydration on exercise performance are recognized by professional sports medicine organizations [24,25,26,27] and sport governing bodies [28] in position statements regarding rehydration, exertional heat illness, and physical performance. Although the precise water deficit at which performance decrements occurs is difficult to determine because of inter-individual differences, there is overall consensus in the literature that dehydration of −2 to −4% represents the range in which endurance exercise performance declines [24,25,26,28,29,30,31]. This effect is illustrated in Figure 3, which presents a performance analysis of 34 published studies involving dehydration; a body water deficit of 1–3% was less likely to impair endurance exercise performance significantly (p < 0.05) than dehydration of 4–7% [32]. These 60 statistical comparisons involved outdoor track running, trail running, outdoor road cycling, indoor treadmill running, indoor rowing ergometry, and indoor cycle ergometry.

It also is relevant that controlled field studies have reported statistically significant decrements of endurance exercise performance at lower levels of dehydration (−1.0 to −2.2%). These controlled protocols evaluated running competitions at (a) three distances on an outdoor all-weather track [33] and (b) an outdoor cycling hill-climbing performance competition [34]. Similar performance decrements were reported for two indoor cycling simulations [35,36] when dehydration ranged from 1.0–1.8%.

Decrements of outdoor cycling performance which occur at ≤4% body mass loss have been disputed in one meta-analysis [37] which evaluated 5 research articles (13 effect estimates). However, virtually all researchers agree that moderate-to-severe dehydration (e.g., 4–7% versus 1–2% body mass loss; Figure 3) results in greater physiologic strain and decreased aerobic exercise performance [24,26,29,30,32,38,39,40,41].

Physiologists agree that dehydration-induced aerobic performance decrements are greater in hot versus cool environments [24,29,41,42], due to greater thermal and cardiovascular strain (i.e., increased skin temperature and blood flow, greater plasma volume loss, decreased cardiac output) that becomes physiologically impactful in the dehydration range of 2–4% of body mass [43,44]. In terms of associated physiological effects, a summary of 8 relevant publications [29] noted that a 4–5% dehydration level reduced maximal oxygen consumption (VO_2max) more in a hot environment (−9 to −27% when Tamb ≥ 30 °C) than in a cool or mild environment (−3 to −7% when Tamb was 15–26 °C). The consequences of this dehydration included a shorter exercise time to exhaustion, an obligatory reduction in exercise intensity, or both.

2.2. Effects of Dehydration on Symptomatology and Illness

The classic desert field observations of Adolph and colleagues [12] during the mid-1940s were among the first descriptions of exercise-induced dehydration that approached and exceeded 10% body mass loss. Although Adolph’s systematic observations of dehydrated soldiers walking in the desert (Tamb > 37 °C) were not subject to statistical analyses, they provide a thorough description of the detrimental health effects of prolonged dehydration. Figure 1 presents a summary of the signs and symptoms that were reported when dehydration exceeded 2%. The following Section 2.2.1 through Section 2.2.3 describe disorders that result from water and salt deficiencies.

3.1. Hyperhydration and Exercise Performance

No evidence suggests that deliberate pre-exercise consumption of excess pure water has an ergogenic effect on exercise performance [24,56]. Glycerol, however, often is ingested prior to exercise (e.g., 1.2 g/kg body mass with a volume of fluid equal to 26 mL/kg body mass) to hyperhydrate athletes by increasing water retention and plasma volume while decreasing urine volume [57]. This act delays reaching a level of dehydration that degrades exercise performance. Research results are not conclusive, in that studies have shown both an ergogenic effect and no effect on exercise performance [58,59]. However, one caveat should be noted. As glycerol dilutes both intracellular and extracellular fluids prior to exercise, it may predispose athletes to low serum Na⁺ as described in the next paragraph, especially if aggressive drinking occurs during exercise. A similar predisposition to low serum Na⁺ has been reported for deliberate overhydration with water and dilute fluids [24,60,61].

3.2. Exertional Hyponatremia (EHN): A Potential Medical Emergency

When overhydration during exercise dilutes blood, an osmotic pressure gradient causes water to move into cells. The resulting cell swelling can result in EHN, one of the few illnesses that is potentially fatal to otherwise healthy athletes during exercise. Recognizable symptoms appear in most athletes at a serum concentration of approximately 130–135 mmol Na⁺/L [26,62] and include lightheadedness, dizziness, nausea, puffiness (e.g., hands and feet), and body weight gain from baseline [63]. The majority of athletes whose serum Na⁺ is below 130 mmol/L (Figure 2) experience symptoms; these may include headache, vomiting, frothy sputum, difficulty breathing, pulmonary edema (i.e., fluid accumulation with swelling), and altered mental status such as confusion or seizure that results from cerebral edema [10,64].

The signs and symptoms of EHN do not necessarily correlate with the serum Na+ in the range shown in Table 3 (≥130 mmol Na⁺/L). The total symptoms score (column 10), rated with a validated Environmental Symptoms Questionnaire [65], was not related to the change of serum Na⁺ (column 5). Indeed, the self-rated symptoms of hyponatremic cyclists LC and AM (serum Na+ of 130 mmol/L) ranked among the lowest in this subject sample [66]. Thus, the severity of symptoms and not the absolute value of serum Na⁺ concentration guide the course of medical treatment [63].

Table 3 includes data regarding two recreational cyclists (LC and AM) who began a summer 164-km event (Tamb, 34 ± 5 °C) with normal serum electrolytes but finished the ride with a serum sodium concentration of 130 mmol/L, indicative of mild EHN. The data of 12 other finishers (A–L, column 1) are presented to allow comparisons. Although they did not ride together, both cyclists consumed a large and similar relative volume of fluid (191 and 189 mL/kg), experienced an identical 11 mmol/L decrease of serum sodium, and reported low thirst sensations. However, one (LC) gained 3.1 kg (+4.3% of body mass) during 8.9 h of exercise (i.e., suggesting a dilutional effect) and the other (AM) maintained body mass (+0.1 kg, +0.1%, 10.6 h), suggesting that no excess fluid was retained. Thus, Table 3 suggests a complex, individualized EHN etiology [66].

After exercise, fluids should be consumed judiciously because symptoms of EHN may develop hours after excess fluid consumption, as described in two published case reports. The first involved a 21-year old man who had aggressively consumed water and gained 5.25 kg of body weight during 5 h of treadmill exercise in a hot environment [61]. He was asymptomatic until he experienced nausea and malaise late that evening, and was transferred to a nearby hospital with a serum Na⁺ of 122 mmol/L. After a night of observation, fluid restriction, and large urine output, he was released at 11:00 a.m. the next morning without symptoms. A more serious EHN case with delayed symptom onset involved a 49-year old runner who finished a 42.2 km marathon in 4 h 22 min then boarded an airline flight to return home [67]. Approximately 5 h after he finished the race, he became ill and experienced a grand mal seizure in the aisle of the cabin. The pilots diverted the aircraft to a nearby city for an emergency landing. The runner was transported to a hospital, where he experienced two additional seizures while unconscious with a serum Na⁺ of 129 mmol/L. A chest x-ray indicated pulmonary edema and a brain scan revealed cerebral edema; he also was diagnosed with renal insufficiency and liver damage. During the ensuing 18 months, this runner learned that he was unable to mentally process information that had previously been routine, and he was unable to perform his professional duties.

4.1. Multiple Factors Influence Drinking during Endurance Exercise

Figure 4 presents several factors that influence thirst and drinking behavior, each of which is monitored and regulated continuously by the central nervous system. Thus, the phrase dynamic complexity applies to the vast, integrated, brain-wide network of nerve circuits and brain regions that regulate thirst, drinking, body water volume, and fluid concentration [92]. The findings of two recent human studies [93,94] suggest that thirst is one of multiple conscious perceptions and subconscious autonomic responses (Figure 4) that evolve simultaneously during dehydration and rehydration to influence drinking behavior. Simply stated, the internal motivation to consume water is influenced by multiple factors that reinforce the perception of thirst.

As dehydration concentrates extracellular fluid, plasma osmolality (P_osm) is recognized as the primary factor that stimulates thirst [2,81,95]. However, as shown in Figure 5, the P_osm threshold at which thirst is perceived varies greatly across individuals (range, 274–293 mOsm/kg) and may be lower than the range of laboratory reference values for P_osm (285–295 mOsm/kg) [96]. The data in Figure 5 were compiled from 5 published human studies; the shape of this frequency distribution implies that the thirst threshold is a multifactorial (polygenic) characteristic [97]. Although inter-individual differences of the P_osm threshold for thirst have not been studied during exercise, Figure 5 suggests that the level of dehydration (i.e., increased P_osm) that initiates drinking during exercise might differ considerably among athletes.

4.2. Inter-Individual Differences

Figure 6 illustrates six factors that influence the change of body mass and serum Na⁺. Interactions of these factors with inherited characteristics and endurance training regimens produce great differences among athletes. In Figure 2, for example, if a male athlete experienced a body mass change of only—0.5% (0.35 kg) during a 13 h Ironman triathlon, his serum Na⁺ could range from 119 (symptomatic EHN) to 157 mmol/L (severe hypernatremia). Conversely, if a male athlete finished with a serum Na⁺ of 140 mmol/L (i.e., in the center of the reference range for healthy adults), his body weight change could range from +5.5% (+3.9 kg) to −12.5% (−8.8 kg). These large serum Na⁺ and body mass ranges represent the effects of numerous perceptual, behavioral, hereditary and fluid-electrolyte variables (Figure 4, Figure 5 and Figure 6), but are influenced mostly by the total volume of sweat produced (e.g., 15.6 L/13 h at a sweat rate of 1.2 L/h) and the total volume of fluid consumed (e.g., 9.1 L/13 h at a drinking rate of 700 mL/h).

Large inter-individual differences also exist among elite runners. Fluid intake rates were determined for 10 men who placed 1st or 2nd (range of finish times, 2:03:59 to 2:10:55) during prestigious city marathons [98]. Each runner’s drinking behavior was recorded on videotape and fluid intake was estimated by multiplying drinking time by 45.2 mL/s (i.e., a value determined by laboratory drinking simulations). Half of these runners consumed fluids at rates that ranged from 30–300 mL/h (73–631 mL total), whereas the rate of fluid intake of the other 5 elite runners ranged from 420–1040 mL/h (886–2205 mL total). Clearly, these data indicate that elite marathon runners (a) ingest fluid at rates that span a wide range during competition (30–1040 mL/h), and (b) some individuals greatly exceed the proposed drinking rate threshold at which symptomatic EHN appears (700 mL/h), described above in Section 3.2.4.

4.3. Personal Beliefs and Sources of Rehydration Information

Personal beliefs about drinking may predispose an athlete to EHN, according to the findings of two case reports. The first [66] involved cyclists LC and AM in Table 3 (who also are designated as athletes H and J in Table 4). Their urine specific gravity measurements on Day-1 (LC, 1.006; AM, 1.004; 31 control cyclists, 1.017) and the morning of the road ride (LC, 1.003; AM, 1.005; controls, 1.019) indicated that both had overhydrated before prolonged exercise. The authors suggested that their pre-event hydration behavior, coupled with high rates of fluid intake during the ride (LC, 1500 mL/h; AM, 1400 mL/h), resulted in both cyclists experiencing asymptomatic EHN. The second publication [61] involved a controlled case report of EHN (serum Na⁺, 122 mmol/L) observed during 5 h of intermittent treadmill walking in a hot environment (Tamb, 41 °C). A physically fit, 21 year-old male began exercise with blood values (serum Na⁺, 134 mmol/L; osmolality, 282 mOsm/kg) slightly below the laboratory reference ranges for healthy males, because he had voluntarily overhydrated throughout the previous day. Overhydration prior to exercise is known to lower serum Na⁺ and therefore increase the risk of dilutional hyponatremia if fluids are aggressively consumed during exercise [24]. This man acknowledged that his pre-exercise overhydrated state, and his high rate of fluid intake during exercise (2061 mL/h; 10.31 L total), were purposeful. He believed that drinking a large volume of water would decrease his risk of exertional heat exhaustion and heatstroke.

To understand the factors that guide drinking behavior, Winger and colleagues [99] surveyed 106 female and 97 male runners who had competed in road races (average experience, 13.0 and 8.3 y, respectively) at distances ranging from 5 to 42.2 km. Seven years later, Wilson [100] surveyed 223 female and 199 male marathon runners (average number completed by both groups, 4). Both studies focused on runner perceptions of fluid replacement, their beliefs about rehydration during exercise, and drinking behaviors. Winger et al. [99] observed that the most important influences on the drinking behavior of runners were: trial and error/personal history, recommendations from running groups or clubs, and advice from friends; sport drink companies were the least influential sources of information. Among marathon runners, Wilson [100] reported that the following sources of information were considered to be the most reliable: social media, print magazines, a personal trainer, and a fellow runner; scientific journals, dietitians/nutritionists, and running coaches were rated as the least reliable sources of information.

4.4. Unique Characteristics of Competitive Events

Distinctive aspects of different sporting events also can influence an athlete’s rate of fluid intake; three examples follow. First, the possibility of inhaling water when respiratory ventilation rate is high (e.g., 120–155 L/min), and potentially having to stop to clear both lungs [21], may deter runners from drinking. Second, whereas fluid intake during endurance footraces is generally limited to the number of aid stations along a race course or a water bottle in hand, cyclists can consume fluids whenever desired because containers are attached to the bicycle frame or held in jersey pockets. Third, during the water stage of a triathlon, endurance swimmers have no access to fluids except the water they inadvertently consume during competition. These factors explain, in part, the consistent observation that endurance competitors replace ≤ 50% of sweat losses when allowed liquids ad libitum [101,102,103].

5.1. Options 1 and 2

The distinction between option 1 (using perceived thirst as the only signal to drink) and option 2 (consuming fluid whenever and in whatever volume desired) is subtle [101]. In fact, some professional organizations and experts have not recognized these as distinct behaviors and some authors use these terms synonymously [6,26,84,98,99,101]. However, a 2014 field study determined that cyclists could identify whether they typically used option 1 or 2 and self-selected into one of two study groups (n = 12 in each). Despite the fact that cyclists understood options 1 and 2 as distinct rehydration behaviors, no between-group differences were observed in the following measurements: total fluid intake (L), body mass change (%), time to ride 164 km (h), urine specific gravity, and rating of thirst at the finish line [110]. Thus, the greatest value of distinguishing options 1 and 2 may exist in communications among athletes.

During prolonged endurance exercise, the effectiveness of thirst and drinking in maintaining whole-body fluid–electrolyte balance is challenged [32] by large water and sodium losses in sweat, perturbations of intracellular volume or concentration, and plasma volume depletion [6]. As one example of several in the literature, Greenleaf et al. [91] reported that test subjects who drank ad libitum during exercise-heat stress, were not thirsty and felt fully recovered despite a body water deficit of 4–5 L. This phenomenon was named involuntary dehydration [6,12] because thirst was not sufficient to maintain body water and fluid intake did not match the level of dehydration. Voluntary dehydration is observed in most endurance athletes [17,102].

A field study involving 26 ultraendurance cyclists explored thirst during exercise [114]. Ratings of thirst were statistically compared to several other variables during 7 h of exercise (mean ground speed, 25.4 km/h; Tamb, 35.5 °C). The total fluid intake of these male cyclists varied greatly, ranging from 2.1 to 10.5 L during the 164 km ride. Post-event ratings of thirst were not significantly correlated with any measured variable, including total fluid intake (i.e., a measure of drinking behavior), body mass index, height, ground speed, body water balance (ingested fluid volume—volume of fluid lost), and change of body mass. In other words, the intensity of thirst did not represent the degree of dehydration or the volume of fluid consumed. This observation is consistent with the known effects of oropharyngeal sensations on drinking behavior. These sensory signals (e.g., mouth dryness, swallowing fluid) rapidly reduce and limit fluid intake by modulating satiety and opposing overdrinking [115]. Thus, whenever fluid is consumed, even if the volume is small, oropharyngeal signals diminish the sensation of thirst [93,115] and theoretically reduce the risk of EHN by reducing fluid intake.

5.2. Option 3

Individualized planned drinking appeals to athletes who prefer to design their exercise rehydration systematically. This approach to rehydration is recommended by the National Athletic Trainer’s Association [26] and the American College of Sports Medicine [24]. The accuracy of this method relies on an objective measurement of sweat rate during outdoor exercise. After sweat rate is determined, the athlete can design a customized fluid replacement plan for use during endurance exercise, realizing that the predetermined rate of fluid intake always should be less than sweat rate, to avoid overhydration and body weight gain.

5.3. Options 4 and 5

The practices of drinking as much as possible (i.e., increasing the risk of EHN) and drinking nothing (i.e., increasing the likelihood of a performance decrement) are discouraged. No professional sports medicine or sports nutrition organization recommends these extreme options during endurance exercise.