Oral Rehydration Beverages for Treating Exercise-Associated Dehydration: A Systematic Review, Part II. The Effectiveness of Alternatives to Carbohydrate-Electrolyte Drinks Open Access

This systematic review was carried out according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses checklist.²⁴  The search strategy is presented in part I of this review.¹⁵  The protocol was submitted at PROSPERO (CRD42020153077).

Screening of titles and abstracts, full texts of articles, and reference lists of included studies was described in part I of this systematic review.¹⁵  Studies were eligible if authors addressed the PICO question and met the inclusion criteria outlined in Table 1. In addition, we did not report on data from studies in which authors reported mean values but no SDs, effect sizes, and P values. Moreover, for most outcomes, we limited relevant time points to between 1 and 2 hours after completion of drinking, essentially as described previously by Borra et al.¹⁵  Finally, the systematic review was limited to rehydration drinks for which more than 1 study was identified.

Data extraction was described in part I of this study.¹⁵ 

The certainty assessment was described previously.¹⁵ 

Of 3485 articles identified, 10 randomized controlled trials (RCTs)^{17,19,20,22,23,25–29}  and 1 nonrandomized study were included in part II of this systematic review. Table 2 summarizes the characteristics and outcome measures of the included studies. Of these studies, 73% (n = 8)^{17,18,22,23,25,27–29}  were published between 2011 and 2021, with 27% (n = 3)^17,18,25  published between 2016 and 2021. All studies involved human volunteers exercising in a controlled environment. No evidence from extreme sporting events met the strict selection criteria. Participants included 148 physically active adolescents or young adults; authors of only 3 studies^26,27,29  included female participants (n = 27; 18.2%). In total, 3 different comparisons of interest were reported in the included studies. Authors of 4 studies^17,18,26,27  elaborated on the potential of skim or low-fat milk compared with water, 4 studies^19,20,28,29  on coconut water, and 3 studies^22,23,25  on beer (with different percentages of alcohol).

The synthesis of findings from all included studies is summarized in Tables 3 through 5, and a narrative overview of the results is given below. Given that the individual participant data were not available from the studies, meta-analyses could not be performed.³⁰  Furthermore, across all comparisons and outcomes, marked heterogeneity in study design and outcomes precluded meta-analysis.

We identified 3 RCTs, including 38 participants, in which authors evaluated the effect of beer on the volume/hydration status after exercise-associated dehydration compared with water (Table 3).^22,23,25  This critical outcome was measured as cumulative urine output, net fluid balance, plasma volume change, and hematocrit percentage. Authors of 2 studies did not demonstrate a difference in cumulative urine output from rehydration with beer with regular (4.5%–5%), low (low-alcohol; 0.5%–2%), or no alcohol content (nonalcohol; 0%) compared with water.^23,25  Flores-Salamanca and Aragón-Vargas showed that drinking regular beer (4.5%–5%) compared with water resulted in an increase in cumulative urine output.²²  Consumption of regular beer for rehydration compared with water was not shown to affect net fluid balance, change in plasma volume, or hematocrit percentage.^23,25  Similarly, low- or nonalcoholic beer was not shown to improve net fluid balance after rehydration compared with water.²⁵  Jiménez-Pavón et al found no difference in serum sodium concentration after rehydration with regular beer compared with water.²³  We did not identify any evidence to address the outcomes of heart rate, need for advanced care, or patient satisfaction.

Authors of 3 RCTs and 1 nonrandomized study including 68 participants studied the rehydration potential of skim or low-fat cow’s milk after exercising compared with water (Table 4).^17,18,26,27  Authors of all included studies showed a decrease in cumulative urine output after rehydration with skim or low-fat milk compared with water. Authors of 3 studies showed an increase in net fluid balance after 1 hour (mean difference [MD] = 655, 368, and 111 mL higher) and 2 hours (MD = 675, 621, and 179 mL higher) after rehydration with skim milk compared with water.^17,26,27  In addition, Sayer et al found an increase in net fluid balance at 30 minutes to 1.5 hours after rehydration with low-fat milk (MD = 0.26 L higher) or at 1.5 to 2.5 hours after rehydration with low-fat milk (MD = 0.36 L higher) compared with water.¹⁸ 

Authors of 2 studies, including 19 participants, evaluated plasma osmolality.^17,18  Sayer et al reported an increase in plasma osmolality at 1.5 to 2.5 hours after rehydration with skim milk compared with water (MD = 3 mOsm/kg higher), but Seery and Jakeman did not demonstrate a difference in plasma osmolality at 1 and 2 hours after rehydration.^17,18  Authors of all 4 studies also evaluated the patient satisfaction of the rehydration drinks.^17,18,26,27 

No difference between milk and water could be demonstrated in the perception of thirst at any time point.^17,18,26,27  However, Sayer et al showed that drinking milk during recovery led to higher perceived levels of gastric fullness and bloating compared with drinking water.¹⁸  Authors of the other studies did not demonstrate a difference between water or skim milk for gastric fullness, bloating, or both at any time after rehydration.^17,26,27 

Authors of 4 RCTs, including 42 participants, described the efficacy of coconut water (ie, fresh coconut water or coconut water from concentrate) for treating exercise-associated dehydration compared with water (Table 5).^19,20,28,29  For the outcomes related to volume/hydration status (ie, cumulative urine output, net fluid balance, and plasma volume changes), no differences were demonstrated at any time of interest between rehydration with fresh coconut water and regular water.^19,20,29  Similarly, no difference was reported for heart rate at 2 hours after rehydration with fresh coconut water or coconut water from concentrate compared with water.²⁸ 

Ismail et al showed an increased serum sodium osmolality (MD = 3 mOsm/kg higher) and concentration (MD = 2 mmol/L higher) 1 hour after rehydration with fresh coconut water compared with water.¹⁹  However, Saat et al reported no difference for serum sodium osmolality and concentration 1 hour after rehydration with fresh coconut water compared with water.²⁰  Furthermore, Kalman et al also did not demonstrate a difference in plasma osmolality 2 hours after rehydration with fresh coconut water compared with water.²⁸  They showed an increase in plasma osmolality 2 hours after rehydration with coconut water from concentrate compared with water (MD = 1.5 mmol/L higher).

Regarding patient satisfaction, no differences in perceived feeling of thirst or gastric fullness were demonstrated between drinking coconut water or regular water at any time recorded.^19,20,28,29  Although Saat et al reported a decrease in nausea and stomach upset immediately and 1 hour after rehydration with fresh coconut water compared with water, authors of most studies did not show a difference in nausea, gastric discomfort, or cramps in the first hour after rehydration.^19,20,28,29  However, at 2 hours after rehydration, Kalman et al showed an increase in gastric discomfort with fresh coconut water or coconut water from concentrate.²⁸  They did not demonstrate a difference in bloating in the 2 hours after rehydration with fresh coconut water or coconut water from concentrate.

An overview of the limitations in study design for each RCT is shown in Table 6. Details of the randomization and allocation process were not provided in all RCTs. In most studies, the blinding of participants, investigators, outcome assessors, or a combination was not described. However, it can be assumed that a double-blind design was not possible due to inherent differences in taste and texture of the study drinks. Moreover, the objective outcomes based on diagnostic records retrieved from blood and urine analyses are less prone to risk of bias from lack of blinding. The trials in which authors evaluated subjective outcomes were most at risk of bias from lack of blinding. We found, in general, no concerns regarding missing outcome data. Lastly, no information was available about the researchers’ prespecified intentions regarding planned outcome measurement and analyses. Protocols of the trials were not published or registered, raising some concerns about bias in selection of the reported results. One study was at high risk of bias from selection of the reported results because the authors collected urine but did not elaborate on the results of the critical volume/hydration status outcome.²⁸  Overall, the included studies had serious limitations in their study designs mainly due to risk of bias regarding measurement of the subjective outcomes.

Looking at the risk of bias of the nonrandomized study, some concerns existed regarding bias due to confounding arising from selection bias, as the sequences of intervention were not randomly assigned.¹⁸  In addition, given that the subjective outcomes were self-reported and the participants were not blinded to the intervention, the measurement of these outcomes was judged to be at serious risk of bias.

The Grading of Recommendations, Assessment, Development and Evaluation (GRADE) assessment of the certainty of the body of evidence is detailed in Tables 3 through 5.³¹  The 11 included studies were experimental, leading to a high initial certainty level. The certainty of evidence was downgraded (−1) for limitations in study design. The overall certainty of evidence was further downgraded because of imprecision (ie, −1) due to limited sample sizes, a lack of data, or both (95% CIs could not be calculated because individual participant data were not available). Moreover, publication bias was strongly suspected (−1) as most trials were funded by industry partners who had a direct interest in the effectiveness of these rehydration beverages. This conflict of interest directly led to concerns about the independence of the findings in these trials. Downgrading due to indirectness or inconsistency was not needed. In conclusion, we found low to very low certainty evidence for the experimental studies concerning the rehydration potential of beverages for exercise-associated dehydration, meaning that any estimate of effect is very uncertain.

The aim of this systematic review was to identify the best available evidence on the effectiveness of any alternatives to CE drinks that can restore the hydration balance in people with exercise-associated dehydration. Although drinking water after sporting events or exercise (especially in hot and humid environments) is generally suggested as a good practice, athletes need to be aware of the risks of drinking hypotonic beverages, including water, in quantities that may lead to hyponatremia.^1,32  Remarkably, very little evidence evaluating the rehydration potential of both commercial and noncommercial fluids is currently available. In performing this systematic review, we acknowledge that, in the setting of exertional dehydration, rehydrating as soon as possible is most important. The choice of hydration fluid will often be made based on what is available, what the dehydrated person is willing to drink, and palatability.

The main finding of this systematic review, although the best available evidence is of low to very low certainty, suggests improvement in several indicators of volume/hydration status after hydration with skim or low-fat cow’s milk compared with water. However, rehydration with milk, without additional food intake, might be associated with feelings of fullness and bloating when compared with rehydration with water. For coconut water, no differences were demonstrated regarding cumulative urine output, net fluid balance, and plasma volume changes when compared with water.^19,20,29  However, although variability exists among the measured outcomes, drinking coconut water may assist in restoring the serum sodium osmolality and concentration.^19,20  Finally, evidence remains insufficient concerning the effect of consuming beer (0%–5% alcohol) on the reported outcomes.

The rate at which the drinks enter circulation is the first step that dictates fluid retention, and researchers have reported that the rate of gastric emptying is dominantly modulated by the energy density of the consumed liquids.^33–35  Skim or low-fat cow’s milk is a nutrient-dense food, containing carbohydrates, protein, electrolytes, and micronutrients at a concentration like that of many commercially available CE drinks. The energy density of milk is like that of CE solutions, and researchers have reported that fluids containing milk protein slowed the rate of gastric emptying and delayed the rate of intestinal water absorption.^33,36  Slower emptying and absorption may sustain a higher plasma osmolality that would attenuate urine production by the kidneys.³⁷  Evidence is limited that the sensation of stomach fullness is greater after ingesting milk compared with water. After fluid is absorbed and released into circulation, the magnitude and time course of change in plasma osmolality are key determinants of water retention in the circulation. Compared with water and many other rehydration drinks available, milk especially contains relatively high concentrations of sodium, potassium, and chloride, and this high electrolyte concentration likely maintains plasma osmolality and thereby prevents the excretion of dilute urine.^26,27  Therefore, the energy density and the relatively high electrolyte concentration of milk explain its beneficial effects on fluid replacement. The composition and effectiveness of a rehydration beverage needs to be balanced with its desirability and palatability. As such, rehydration with milk may be associated with other issues of patient satisfaction or compliance when compared with water. In addition, rehydration with milk, without additional food intake, by people with lactose intolerance may induce adverse effects such as diarrhea, which could hinder the management of rehydration.

Evidence was insufficient to show that rehydration with coconut water, naturally rich in potassium and carbohydrates, resulted in improved outcomes compared with water for postexercise rehydration.²⁹  The presence of potassium in a drink may not be as effective as sodium in improving hydration status. Coconut water has been widely used in the tropics as a simple, palatable drink, but it may be more costly in geographic regions where fresh coconuts are not readily available and coconut water is commercialized as a natural sports drink.^28,29  In addition, some people may find coconut water less palatable than water.

Lastly, although alcoholic beverages are commonly consumed after exercising in a nonprofessional context in Western countries, researchers debate that the consumption of alcoholic beverages may have other unwanted effects such as adverse effects on performance, behavioral changes, and potential intoxication.³⁸  Moreover, alcohol increases the diuretic response in the body, which could interfere with adequate rehydration after exercise.³⁹  However, because of its nutritional composition and the absence of alcohol, one could suspect that nonalcoholic beer may have greater fluid-retention properties than water.²³  We were unable to detect any difference between the hydration responses to low-alcohol or nonalcoholic beer and water. Although the carbohydrate concentration of beer, in particular of nonalcoholic beer, is close to that found in commercial CE drinks, researchers have previously hypothesized that the electrolyte concentrations were too low to make any difference.^22,23,40  Indeed, investigators have shown that the addition of 25 to 50 mmol of sodium per liter to a low-alcohol beer (2.3% alcohol) may improve its fluid-retention properties.^41,42  The current limited evidence is insufficient to formulate any recommendation on the rehydration potential of low-alcohol or nonalcoholic beers. However, high alcohol intake should not be recommended, as the physiological and health consequences could be unpredictable.

The main strength of this systematic review is that the best available evidence on the efficacy of rehydration drinks was, for the first time, identified and assessed via the rigorous and transparent Cochrane and GRADE methodology. The evidence from this review was used as a direct scientific basis by the First Aid Task Force of the ILCOR to develop a Consensus on Science with Treatment Recommendations.⁴³  This consensus document, in turn, will be used by international medical and first aid guideline organizations to provide recommendations for professional and recreational athletes, their families and coaches, and first-aid responders. In this context, the findings of this 2-part systematic review have several practical implications in the field. First aid providers are commonly recruited to assist at first aid stations located at sporting and challenge events where exercise-induced dehydration is a common problem. Providing the most effective rehydration options to athletes will ensure complete fluid restoration and will avoid the complication of simple dehydration progressing into a potentially life-threatening medical emergency. It is crucial to note that oral hydration may not be appropriate for individuals with severe dehydration associated with hypotension, hyperpyrexia, or mental status changes.

The major limitation of this review is that it only provides evidence of low to very low certainty because of serious concerns regarding the design of the included studies, imprecise results, and publication bias. Moreover, the applied experimental methods varied widely between included studies, and, for example, differences existed in level of dehydration, amount of liquids consumed, and monitoring time after rehydration. These variations may compromise the reliability of the findings of this review. All included studies were crossover trials, and unfortunately, the individual participant data required to include a paired analysis in a meta-analysis were not published. Furthermore, we limited the systematic review to rehydration drinks for which more than 1 study was identified. In addition, the inclusion of studies was made based on reported outcomes. Thus, no studies were included for any type of yogurt drink or tea.^44–48  Authors of all included studies conducted exercise in a controlled environment and for a specific period. The laboratory environment may not adequately mimic conditions faced by recreational athletes undertaking habitual training and competition, such as variability in environmental temperature, humidity, and wind strength and differences in the intensity of training and competition.

Future studies should be done to target individuals with simple dehydration caused by participation in sporting events, personal exercise, or both. In this systematic review, we focused exclusively on the effectiveness of various drinks for replacing fluid losses after exercising and correcting dehydration. Although this first aid strategy might be effective over a short recovery period (<4 hours), foods are commonly consumed after exercising, and the electrolyte and water contents of these meals contribute to fluid replacement and can enhance fluid retention.^1,49  Future research might be done to focus on the addition of different foods combined with CE fluid consumption. Nevertheless, the NATA has recommended rehydrating with a fluid volume of 125% to 150% of the body mass lost to account for extra urinary losses.¹  At mass sporting events, specialized medical providers may take precompetition and postcompetition body mass of athletes to provide valuable information for postcompetition rehydration. However, this is not always the case and certainly not at smaller events. In these cases, it remains unclear how medical providers can determine the amount of liquid required for rehydration. Finally, we only focused on proposed rehydration beverages other than CE drinks compared with drinking water for effective rehydration. Based on our findings, the potential of alternative beverages, especially milk, for treating exercise-associated dehydration could be compared with the effectiveness of drinking CE drinks in a future review.

We found very low certainty evidence that skim or low-fat cow’s milk has beneficial effects on rehydration after exercise. The water, energy, and macronutrient content of milk appears to be like that of sports drinks. However, rehydration with milk may be associated with other issues of patient satisfaction or compliance when compared with water. For coconut water, no differences were demonstrated regarding rehydration outcomes in comparison with water. Lastly, insufficient scientific evidence is available to support a recommendation for the use of beer (in any alcohol percentage) as a rehydration drink. However, the use of alcoholic beverages may have unwanted effects, and therefore, beer is not recommended as a rehydration drink for athletes.

Drs Borra, De Brier, and De Buck are employees at the Belgian Red Cross and receive no other funding. One of the activities of the Belgian Red Cross is providing first aid training to laypeople. This work was made possible through funding from the Foundation for Scientific Research of the Belgian Red Cross (Mechelen, Belgium). The ILCOR and the American Heart Association provided the resources to assemble the Task Force and manage records and data. In addition to Drs Borra, Berry, Singletary, and Zideman, members of the ILCOR First Aid Task Force include Jestin N. Carlson, Pascal Cassan, Wei-Tien Chang, Nathan P. Charlton, Therese Djarv, Matthew Douma, Jonathan L. Epstein, Craig Goolsby, David S. Markenson, Daniel Meyran, Peter Morley, Michael Nemeth, Aaron Orkin, and Tetsuya Sakamoto. PROSPERO registration CRD420201530.