During continuous exercise in a hot environment, humans typically consume only 12-84% of the fluid that they lose in sweat (23). While dehydration develops, water moves within the body as the result of interacting extracellular and intracellular osmotic and hydrostatic forces. For example, plasma water moves from blood to the interstitial space (between cells) at the beginning of exercise, primarily because of increased blood pressure (hydrostatic force). This plasma water loss may be large (i.e., −14 to −23% during intense exercise) (32) and may impair cardiovascular function and endurance exercise performance and induce thirst (3,17,19,24). A portion of this plasma water moves into active skeletal muscle, drawn by increased intracellular concentration as a result of lactate accumulation (26). As a result of the net efflux of plasma, blood constituents become more concentrated and oncotic pressure (i.e., because of protein concentration) increases, encouraging fluid movement back to the circulation. These responses demonstrate that fluid movements during exercise are complex, dynamic, and influenced by osmotic, oncotic, and hydrostatic forces (15,21). During upright exercise in the heat, circulatory and thermoregulatory strain are greater than in a cool environment because (a) central blood volume is reduced by increased skin capillary filling (i.e., to cool the body), and (b) gravitational pooling of blood reduces venous return to the heart, ventricular filling, and cardiac output.
Cells generate and transport organic osmolytes such as betaine ((2(N,N,N-trimethyl)ammoniumacetate)) to regulate intracellular fluid homeostasis. One form of this molecule, glycine betaine, defends intracellular volume (6) and human kidneys (8) against great osmotic stress that occurs during progressive dehydration (11). Glycine betaine also protects simple organisms from very hot and very cold temperature extremes, and defends the critical Kreb's cycle enzyme citrate synthase against thermodenaturation (7). Thus, it is possible that betaine consumption acts to counteract dehydration, plasma volume loss, hyperthermia, metabolic dysfunction, and/or performance decrements during exercise.
Little information exists regarding the effects of human betaine consumption on intravascular volume and tonicity during any form of exercise (26). This is peculiar because an affluent Western diet provides up to 2.5 g betaine per day, especially when whole wheat bread or bran, spinach, beets, and seafood are included (11). Although the organic osmolyte betaine regulates the movement of water and electrolytes across the intestine (11), the extent to which betaine intake counteracts cardiovascular stress in humans during exercise, or influences intracellular and extracellular osmotic forces, is unknown. Therefore, the primary intent of the present investigation was to determine whether the nutritional supplement betaine would support cardiovascular function and thermal homeostasis, thereby acting as an ergogenic aid during exercise in the heat. This goal is relevant to athletes and soldiers who seek to counteract the detrimental effects of dehydration and environmental heat stress on exercise performance.
Experimental Approach to the Problem
In designing the present experiments, we realized that betaine could osmotically support plasma volume and circulatory function, leading to the development of three hypotheses. First, we hypothesized that rehydration with fluids containing betaine would elicit superior physiological and perceptual responses when compared with a non-caloric, flavored water solution. Second, we hypothesized that betaine would positively influence prolonged endurance exercise by optimizing plasma volume shifts (2,17). Third, we hypothesized that rehydration fluids containing betaine would elicit different responses (versus water) when combined with a 6% carbohydrate-electrolyte mixture. Therefore, this investigation examined the effects of rehydration with four different fluids (i.e., containing combinations of betaine, carbohydrate, electrolytes, and water) after −2.7% dehydration. The outcome variables included running performance (sprint duration); oxygen consumption; heart rate; sweat rate; rectal and skin temperatures; plasma lactate, glucose, sodium, and potassium concentrations; percent change of plasma volume; plasma osmolality; and psychophysiological ratings of perceived exertion, thirst, and thermal sensation.
The 10 healthy male runners who participated in this study exhibited the following characteristics: age, 20 ± 2 years; height, 177 ± 6 cm; weight, 70.6 ± 6.8 kg; body fat, 6.2% ± 2.1%; and maximal aerobic power, 63.5 ± 4.1 mL O2·kg−1·min−1. This latter value is considerably higher than an average college man (i.e., maximal aerobic power of approximately 50 mL O2·kg−1·min−1), and these subjects were best classified as well-trained, competitive distance runners. They were included in this study because they could complete the exercise protocol (see Procedures) comfortably. The university institutional review board approved all procedures, and each participant provided informed consent. A briefing was conducted to explain procedures, risks, and benefits. Subjects completed a medical history questionnaire to rule out contraindications to participation, such as a history of musculoskeletal injury or cardiovascular, metabolic, respiratory, and heat illness.
Test subject sample size was calculated before this investigation to determine the number of subjects needed to permit statistical judgments that were accurate and reliable, and the likelihood that statistical tests would detect significant effects of a given size in a particular situation. Using representative variables, an α level of 0.05 and a desired power of 0.80, the number of subjects was determined to be eight, but we proceeded with 10 subjects.
Peak oxygen consumption and the relative exercise intensities (50, 65, and 84%V̇o2max) were determined by graded exercise testing while running on a motorized treadmill (Quinton, Model 24-72, Seattle, WA). Maximal aerobic power (V̇o2max) and oxygen consumption (V̇o2) were measured with a metabolic cart (Medical Graphics Corp, St. Paul, MN). Body composition was assessed from three-site skinfold thicknesses and calculated using the Brozek equation (5). Subjects recorded their training (7 days) plus food and fluid intake (3 days) before the initial experimental trial. Analysis of these dietary records (Nutritionist IV software; First DataBank Inc., San Bruno, CA) provided investigators with a 24-hour dietary plan and 24-hour workout schedule that subjects repeated during subsequent trials. Subjects were instructed to avoid consumption of alcohol, caffeine, or drugs within 48 hours of testing and to abstain from physical training within 24 hours of experiments.
Subjects visited the laboratory on the morning before each experiment. Investigators obtained a urine specimen and blood sample, then measured body mass and urine specific gravity to verify euhydration (27). On the day of each experimental trial (day 1), subjects reported at 0700 hours after a 12-hour overnight fast (i.e., food and fluid); this resulted in a body mass change of −0.5 to −0.7%. Body mass and a urine sample were obtained before subjects consumed a standardized breakfast consisting of one bagel, one tablespoon of cream cheese, a banana, and 100 mL water. An indwelling venous catheter was placed into a superficial antecubital vein for subsequent blood sampling. Subjects then entered a heated environmental chamber (dry bulb, 31.1 ± 0.7°C or 88.0 ± 1.3°F; wet bulb, 19.9 ± 1.0°C or 67.8 ± 1.8°F; relative humidity, 34.7 ± 5.5%) (Model 200; Minus Eleven Inc., Malden, MA) for the duration of the experiment. Body mass, a urine specimen, and a blood sample were taken after 20 minutes of standing equilibration. Subjects dehydrated by performing exercise (30 minutes of cycling, 30 minutes of walking, 30 minutes of cycling) at 50%V̇o2max, until −2.7% body mass was achieved. Subjects then consumed 1 L rehydration fluid within 25 minutes while seated, returning mean body mass to −1.4% dehydration level. After 20 minutes of standing in an upright posture, a blood sample was drawn. Subjects began the running performance trial 45 minutes after rehydration ended.
On each of four experimental days, subjects consumed one of the following rehydration mixtures: a commercial 6% carbohydrate fluid containing electrolytes (C); C + 5 g·L−1 betaine (C+B); a flavored, non-caloric water placebo solution similar to C in flavor, color, and sweetness (W); or W + 5 g·L−1 betaine (W+B). The composition of each rehydration fluid is described in Table 1. Betaine (99% purity as a dry solid) was provided by Danisco USA Inc. (Ardsley, NY). Betaine is a FEMA GRAS flavor that can be used in foods up to 0.5%, making 5 g·L−1 a practical level. Ingestion of 5 g betaine in 1L water provides a safe level. Many studies have been conducted with 3-6 g betaine (11) and shown to provide physiological benefits such as reduced serum homocysteine. The optimal level of betaine for an ergogenic effect remains unknown.
The time of day was controlled among trials (±5 min). Before the experimental exercise bout, the subject again stood for 20 minutes, venous blood was sampled, and a urine specimen was collected. The four experimental trials were separated by a minimum of 7 days using a randomized, double-blind, cross-over design.
After this dehydration and rehydration, the experimental exercise trial (beginning at a dehydration level of −1.4% of baseline body mass) involved treadmill running for 75 minutes (65% V̇o2max) followed by a timed sprint to volitional exhaustion (84% V̇o2max). Subjects were instructed to give a best effort during each experiment, and verbal encouragement was provided during each sprint. Treadmill speeds were measured with a tachometer to achieve the appropriate relative exercise intensity based on preliminary testing. A fan was placed in front of subjects to provide air movement (2.5 m·s−1). Venous blood samples were obtained immediately before exercise (Pre), at 36 minutes (36) and 72 minutes (72) of the continuous run, immediately post-sprint (IP), and 15 minutes post-sprint (15P). The duration of the final sprint was measured to the nearest 0.1 second with a handheld stopwatch. Expired gases were sampled and analyzed for oxygen consumption and respiratory quotient (V̇co2/V̇o2) three times. Heart rate and rectal temperature were recorded periodically throughout the entire exercise protocol.
The reliability of two of these variables has been previously reported (1), and they are included as representative examples. First, oxygen consumption has an analytical variation (coefficient of variation (CV) = SD/X · 100) of 0.41% and a CV because of differences of metabolism of 3.8% during day-to-day measurements. Second, blood lactate concentration has an analytical CV of 3.3% and a metabolic CV of 13.0% during day-to-day measurements.
The subjects' overall rating of perceived exertion was measured four times using a six- to 20-point category-ratio scale (4). Perceptual ratings of thirst were recorded at five time points using a nine-point scale containing five thirst ratings (e.g., 1 = not thirsty at all; 5 = moderately thirsty; 9 = very very thirsty) (13). Thermal sensations were evaluated using a rating scale developed by Young and colleagues (33). This scale progresses from 0.0 (unbearably cold) through nine categories to 8.0 (unbearably hot); intermediate categories include very cold, cold, cool, comfortable (4.0), warm, hot, and very hot.
Hematocrit was measured with a microcapillary reader (International Equipment Co., Needham Heights, MA) in triplicate after centrifugation. Hemoglobin was determined spectrophotometrically in duplicate samples (Spectronic 401; Spectronic Instruments, Rochester, NY) using cyanmethemoglobin reagent. Percent change of plasma volume was calculated from hematocrit and hemoglobin values as described by Dill and Costill (12). Plasma glucose and lactate concentrations were determined in duplicate with an automated analyzer (Model 2300STAT; Yellow Springs Instruments, Yellow Springs, OH). Plasma sodium and potassium concentrations were assessed via ion-selective electrodes (Model 984-S; AVL Scientific Corporation, Roswell, GA). Plasma osmolality was determined without freezing via freezing-point depression osmometer (Model 3DII; Advanced Digimatic, Norwood, MA). Betaine was analyzed in plasma samples that had been preserved in EDTA and frozen at −85°C (−121°F); high-performance liquid chromatography analyses utilized a silica column in a mixed partition and ion exchange mode, following the method of Laryea et al. (18).
Means and SD were computed with statistical software (SPSS Inc., Chicago, IL). No statistical comparisons were made between fluids containing carbohydrate (C and C+B) and those without carbohydrate (W and W+B). Analysis of variance (treatment × time) and paired sample t-tests were used to compare differences among the four rehydration trials at all time points, and Tukey's post hoc test was used to determine pairwise differences in the event of a significant F-ratio at the p ≤ 0.05 confidence level. Effect size was calculated for selected measurements as an indication of the strength of the relationship between two variables using Cohen's d test (9). All variables were analyzed for an order effect; none existed.
Dietary analyses indicated that there were no significant differences of daily macronutrient and micronutrient consumption (i.e., 2847 ± 290 Kcal, 4127 ± 557 mg sodium, 3539 ± 373 mg potassium, 18% ± 1% protein, 57% ± 3% carbohydrate, and 25% ± 2% fat) during the 72 hours before laboratory testing on all days. The morning hydration of subjects (i.e., represented by urine specific gravity at 0700 hours) was statistically similar on all days before exercise testing (1.020 ± 0.002); it also was similar after dehydration on all days of testing (1.026 ± 0.001). This increased urine specific gravity reflected both overnight fasting and exercise-induced dehydration. All subjects reported that their physical exercise training was similar during the 24 hours before each exercise test.
The betaine compound was verified as 99% pure by laboratory tests. During the W+B and C+B tests, plasma betaine levels (Figure 1) were markedly higher than during the W and C tests, as expected, but were statistically similar. Plasma betaine concentrations were maximal (C+B, 954; W+B, 933 μmol·L−1) approximately 1 hour after ingestion; this observation agrees with previously published pharmacokinetics data that involved a similar dosing protocol (25). Because of the similarities in flavor and color, test subjects reported that they could not distinguish which fluid they consumed on any day.
At the conclusion of the final sprint (3.1-3.8 minutes), the strenuous nature of this exercise-heat protocol was evident in all experiments as a mean heart rate of 178-188 b·min−1, mean rectal temperature of 39.1-39.3°C (102.4-102.6°F), and mean rating of perceived exertion of 19-20. Subjects ended the sprint phase with a −3.9% body mass loss. During exercise and recovery, all of these variables plus thirst, skin temperature, and whole body sweat rate (1.40 ± 0.06 L·h−1) were similar among fluid types.
Because of great exercise-heat stress, the following variables increased significantly within experiments (p < 0.05) during the 78.5-minute exercise performance test: heart rate, oxygen consumption, rectal temperature, rating of perceived exertion, thermal sensation, thirst rating, plasma osmolality, and plasma concentrations of lactate, glucose, sodium, and potassium; skin temperature was a noteworthy exception. These significant differences were anticipated, are intuitive, and are not marked in Tables 2-5 because they are numerous and would obscure our focus on the experimental hypotheses and objectives of this investigation.
The sprint time to exhaustion was not statistically different among fluid treatments (Figure 2). However, a nonsignificant trend of sprint duration occurred during both betaine experiments. When C+B (228 ± 173 seconds) was compared with C (196 ± 119 seconds), the difference between mean sprint times to exhaustion was 32 seconds (16%; p = 0.12). The W+B (223 ± 165 seconds) test differed from W (185 ± 71 seconds) by 38 seconds (21%; p = 0.22). Effect size of the relationship between W and W+B or C and C+B was small for both (9). The sprint data of one test subject were removed before conducting these statistical analyses (Figure 2) because his shortest and longest sprint times varied by 111%; no other runner had this discrepancy in performance, and we interpreted this to mean that one or more of his sprints were not his best effort.
The mean oxygen consumption for all treatments, measured three times during exercise (Table 2), was similar except that whole-body oxygen uptake during C (52.3 ± 2.7 mL·kg·min−1) was significantly lower than C+B (54.96 ± 5.65 mL·kg·min−1) during the final sprint. Statistical comparisons of the corresponding respiratory quotients (V̇co2/V̇o2) during the final sprint of C (1.04 ± 0.01) and C+B (1.04 ± 0.03) indicated that they were not different; this also was true of the W (1.01 ± 0.03) and W+B (1.03 ± 0.02) experiments. However, the rating of perceived exertion (Table 2), which is statistically correlated to heart rate and oxygen consumption, was similar for all treatments at all time points.
Blood glucose concentration (Table 4) was similar among all treatments at all exercise time points. The plasma lactate concentration of the C trial was significantly lower (p < 0.05) than C+B during the IP and 15P measurements. No other statistical differences resulting from fluid type were detected for plasma lactate.
Plasma Volume, Concentration, and Electrolytes
The mean percent changes of plasma volume (standing at rest and during exercise), as determined from hematocrit and hemoglobin analyses (12), are depicted in Figure 3. Significant differences (p < 0.05) were identified between W and W+B but not between C and C+B. These significant differences in fluid shifts (W vs. W+B) occurred between the end of dehydration and the end of rehydration (pre-exercise), at 36 (65%V̇o2max), and at 15P.
No between-treatment differences existed for plasma osmolality, sodium, or potassium concentrations (Tables 4 and 5) at any time point. The plasma osmolality at the beginning of experiments C and C+B (Time 0, Table 5) tended to be higher (not significant, p > 0.05) than during their corresponding water control trials (W and W+B); this was likely the result of the greater osmolality of the carbohydrate-electrolyte mixtures consumed during C and C+B (Table 1).
Thirst and drinking behavior are mechanistically linked to plasma osmolality (17). Despite differences of sodium content and osmolality for the four fluids (Table 1), fluid composition had no effect on thirst rating.
We initially hypothesized that rehydration with fluids containing betaine would elicit (a) superior physiological and perceptual responses when compared with a water placebo, (b) improved running performance by optimizing plasma volume shifts and maintaining internal temperature, and (c) different responses (vs. water) when combined with a 6% carbohydrate-electrolyte mixture. With few exceptions, hypotheses A and B generally were not supported, but hypothesis C was verified with some variables. Rehydration fluids containing the osmolyte betaine affected small but significant differences (p < 0.05) of plasma volume shifts (before, during, and after exercise; W vs. W+B), oxygen consumption (during the sprint; C+B>C), plasma lactate concentration (IP and 15P; C+B>C), and thermal sensation (pre-exercise, C>C+B; IP, W>W+B and C>C+B).
The 10 male runners who participated in this study had a mean maximal aerobic power of 63.5 ± 4.1 mL O2·kg−1·min−1. Although this latter value is considerably higher than that of an average college man (i.e., maximal aerobic power of approximately 50 mL O2·kg−1·min−1), these subjects were best classified as non-elite, recreationally competitive runners who participated in road races occasionally or regularly. They were included in this study because they could complete the exercise protocol (see below) without disqualifying hyperthermia (final rectal temperature > 39.5°C) and were capable of running 8-10 miles continuously. The SDs of sprint duration were large (Figure 2). The magnitude of this variation was an important factor in our observation that rehydration with betaine did not enhance performance during strenuous sprinting in the heat, as evaluated at the 0.05 level of confidence. Had a homogeneous sample of elite runners been tested, the variability of mean sprint duration values likely would have been smaller, and significant findings may have resulted. However a small, nonsignificant (p > 0.05) trend toward improved mean sprint duration was observed (Figure 2). Specifically, the C+B sprint time was greater than C by 32 seconds (+16%), and the W+B sprint time was greater than W by 38 seconds (+21%). These trends corresponded to differences of thermal sensation (p < 0.05), as explained below.
The metabolic measurements, however, did not explain why a favorable ergogenic outcome might result from betaine rehydration. First, no between-treatment differences were detected for plasma glucose concentrations; this indicated similar glucoregulatory responses to all fluid types. Second, plasma lactate levels were greater during C+B (p < 0.05) than during C immediately after and 15 min after the sprint. Although an increase of plasma lactate may involve increased muscle glycolysis or decreased muscle lactate concentration because of enhanced intracellular lactate clearance, the dynamics of lactate turnover complicate this interpretation because the plasma lactate concentration represents skeletal muscle production, efflux via lactate transporters in muscle membranes, and removal of lactate as a substrate by active and inactive tissues. Using only plasma lactate values (Table 4), the mechanism cannot be determined by the present protocol, but because this effect was not observed during continuous running at 65% V̇o2max (see Table 4 at 36 and 72 min), it apparently was related to exercise intensity. Third, oxygen consumption was 4-5% greater in C+B vs. C (Table 2) during the final sprint to exhaustion (84% V̇o2max). This means that betaine increased energy production via aerobic metabolism, because the treadmill speed was controlled in all experiments (65% and 84% V̇o2max). Because this effect was not observed during continuous running at 65% V̇o2max (see Table 2 at 12 and 72 min), it apparently was related to exercise intensity. Alternatively, betaine may have influenced metabolism in ways that are not completely understood (14). Although the respiratory quotients (V̇co2/V̇o2) for all treatments were statistically similar, indicating that the ratio of carbohydrate-to-fat metabolism was not affected by fluid composition, respiratory quotient does not account for protein metabolism. Thus, it is theoretically possible that betaine increased oxygen consumption 4-5% in C+B vs. C because of enhanced protein metabolism; previous animal studies have demonstrated that betaine increases protein turnover and amino acid metabolism (11,16).
Betaine has been shown to protect simple organisms from hyperthermia (7). In the present study, however, rehydration with betaine did not alter internal heat storage, as represented by rectal temperature, or peripheral tissue heat, as represented by skin temperature (Table 3). Skin temperature was also one of the few physiological variables that did not increase during the running protocol. This observation is explained by the cutaneous cooling afforded by sweat evaporation (1.04 L·H−1) and the widely accepted fact that skin temperature remains constant among a wide range of air temperatures, including that maintained within our environmental chamber (31.1°C). Although rehydration with betaine did not alter rectal or skin temperatures during the present protocol, it may provide other thermal benefits, including defense of intracellular volume (6) and human kidneys (8) when dehydration is severe (11). Specific evaluation of these benefits was beyond the scope of the present study.
Thermal sensation (33), perceived exertion (4), and thirst (13) were measured because each of these sensations, when extreme, may contribute to a loss of motivation or a decision to discontinue high intensity exercise. The strenuous nature of this exercise protocol was verified by the high to near-maximal ratings for these three variables at 76 and IP (see RPE, Rth, and St in Tables 2, 3, and 5). The four fluid treatments, however, resulted in similar ratings of perceived exertion and thirst during exercise. In contrast, thermal sensation was lower (p < 0.05) (Table 3) in both betaine (W+B and C+B) experiments (vs. control fluids) at the end of the sprint (IP). It is possible that thermal sensation influenced the increased mean sprint duration, as described above (Figure 2).
After rehydration with 1 L fluid (pre-exercise), betaine supplementation in W+B resulted in greater resting plasma volume expansion vs. W (see Figure 3a). We attribute this to the concentration difference of the consumed fluids (W, 28 mOsm·kg−1; W+B, 76 mOsm·kg−1) (Table 1), which caused more water to move from the interstitial space into the blood during W+B. This means that test subjects began exercise with an expanded plasma volume-a theoretically positive condition-during betaine supplementation.
Interestingly, this relationship reversed during the initial 36 minutes of exercise (see Figure 3b), as the mean plasma volume expanded slightly (+3%) during the W trial but contracted mildly (+1%) during the W+B trial. At the end of the sprint (IP, 3.9% whole-body dehydration), the increased blood pressure resulting from exercise-heat stress (84% V̇o2max) overrode all osmotic or oncotic forces and caused similar large plasma volume shifts (−6% to −8%) in all tests. During recovery (15P) (see Figure 3), the loss of water from plasma was greater (i.e., more negative) for W+B than for W, although both involved plasma volume losses. These extracellular water movements into or out of the circulation involved the algebraic sum of complex osmotic, hydrostatic, and oncotic (i.e., related to protein) forces (15,21).
We did not anticipate these plasma volume changes in experiments W and W+B; these are unique among published studies involving exercise-heat stress (26). The juxtaposition of W and W+B (Figure 3) from pre-exercise to recovery (15P) may be explained by betaine transport. If betaine were transported across the skeletal muscle membrane (6,11), an osmotic gradient (i.e., intracellular > interstitial > plasma osmolality) would move fluid from the extracellular compartment into muscle tissue, and plasma volume would decrease. This interpretation is reasonable because we know that betaine is largely confined to the intracellular space and is moved into cells when consumed in the diet (20). However, this scenario is not consistent with the plasma volume shifts during experiments C and C+B (i.e., no significant differences at any time point) (Figure 3), perhaps because the osmolalities of mixtures C and C+B (Table 1) were considerably greater than fluids W and W+B, creating a different osmotic gradient between the intracellular and extracellular compartments. The greater osmolalities of C and C+B may have also delayed gastric emptying and subsequent intestinal absorption; high osmolality is one of several fluid characteristics that delay gastric emptying and the appearance of an ingested fluid in blood (10). Although the exact mechanism of these plasma volume changes cannot be determined (i.e., total plasma protein concentration, plasma oncotic pressure, and intracellular variables were not measured), it is unlikely that sodium or potassium concentrations played a significant role because their concentrations were similar among all experiments (Tables 4 and 5).
The above findings generated the following research recommendations, which are readily tested.
- Although rehydration with fluids containing betaine did not enhance running performance in the heat, the present protocol involved running and does not exclude the possibility that betaine may enhance the performance of other types of exercise or labor. Support for this hypothetical concept originates from the laboratory study published by Pivarnik et al. in 1984 (22), which involved endurance-trained men. Sixty minutes of low-intensity treadmill running (37% V̇o2max) caused final plasma volume to increase slightly (+3.3%), whereas running at 56 and 74% V̇o2max produced stable (+0.5%) and slightly decreased plasma volumes (−3.7%), respectively. Thus, because the slowest treadmill speed involved the most stable circulatory state, we recommend that betaine be studied with prolonged, low-intensity exercise, such as long-distance hiking or marching. It seems that betaine plus water would support these modes of exercise when consumed before exercise (see Figure 3a).
- Interestingly, the metabolic differences involving plasma lactate and oxygen consumption occurred in comparisons of C+B vs. C, not W+B vs. W. Thus, betaine interacted uniquely with the mixtures of carbohydrate and electrolytes (Table 1). This interaction has not been previously reported and warrants future research involving high-intensity exercise protocols.
- In view of the protection that betaine provides to cells (i.e., against dehydration, high intracellular concentration, and high internal temperatures) (7,11) and its role in human renal regulation of water and salt balance (28), it is possible that betaine has greater ergogenic potential when the level of dehydration is great (i.e., similar to those experienced by patients with heat exhaustion; −5% body weight or greater).
- Because betaine is primarily an intracellular constituent (20), it is possible that a single acute dose, as provided in the present protocol, is not adequate to induce measurable performance effects. We recommend that animal and human studies establish whether repeated days of betaine loading establish unique physiological capabilities. This approach is modeled after creatine, which enhances sprint and power performance. Repeated days of high creatine consumption (3-20 g·d−1) are required to bring intramuscular creatine to a stable, elevated level that provides body composition and ergogenic effects (29-31).
When betaine was consumed with water (W+B), plasma volume decreased slightly more than the control solution (W) during exercise and recovery. This effect was undesirable because plasma volume maintenance is essential during endurance exercise in the heat. When betaine was added to a carbohydrate-electrolyte mixture (C+B), metabolism increased during the final sprint (vs. C). Because successful endurance performance requires exercise efficiency (i.e., minimizing energy cost), this effect was undesirable. After 75 minutes of endurance running, the nutritional supplement betaine did not significantly enhance sprint performance (Figure 2). However, we interpret the increases of both aerobic and anaerobic metabolism during C+B (Tables 2 and 4) to mean that betaine supplementation may be beneficial during some forms of high-intensity exercise (i.e., middle-distance running, sprint cycling, repetitive lifting) because such exercise is less dependent on plasma volume maintenance than endurance activities.
The authors gratefully acknowledge the technical assistance of Heidi L. Hatch, Nicholas V. Mahood, Julie M. Clements, Lindsay Clarkson, Ashley Miller, Alex Seen, Katriina Pulliainen, and Jaci VanHeest. This investigation was funded in part by Danisco USA, Inc., a commercial supplier of betaine and the employer of Stuart A.S. Craig, Ph.D. The results of the present study do not constitute an endorsement, by the authors or the National Strength and Conditioning Association, of betaine as a product or as a nutritional supplement.
1. Armstrong, LE and Costill, DL. Variability of respiration and metabolism: responses to submaximal cycling and running. Res Q Exerc Sport
56: 93-96, 1985.
2. Armstrong, LE and Maresh, CM. Effects of training, environment, and host factors on the sweating response to exercise. Int J Sports Med
19(Suppl. 2): S103-S105, 1998.
3. Armstrong, LE, Maresh, CM, Gabaree, CV, Hoffman, JR, Kavouras, SA, Kenefick, RW, Castellani, JW, and Ahlquist, LE. Thermal and circulatory responses during exercise: effects of hypohydration, dehydration, and water intake. J Appl Physiol
82: 2028-2035, 1997.
4. Borg, GA. Psychophysical bases of perceived exertion. Med Sci Sports Exerc
14: 377-381, 1982.
5. Brozek, J, Grande F, Anderson, JT, and Kemp, A. Densitometric analysis of body composition: revision of some quantitative assumptions. Ann N Y Acad Sci
110: 113-140, 1963.
6. Burg, M. Molecular basis of osmotic regulation. Am J Physiol
268: F983-F996, 1995.
7. Caldas, T, Demont-Caulet, N, Ghazi, A, and Richarme, A. Thermoprotection by glycine betaine and choline. Microbiology
145: 2543-2548, 1999.
8. Chambers, ST and Kunin, CM. Isolation of glycine betaine and proline betaine from human urine: assessment of their role as osmoprotective agents for bacteria and the kidney. J Clin Invest
79: 731-737, 1987.
9. Cohen, D. A power primer. Psych Bull
112: 155-159, 1992.
10. Costill, DL. Gastric emptying of fluids during exercise. In: Perspectives in Exercise Science and Sports Medicine
(vol. 3). Fluid Homeostasis During Exercise
. Gisolfi, CV and Lamb, DR, eds. Carmel, In: Benchmark Press; 1990. pp. 97-128.
11. Craig, SAS. Betaine in human nutrition. Am J Clin Nutr
80: 539-549, 2004.
12. Dill, DB and Costill, DL. Calculation of percentage changes in volumes of blood, plasma and red cells in dehydration. J Appl Physiol
37: 247-248, 1974.
13. Engell, DB, Maller, O, Sawka, MN, Francesconi, RN, Drolet, L, and Young, AJ. Thirst
and fluid intake following graded hypohydration levels in humans. Physiol Behav
40: 229-236, 1987.
14. Fernandez-Figares, ID, Wray-Cahen, D, Steele, NC, Campbell, RG, Hall, DD, Virtanen, E, and Caperna, TJ. Effect of dietary betaine on nutrient utilization and partitioning in the young growing feed-restricted pig. J Anim Sci
80: 421-428, 2002.
15. Guton, AC and Hall, JE. The microcirculation and the lymphatic system: capillary fluid exchange, interstitial fluid, and lymph flow. In: Textbook of Medical Physiology
(9th ed.). Philadelphia: W.B. Saunders; 1996. pp. 183-197.
16. Haussinger, D. The role of cellular hydration in the regulation of cell function. Biochem J
313: 697-710, 1996.
17. Hubbard, RW, Szlyk, PC, and Armstrong, LE. Influence of thirst
and fluid palatability on fluid ingestion during exercise. In: Perspectives in Exercise Sciences and Sports Medicine. Fluid Homeostasis During Exercise
. D.R. Lamb and C.V. Gisolfi, eds. Indianapolis: Benchmark Press; 1990. pp. 39-95.
18. Laryea, MD, Steinhagen, F, Pawliczek, S, and Wendel, U. Simple method for the routine determination of betaine and N, N-dimethylglycine in blood and urine. Clin Chem
44: 1937-1941, 1998.
19. Maughan, RJ. Impact of mild dehydration on wellness and on exercise performance. Eur J Clin Nutr
57(Suppl. 2): S19-S23, 2003.
20. McNeil, SD, Nuccio, ML, and Hanson, AD. Betaines and related osmoprotectants: targets for metabolic engineering of stress resistance. Plant Physiol
120: 945-949, 1999.
21. Nose, H, Morimoto, T, and Ogura, K. Distribution of water losses among fluid compartments of tissues under thermal dehydration in the rat. Jpn J Physiol
33: 1019-1029, 1983.
22. Pivarnik, JM, Leeds, EM, and Wilkerson, JE. Effects of endurance exercise on metabolic water production and plasma volume
. J Appl Physiol
56: 613-618, 1984.
23. Rothstein, A, Adolph, EF, and Wills, JH. Voluntary dehydration. In: Physiology of Man in the Desert
. E.F. Adolph, ed. New York: Interscience Publishers; 1947. pp. 254-270.
24. Sawka, M. Physiological consequences of hypohydration: exercise performance and thermoregulation. Med Sci Sports Exerc
24: 657-670, 1992.
25. Schwahn, BC, Hafner, D, Hohfeld, D, Balkenhol, N, Laryea, MD, and Wendel, U. Pharmokinetics of oral betaine in healthy subjects and patients with homocystinuria. Br J Clin Pharmacol
55: 6-13, 2003.
26. Senay, LC and Pivarnik, JN. Fluid shifts during exercise. Exerc Sport Sci Rev
13: 335-387, 1985.
27. Shirreffs, SM. Markers of hydration status. Eur J Clin Nutr
57(Suppl. 2): S6-S9, 2003.
28. Sizeland, PCB, Chambers, ST, Lever, M, Bason, LM, and Robson, RA. Short-term response of nonurea organic osmolytes in human kidney to a water load and water deprivation. Am J Physiol
268: F227-F233, 1995.
29. Terjung, RL, Clarkson, P, Eichner, ER, Greenhaff, PL, Hespel, PJ, Israel, RG, Kraemer, WJ, Meyer, RA, Spriet, LL, Tarnapolsky, MA, Wagenmakers, AJ, and Williams, MH. American College of Sports Medicine roundtable: the physiological and health effects of oral creatine supplementation. Med Sci Sports Exerc
32: 706-717, 2000.
30. Van Loon, LJ, Oosterlaar, AM, Harthens, F, Hesselink, MK, Snow, RJ, and Wagenmakers, AJ. Effects of creatine loading and prolonged creatine supplementation on body composition, fuel selection, sprint and endurance performance in humans. Clin Sci (Lond.)
104: 153-162, 2003.
31. Volek, JS and Rawson, ES. Scientific basis and practical aspects of creatine supplementation for athletes. Nutrition
20: 609-614, 2004.
32. Whittlesey, MJ, Maresh, CM, Armstrong, LE, Morocco, TS, Hannon, DR, Gabaree, CV, and Hoffman, JR. Plasma volume
responses to consecutive anaerobic exercise tests. Int J Sports Med
17: 268-271, 1999.
33. Young, AJ, Sawka, MN, Epstein, Y, Dechristofano, B, and Pandolf, KB. Cooling different body surfaces during upper and lower body exercise. J Appl Physiol
63: 1218-1223, 1987.
Keywords:© 2008 National Strength and Conditioning Association
thirst; osmolality; plasma volume; plasma sodium; plasma lactate; oxygen consumption