Hypohydration Effects on Skeletal Muscle Performance and Metabolism: A 31P MRS study

Scott J. Montain1, Sinclair A. Smith2, Ralph P. Mattot1, Gary P. Zientara3,
Ferenc A. Jolesz 3, Michael N. Sawka1

1 U.S. Army Research Institute of Environmental Medicine, Natick, MA

2 Boston University, Boston, MA

3 Brigham and Women's Hospital-Harvard Medical School, Boston, MA

Address for Correspondence:

Scott J. Montain, Ph.D.
Thermal & Mountain Medicine
Division U.S. Army Research Institute of Environmental Medicine
Natick, Massachusetts 01760
voice (508) 233-4564
fax (508) 233-5298
email: smontain@natick-ccmail.army.mil

Running Title: Dehydration and Muscle Metabolism

(The full text is avaliable as a PostScript or pdf)


The purpose of this study was to determine if hypohydration reduces skeletal muscle endurance and whether increased H+ and Pi might contribute to performance degradation. Ten physically active volunteers
(age 21-40 y) performed supine single-leg knee extension exercise to exhaustion in a 1.5 Tesla whole body magnetic resonance spectroscopy (MRS) system when euhydrated and when hypohydrated (4% body weight). 31P spectra were collected every 10 seconds at rest, exercise and recovery. The desired hydration level was achieved by performing 2-3 h exercise in a warm room (40 C db, 20% rh) with or without fluid replacement 3-8 h prior to the experiment. Time to fatigue was reduced (P<0.05) 15% when hypohydrated (EU=251ñ67, HY=213ñ52 sec; ñsd). Muscle strength was generally not affected by hypohydration. Muscle pH and Pi/á-ATP were similar during exercise and at exhaustion regardless of hydration state. The time constant for PCr recovery were also not different between trials. These data demonstrate that hypohydration reduces muscle endurance and the effects appear independent of H+ and Pi concentration.

Keywords: dehydration, fatigue, acid-base balance


The effects that hypohydration (body water deficit) has on increasing heat strain and cardiovascular strain and aerobic exercise performance are well documented. Less understood are the effects of hypohydration on skeletal muscle performance and metabolism. While two studies reported that hypohydration reduced muscle endurance another study found no difference in fatiguability during handgrip exercise . Likewise, anaerobic exercise performance has been reported to be decreased or not altered by hypohydration. These previous studies are somewhat confounded, however by not controlling for prior exercise and/or heat exposure and different caloric intake prior to the performance tests. Research is needed that controls for these confounding variables, to determine if hypohydration has direct effects on skeletal muscle that contribute to the well documented reductions in aerobic performance. The physiological mechanisms responsible for reduced muscle performance might include both peripheral and central factors. Within muscle, hypohydration might accelerate depletion of energy stores, accumulation of metabolites (e.g. lactate, hydrogen ions, inorganic phosphate), changes in intracellular electrolyte concentrations, and reduce buffering capacity. Studies examining the effects of hypohydration on muscle glycogen use have found either no affect or a small increase in muscle glycogen utilization. Similarly, hypohydration has been reported to not alter or increase muscle lactate concentration. The effects of hypohydration on intracellular hydrogen ions (H+) or inorganic phosphate (Pi) concentrations in skeletal muscle have not been studied. Elevated H+ and Pi concentrations have been shown to reduce muscle force production during repeated contractions, and intracellular concentrations would be increased by simply reducing intracellular water. These two metabolites can be measured non-invasively and repeatedly during exhaustive exercise using 31P magnetic resonance spectroscopy (MRS). The purpose of this study was to determine if hypohydration reduces skeletal muscle performance and to determine whether increased H+ and Pi concentrations might contribute to peformance degradation. We hypothesized that hypohydration would reduce skeletal muscle endurance and act via increased H+ and Pi concentrations. To test these hypotheses, 31P MRS was used to noninvasively measure high energy phosphates and pH continuously during exhaustive single-leg knee extension exercise when subjects were euhydrated and hypohydrated. The employed ergometer was designed for MRS studies (i.e., fit within magnet bore and made of nonferrous materials) and isolated a relatively large muscle mass to enhance the signal to noise ratio.



Ten healthy physically active persons (5 men and 5 women), 21 to 40 years of age, participated in this study. The study was approved by the appropriate institutional review boards and all volunteers gave their voluntary and informed consent prior to participation.

Experimental Procedure

Following several practice sessions to familiarize the volunteers with the experimental procedures and to determine the appropriate exercise intensity for the experimental trials, the volunteers reported to the laboratory on two occassions separated by a minimum of one week. Upon arrival, at 1100 h to 1300 h, an initial nude body weight was obtained to establish baseline body weight. The volunteers then entered a hot room (40 C, 20% rh) to perform 2 to 3 h of moderate intensity treadmill and cycling exercise. For the euhydrated trial (EU), water was available ad libitum during the exercise. For the hypohydration trial (HY), drinking was restricted to produce a 4% body weight loss (BWL). The exercise mode, duration and intensity were held constant for each trial. In the event that the exercise protocol did not elicit the desired body weight loss, supplemental sauna exposure was included. Trial order was randomly assigned and balanced across subjects. After exercise, the volunteers were given a small standardized meal (approximately 400 kcal; 70% carbohydrate) and 200 ml of fruit juice. Three to eight hours of recovery separated the dehydration sessions and experimental testing in the magnetic resonance system. During the recovery period the volunteers were provided ad libitum access to water and other beverages that did not contain sugar or caffeine if they were performing the euhydration trial, but fluids were restricted to maintain the desired water deficit for the hypohydration trial. For the experimental trials, volunteers performed single-leg knee extension exercise to exhaustion while lying supine inside a whole-body 1.5 tesla magnetic resonance system (GE Medical Systems, Milwaukee, WI). The experimental setup is illustrated in Figure 1. Single leg knee extension exercise was performed at 37 contractions/min through a range of motion of approximately 110 to 140 knee extension. The resistance was determined from practice sessions and set to elicit exhaustion in approximately 4-5 min. The same resistance was used for both trials and was achieved by adding frictional resistance with known quantities of lead weight suspended over a *** kg flywheel (set in motion by the knee-extension exercise). An elastic cord returned the lever arm to the starting position after each knee extension motion. Force, range of motion and kick duration were measured by a computer-based data equisition system (Strawberry Tree, Sunnyvale, CA) interfaced to a force transducer and 360 potentiometer located in-line between the knee extension lever arm and the flywheel. The average power output was 19ñ3 Watts. The volunteers were instructed to perform the knee-extension task as long as possible. Endurance time was defined as the time when the power generated each kick declined 20% below the average value during the initial minute of exercise. The subjects were given verbal encouragement to produce maximal effort. Both legs were tested in each hydration condition and the results treated as independent observations. Prior to exhaustive exercise and at select times during recovery, measurements of muscle efficiency (left leg) or muscle strength (right leg) were obtained. For the muscle efficiency tests, subjects performed 6 knee-extensions within the 10 sec 31P MRS sampling periods 100 sec and 50 sec before exhaustive exercise, at 30 sec of recovery, and every min thereafter through 5 min of recovery. Muscle strength was measured by having the subjects perform a maximal voluntary isometric contraction for 5 sec with the knee at ~110 extension. The procedure was performed 100 sec and 50 sec before exhaustive exercise, every 30 sec of recovery for 2 min, and every minute thereafter through 5 min of recovery. 31P spectra were collected at rest and during exercise through an 11 cm coil (USAsia, Inc, Columbus, OH) placed over the quadriceps muscles. Data was acquired using a hard pulse 25.85 MHz excitation (pulse width 600 usec), Tr = 1,000 msec, spectral width 2,000 Hz and 1,024 sampled points. Ten FID signals were averaged producing 1 spectra every 10 seconds. Post processing consisted of 10 Hz line broadening, zero-filling to 4,096 points and Fourier transformation, followed by zero and first order phasing. Peak areas from the phosphocreatine (PCr), Pi and á-ATP peaks were used to determine phosphorous ratios. Muscle pH was calculated from the frequency shift between Pi and PCr using the following equation: pH = 6.73 + log10[(a-3.275)/5.685-a)] where a is the chemical shift from Pi to PC. Recovery kinetics for Pcr resynthesis were determined by calculating the time constant for the left leg PCr/á-ATP data. Recovery data were fit to a mono-exponential curve and time constant calculated from the derived rate constant. MRS system calibration was periodically verified using known standards.

Statistical Analysis

The data were analyzed using one- and two-way repeated measures analysis of variance, where appropriate. For all analyses, the data obtained from each leg were treated as independent measures. If fatigue was not achieved and/or spectra were of inadequate quality, the data for that leg were dropped from statistical analysis. Tukey's highly significant difference procedure was used to identify differences between means when statistical significance was achieved. Statistical significance was tested at the P<0.05 level. Data in the text are reported as means ñ sd.


Body Weight Loss

Prior to performing exercise in the hot room, body weights were 65.9ñ12.8 kg and 66.1ñ13.0 kg for EU and HY, respectively. The dehydration-rehydration procedures resulted in 0.6ñ0.7% and 4.0ñ0.5% BWL prior to the MRS tests.

Muscle Endurance

Figure 2 presents the individual leg and mean endurance times to exhaustion. Hypohydration reduced endurance time $ 8% (coefficient of variation for time to fatigue) in 12 of 19 of the trials performed and mean endurance was reduced (P<0.05) from 251ñ67 sec to 213ñ52 sec (15%). Four of 10 subjects had reduced endurance time in both legs when hypohydrated, while in 3 others only one leg was affected. For these three subjects, the reduced exercise performance occured in the second leg tested. For one subject, endurance time was reduced in one leg but not tested in the other leg due to technical problems. These results were similar to our pilot work (n=5 subjects) where 4-5% BWL reduced endurance in 8 of 10 trials and reduced (P<0.05) mean endurance time from 230ñ108 to 192ñ101 seconds (17%). These combined results demonstrate that hypohydration decreased mean endurance time (20 of 29 tests) by 15-17% using this exercise paradigm.

Muscle Strength

Figure 3 presents maximal voluntary contraction (MVC) data. Hypohydration did not alter pre-endurance exercise maximal isometric force. Hypohydrated persons produced a 16% higher (P<0.05) maximal isometric force 30 seconds after exhaustive exercise. No other difference between trials existed during recovery from exhaustive exercise. Furthermore, there were no relationships between the increased MVC at 30 sec post exercise and the reduction in endurance time when hypohydrated. 31P MRS. Figure 4 presents Pi/á-ATP, pH and Pi/PCr collected during all experimental trials and these variables were not altered by hydration. Pi/á-ATP were similar at rest, averaging 1.20ñ0.33 and 1.14ñ0.25 during EU and HY, respectively. During exhaustive exercise, the Pi/á-ATP rose progressively to peak values of 5.66ñ1.39 and 5.55ñ1.30 during EU and HY, respectively. Similarly, Pi/PCr rose from resting values averaging 0.17ñ0.05 to 3.79ñ1.83 and 3.46ñ1.54 at exhaustion during EU and HY, respectively. The pH fell progressively from 7.04ñ0.07 at rest to 6.49ñ0.33 at exhaustion during EU and HY. Similar to exercise data, hypohydration did not alter (P<0.07) the time constant of PCr synthesis after exhaustive exercise (EU=63ñ19 sec; HY=72ñ22 sec). To further investigate whether elevated levels of Pi or H+ could contribute to reduced endurance time, the data were subdivided to only compare spectra from trials when endurance times were reduced. Figure 5 presents this subset. Note that Pi/á-ATP and pH were similar (P>0.05) between EU and HY during rest and exercise. In contrast, the Pi/PCr ratio increased more rapidly and to a higher (P<0.05) level at the time of HY exhaustion during HY than EU. Examination of the individual data revealed that the higher Pi/PCr when hypohydrated was largely attributable to 3 of 10 trials, and there was no apparent relationship between the higher Pi/PCr ratios and reduced endurance times.


This study was the first to simultaneously examine the impact of hypohydration on skeletal muscle performance and muscle metabolism. The level of hypohydration studied is commonly achieved by athletes during competition. To minimize the likelihood of hypoglycemia and to replace some of the carbohydrate metabolized during the dehydration procedures, a small meal was given during the recovery period prior to the experimental trials. To isolate the effects of hypohydration on muscle from the potentially confounding effects of elevated body temperature, a minimum of three hours rest separated the heat exposures from experimental testing, and the MRS experiments were conducted in a cool room (~18 oC). In fact, no subjects reported any percieved heat strain and all felt cool. We found that hypohydration reduced muscular endurance by 15% but had no affect on muscle strength. These findings support earlier studies which reported that hypohydration can impair muscle endurance but had no affect on muscle strength. It also agrees with studies demonstrating that hypohydration can reduce aerobic endurance. Our results extend the findings of these earlier studies by separating the effects of hypohydration from the confounding effects of elevated body temperature, cardiovascular strain, heat exposure, and differing quantities of exercise prior to experimental testing. This study also demonstrated that hypohydration has no affect on recovery of muscle strength after exhausting exercise. During exercise, we employed 31P MRS to assess whether hypohydration would accelerate the accumulation of H+ or Pi during exhaustive exercise. These two variables were chosen as both have been shown to reduce cross-bridge formation and force production, and are the two variables within muscle often considered to be responsible for fatigue during high intensity exercise. We hypothesized that if hypohydration had direct effects on muscle metabolism then the hypohydration trials would likely be associated with elevated exercise H+ and/or Pi concentrations. The results of this study did not support this hypothesis, however, as Pi/á-ATP and pH responses to exercise were not affected by 4% BWL. The lone observation that suggested that hypohydration had an effect on muscle metabolism was the accelerated increase in Pi/PCr in the subgroup of trials with shortened time to fatigue. This would suggest that hypohydration required greater reliance on creatine kinase to sustain muscle ATP in these trials. The fact that Pi/PCr ratio was not consistently elevated even in this subgroup or correlated with maintenance of endurance time, however, would further support the contention that moderate levels of hypohydration had little or no affect on muscle metabolism. How hypohydration reduces muscle endurance remains an intriguing question. Alternative mechanisms within muscle include altered cell depolarization and changes in calcium release and/or uptake by the sarcoplasmic reticulum. Dehydration-induced changes in the ionic status of the T tubular lumen and intracellular compartments could contribute to the development of fatigue by negatively affecting the T tubular charge movement . Similarly, longer calcium transients might reduce calcium flux upon depolarization and reduce force production. The possiblity that elevated intracellular magnesium ions play a role appears unlikely as Costill and Saltin found no difference in intracellular magnesium concentration during exercise when euhydrated vs hypohydrated by 4% of initial body weight. An alternative explanation for the detrimental effects of hypohydration on muscle endurance is that hypohydration alters central nervous system function. In the subgroup of trials in which muscle strength was measured by performance of MVC before and after exercise, muscle endurance time was reduced (P<0.05) 14% yet volunteers were able to generate greater absolute force during the initial period of recovery; suggesting that the subjects were either less willing or unable to sustain voluntary concentric exercise when hypohydrated despite having adequate muscle strength. An unwillingness or inability to generate or maintain adequate CNS drive to the working muscle is thought to be responsible for the debilitating fatigue that accompanies many infections and illnesses, recovery from injury, and chronic fatigue syndrome. Hypohydration may impair performance in a similar manner. These conditions are characterized by an increased perception of effort during physical activity, yet the afflicted are capable of generating maximal force. Body water loss also increases perception of effort during physical activities, yet has no apparent effects on maximal strength. In addition, hypohydration is known to alter neuronal firing of osmoreceptive cells located in the organum vasculosum laminae terminalis and cells near the preoptic/anterior hypothalamic areas of the brain. Neuronal activation mediated by hypohydration might also alter the magnitude of corollary discharge from the motor cortex.
In summary, we found that moderate hypohydration:

  1. decreases skeletal muscle performance by reducing endurance by 15%
  2. does not alter muscle strength or recovery of muscle strength after exhaustive exercise
  3. does not alter pH and Pi/á-ATP response to exhaustive exercise

These findings clearly identify another physiologic system by which hypohydration adversely affects human exercise perfromance, however, the mechanisms for this action are unclear.