Since wrestlers compete in weight categories, there is motivation for some wrestlers to temporarily lose weight to be eligible for a lower weight class. Steen and Brownell (17) reported that 89% of the collegiate wrestlers they studied lost weight to achieve their weight class with an average weight loss of 4.4 kg over 3 d. The majority (52%) practiced total food restriction at least once per week. Only 7% of these college wrestlers reported that they never restricted food during the season.
The effect of acute weight loss on physical performance in wrestlers has been inconsistent between research groups and even within the same study using different performance tests(5-8,14,22). The studies may have found different results due to different dietary and dehydration interventions. Since dehydration was typically used with energy restriction, it is not possible to determine what specific behavior affected performance. Comparison among studies is difficult due to the variety of performance tests utilized. In general, those studies which used tests lasting less than 1 min did not find reduced performance while those lasting between 1 and 6 min found effects of weight loss on performance. Hickner et al. (6) developed an intermittent, intense arm ergometer test that was perceived as more similar to the demands of a wrestling match than a 30-s arm Wingate test, a treadmill, or an arm crank ˙VO2max test. The duration of this test, 6 min, and the intermittent nature of the effort seem to be more similar to a wrestling match than many of the brief anaerobic tests.
Horswill et al. (7) suggested that alteration in acid/base status mediated the decrease in anaerobic performance coincident with weight loss in wrestlers in their study. Evidence for this hypothesis was the significantly greater reduction in resting base excess and anaerobic performance in wrestlers who lost weight while consuming a low carbohydrate diet (41.9%) compared with those who consumed a higher carbohydrate diet(65.9%). Other possible mechanisms for impaired performance with energy restriction include reduced glycogen or creatine phosphate stores, altered enzyme activity, and impaired sacroplasmic reticulum function.(20).
Many of these studies have examined performance while the subjects were in a negative energy balance state. However, wrestlers do not compete just after the weigh-in. They typically have between 3 and 20 h to eat and drinkad libitum prior to their match. There is some evidence that this is not enough time for total recovery. Houston et al. (8) confirmed that a 3 h ad libitum feeding period did not totally replenish muscle glycogen. Concentration of muscle glycogen remained 38% depressed relative to levels prior to weight loss.
Little research has focused on the diet during the period between weigh-in and the competition. Data collected by Steen and McKinney(18) indicated that wrestlers did not choose high carbohydrate foods. Pre-match foods most often cited by these wrestlers were steak, eggs, and protein bars. These dietary preferences after weight loss may be counterproductive for an athlete attempting to regain muscle glycogen and performance over a short time period.
The purpose of this study was to determine the effect of 3 d of energy restriction without dehydration on anaerobic performance in wrestlers. In addition, we examined the effects of two refeeding diets differing in carbohydrate content on performance. Indicators of blood acid/base status were measured to determine whether they played a role in any performance changes.
Twelve collegiate wrestlers who had been informed of all risks volunteered to participate in the study that had been approved by the Institutional Review Board. The subjects were tested from 3-4 wk prior to the beginning of the competitive season. Descriptive information on the subjects is given inTable 1. The subjects kept diet records during the 2 d prior to the experimental period. The athletes' body weight was measured and bodyfat was calculated from skinfolds measurements using the equation of Lohman (13) 2 d prior to the dietary intervention. A practice trial of the anaerobic performance test was also completed on this day to acquaint the subjects with the equipment and procedure. Another performance test (TEST1) was performed the day prior to the intervention. A venipuncture in an arm vein for blood collection was performed within 5 min before and at approximately 1 min after each of the performance tests except the practice trial.
A 72 h formula hypoenergy diet phase began immediately after the baseline performance test. Each individual was given a diet containing 18 kcal·kg-1 with 60% carbohydrate, 20% protein, and 20% fat(Exceed, donated by Ross Laboratories, mixed with skim milk). They picked up their diet each day and were weighed to the nearest 0.1 kg. The subjects were specifically instructed to consume copious noncaloric fluids so that dehydration would not confound the results. All subjects were members of the same collegiate wrestling team and performed their customary team workout during the study. They were asked not to perform any other exercise during the study. Another performance test (TEST2) was scheduled at the end of the 72 h weight-loss period.
The 5-h refeeding period began just after TEST2. All subjects remained sedentary and in the laboratory during this period. They were all fed a total of 21 kcal·kg-1 over the 5 h in a scheduled pattern: 20% fed immediately post, 20% 30 min post, 20% 60 min post, 20% 90 min post, 10% 120 min post, and 10% 150 min post. Subjects were allowed water ad libitum during the recovery period. A final performance test (TEST3) was performed at the end of the 5-h recovery period.
The subjects had been assigned to two groups by listing the initial body weights from highest to lowest and then assigning to alternate groups. Both groups received the same diet during the energy restriction period but different diets during the refeeding period. One group received a high carbohydrate diet (75% carbohydrate, 15% fat, 10% protein) while the other consumed a moderate carbohydrate diet (47% carbohydrate, 37% fat, and 16% protein) during this 5-h period. The diet for the high carbohydrate group (HC) consisted of Exceed Sports Meal in one cup and Exceed High Carbohydrate (both donated by Ross Laboratories) in another. The diet of the moderate carbohydrate group (MC) contained Exceed Sports Meal mixed with Pulmocare (28% carbohydrate, 55.2% fat, and 16.7% protein), and an artificially sweetened beverage colored to mimic the Exceed High Carbohydrate drink in another cup.
Performance test. The performance test was a modification of that designed and validated by Hickner et al. (6) as a test most similar to a wrestling match compared with other traditional anaerobic tests. A Monark bike ergometer was mounted on a heavy table with the pedals replaced with hand grips. The ergometer was integrated with a microcomputer that was used to collect and record the power output data. This ergometer, microcomputer and software set up has been described by Williams et al.(24). The software computes power output for each one-half pedal revolution and calculates peak power, average power, and work done for each high-intensity interval. The subjects' legs were strapped to the seat to isolate use of the arms during the test. The resistance on the ergometer was set at 0.04 kg·kg-1 body weight. The test began with 15 s of maximal effort cranking against the resistance followed by a 20 s unloaded low-intensity interval. This sequence was repeated eight times. Pilot work was conducted on this test in our laboratory on six subjects tested three times with 1 d between each test. Analysis of the total work done at each test showed a r of 0.92 between the first and second test and an r of 0.96 for the comparison between the second and third test. Therefore, we used one practice trial before collecting the dependent measures for performance in this study. The high reliability of the performance test in this pilot study reduced the need for a control group in the study design.
Blood collection and analysis. Blood was withdrawn using a 3-ml heparinized syringe and 22-gauge needle from the antecubital forearm vein within 5 min prior to the start of each performance test and within 1 min after the performance test. All air was expelled from the syringe. Half of the sample was transferred into a test tube and a hematocrit tube was immediately collected. The blood in the test tube was centrifuged and the plasma was frozen for later analysis of lactate and glucose in duplicate using an automated glucose and lactate analyzer (YSI 2300, Yellow Springs, OH). The syringe with the remaining blood was plugged and placed on ice. This sample was analyzed for PCO2, and pH with an automated blood gas analyzer(1306 Instrumentation Laboratories, Lexington MA) within 30 min. Hemoglobin was measured with an automated system (Instrumentation Laboratories 482 co-oximeter, Laxington, MA). Bicarbonate and base excess were calculated from pH and PCO2 values using the Siggard-Anderson alignment nomogram. Since the blood samples are venous rather than arterial or arterialized, the lactate and pH values reported may underestimate the degree of change from resting samples. However, this should not affect differences between groups.
Statistical analysis. The performance data was analyzed with a repeated measures analysis of variance (diet and test as factors with two and three levels, respectively). The blood measurements were analyzed with three-way analysis of variance (group, test, and pre/post performance test as factors). The Newman-Kuels post-hoc test was used to discriminate between averages when a significant F ratio was calculated. AP-value < 0.05 was considered significant while thoseP-values < 0.1 are reported to indicate trends.
There were no significant differences between groups for age, height, body weight, percent body fat, or initial work done on anaerobic power test. Analysis of the diet records for the 2 d prior to the dietary intervention showed no significant difference in percent carbohydrate, fat, or protein intake between groups (Table 2). There was a trend(P = 0.09) for HC to eat more energy than MC. The average body weight was relatively stable with a slight increase of 0.11 kg over the 2 d before the energy restriction period. Five subjects showed a slight increase, five a slight decrease, one had no change, and one had no body weight data on the 2 d before the intervention. There was no difference in body weight change for the two groups over these 2 d. As planned, all athletes lost weight during the energy restriction period. There was an average reduction in body weight of 2.4 kg ± 1.0 to 96.7% of initial body weight with no difference between the groups. The average gain in body weight over the 5 h refeeding period was 1.00 ± 0.07 kg for HC and 0.83 ± 0.21 for MC with no difference between the groups.
Work and power. There was a significant effect of test on total work done during the high-intensity portion of the performance test(Fig. 1). All subjects did less work at TEST2 than they had done at TEST1. The average work done at TEST2 when groups were collapsed was significantly different than TEST1 at 92.4% of this value. The groups tended to be different over the tests (P = 0.10). The calculated power for this statistical test of interaction of test and group for our subject number was 0.7. The trend for interaction between test and group is reflected in the fact that the work done by HC subjects was 99.1% of TEST1 at TEST3 but that of MC remained similar to the reduced level observed at TEST2(91.5% of TEST1). Five of the six HC subjects had a increase in work done from TEST2 to TEST3 while only two of the six in MC did an increased amount of work at the last test compared with TEST2.
The results for average power during the high-intensity intervals of the performance test was similar to that for total work done(Fig. 2). There was a significant effect of test withpost-hoc test, indicating that average power produced was significantly reduced at TEST2. The group by trial interaction had aP-value of 0.15. There were no significant effects of any factor when peak power was analyzed (Fig. 3).
Blood lactate and glucose. Statistical analysis of plasma lactate indicated that there was a significant effect of test, pre/post, test by group interaction, test by pre/post interaction, and a tendency for a group effect(P = 0.08, Fig. 4). The pre/post effect was most obvious with an average resting lactate of 1.72 mmol·l-1 when collapsing the values from the three tests and post lactate of 21.91 mmol·l-1. The significant effect of test related to the reduced lactate at TEST2, especially at the post blood sample compared to that at the other performance test (22.6, 20.0, and 23.11 mmol·l-1 for post lactate collapsed for groups at TEST1, 2, and 3, respectively). The differences in plasma lactate between groups widened over the tests; this resulted in the significant interaction between groups over the tests.
Plasma glucose significantly increased at each performance test from the pre to the post measurement (4.88 and 5.25 mmol·l-1, respectively) when collapsed for all tests (Fig. 5). The concentration of glucose was significantly higher for the high carbohydrate group (5.24 mmol·l-1) than the low carbohydrate group (4.80 mmol·l-1) when the values were collapsed by group. Although there was a significant effect of test, none of the averages were different bypost-hoc analysis.
Both hemoglobin and hematocrit increased significantly from the pre to post exercise sample. There was no main effect of group or test on either factor although there was a trend for an interaction between test and group for hematocrit (Table 3).
Acid-base variables. The performance test caused the blood pH to significantly drop (grand mean 7.35 and 7.06 for pre and post, respectively,Table 4). There was an effect of dietary treatment in that the pH was significantly lower overall for HC as compared to MC (7.18 and 7.23, respectively). Analysis of the significant effect of test indicated that the pH was higher in TEST2 compared to TEST1 (7.19, 7.22, 7.21 for TEST 1, 2, and 3 respectively). The significant interaction between group and pre/post was indicated by less of a decrease in pH from pre to post in MC (7.35 and 7.11, respectively) versus HC (7.35 and 7.02, respectively).
The only significant factor in the analysis of blood bicarbonate was the significant reduction in the post sample as compared to the sample taken just before the performance tests (Table 4). Blood base excess significantly dropped when post samples were compared to those at pre (-12.77 and 2.82 mmol·l-1, respectively, Table 4). There was a significant effect of test on this factor but post-hoc analysis did not reveal a difference between means of base excess for the different performance tests.
This study showed that acute energy restriction without dehydration impairs performance of an anaerobic exercise test in wrestlers. Although a variety of studies have been done looking at performance in wrestlers, they are difficult to compare since a variety of weight loss methods have been employed. Several studies reported no effect of acute weight loss (5) or weight loss over the season (15,22,23) on performance tests that included muscle strength and arm or leg Wingate tests in wrestlers. Some studies find no effect of weight loss on some performance tests but decrements in others (8,22).
Carbohydrate content of the weight loss diet of wrestlers has been shown to influence physical performance in two studies(7,14). No change was noted in performance of a 30-s Wingate test (14) or a 6-min intermittent arm ergometer test (7) if the subjects had eaten a high carbohydrate diet of 70% (14) or 65.9%(7), but performance on the same tests was inferior if the wrestlers had eaten a lower carbohydrate diet (55% and 41.9%, respectively). We have previously shown an effect of carbohydrate content of a weight reduction diet in resistance weight trainers. Energy restriction reduced quadriceps endurance in those who consumed a 50% carbohydrate hypoenergy diet but was maintained after weight loss if the dietary carbohydrate was 70% of the energy (21).
The 60% carbohydrate diet used in this study did not prevent a drop in anaerobic performance of the wrestlers. It is possible that a higher carbohydrate or energy diet may have maintained performance. However, the evidence from this study and others suggests that anaerobic performance of wrestlers and resistance weight trainers is most likely to be impaired if the carbohydrate is less than 60% of the energy in a weight-reducing diet.
The mechanism for an effect of dietary carbohydrate on anaerobic performance is unclear. The role of muscle glycogen in fatigue during high-intensity exercise is controversial since brief, anaerobic exercise does not typically deplete this muscle fuel. Houston et al.(9) confirmed a 21.5% reduction in muscle glycogen as a result of a wrestling match. Muscle glycogen is likely to be even more depressed with the exercise protocol used in this study since post-bout blood lactate levels in the Houston et al. (9) study were less than half of those found with this study. The additional reduction in muscle glycogen which is likely from 3 d of energy restriction may have contributed to muscle glycogen becoming a limiting factor for this exercise protocol. Muscle glycogen was not assessed in this study, but the reduced post-exercise lactate concentrations after the exercise bout in TEST2 suggest that these concentrations were lower after weight loss.
Although it is valuable to determine the effect of acute energy restriction, it is recognized that wrestlers almost never compete just after weigh-in. They are typically given at least 3 h to eat and drink ad libitum. Thus, a critical question is whether the performance decrements can be overcome by refeeding. Houston et al. (8) used a 3-h refeeding period in wrestlers after a 4-d weight-loss period. The reduction in muscle torque noted following weight loss was not relieved following the refeeding period. The diet during this refeeding period was not controlled or measured.
The content of the diet following weight loss may be critical in the rate of recovery of performance in athletes. The rate of muscle glycogen resynthesis after depletion by exercise has been shown to be affected by dietary intake. For example, Blom et al. (1) found that glycogen resynthesis is dependent on amount of carbohydrate ingested following aerobic exercise. Although most of the research on recovery of muscle glycogen following exercise has been done using aerobic exercise, Pascoe et al.(16) confirmed that glycogen resynthesis after a resistance exercise bout was higher over 6 h when the subjects consumed 1.5 g·kg-1 carbohydrate immediately and then at 1 h after exercise compared with those who consumed only water. Although muscle glycogen and anaerobic performance are not necessarily directly associated(4), it is possible that enhanced resynthesis of muscle glycogen would allow recovery of physical performance. Muscle glycogen and performance were associated in the wrestlers studied by Houston et al.(8), who produced less muscle torque when they had lower muscle glycogen after weight loss.
The performance of the wrestlers we studied tended to be affected by the diet they were fed between the simulated weigh-in and the final performance test 5 h after this. A high dietary carbohydrate intake tended to be helpful in restoring performance on this anaerobic test. Several other studies have found a benefit of a high carbohydrate intake on anaerobic performance(3,7,10,12,14). Jenkins et al.(10) reported that a 83% carbohydrate diet for 3 d allowed subjects to do more work during five 60-s sprints with 5-min rest between than 3 d of a 12% carbohydrate diet. A 59% carbohydrate diet resulted in intermediate amount of work done, significantly different from the low carbohydrate diet.
It is difficult to compare the particular diets used in this study with those of competing wrestlers since little has been published on the weight reduction diet and even less on the diet of wrestlers between the weigh-in and the match. The seven wrestlers studied by Webster et al.(22) consumed an average of 1080 kcal over a 36-h weight-loss period with 48.5% as carbohydrate. The judo and wrestling athletes studied by Fogelholm et al. (2) who were asked to lose 6% of their body weight over 2.4 d using diet and fluid restriction chose a diet of about 1300 kcal and 55% carbohydrate. These same athletes kept food records during 5 h after the weight-loss period. They consumed an average of 285 g of carbohydrate, which was approximately 56% of the 2020 kcal they ate. A pilot study (unpublished) conducted by the first author showed that the diet consumed between weigh-in and match by six collegiate wrestlers, each studied at one to two matches for a total of 11 food records varied from 51% to 72% carbohydrate. Keizer et al. (11) demonstrated that athletes may not consume the optimum amount of carbohydrate during recovery from strenuous exercise. The ad libitum dietary carbohydrate intake during 5 h after exhaustive aerobic exercise for subjects in that study was only 44.8%. These athletes had less muscle glycogen resynthesis at 5 h compared with groups fed a controlled, higher carbohydrate diet (>70% of energy). Thus, the 47%, lower carbohydrate diet we fed to the MC group may be close to that consumed by some athletes after weight loss or strenuous exercise. These athletes may not be achieving maximum recovery of glycogen and performance.
Greenhaff et al. (3) and Horswill et al.(7) suggested that the negative impact of a lower carbohydrate diet on performance of high-intensity exercise is mediated by changes in the body's acid/base environment. These studies identified a reduced resting base excess after 3-4 d of a reduced carbohydrate diet. They hypothesize that this would reduce the ability of the blood to accept hydrogen ions produced in the muscle during anaerobic work and lead to earlier fatigue. We did not find a significant effect of the 3 d energy restriction using a 60% carbohydrate diet on resting base excess nor a differential effect of the two diets over the 5 h refeeding period on base excess. Base excess tended to increase similarly for both groups during this period. Thus, neither the reduced performance after weight loss nor the tendency for improved performance in the group fed higher carbohydrate during recovery can be explained by the acid/base parameters measured.
Plasma lactate values after this performance test were higher than those reported by Horswill et al. (7) for a similar protocol using an isokinetic bike and a longer low-intensity interval period (30 s). Our average post blood lactate, ranging from 17.5 to 26.4 mmol·l-1, are closer to the average of 20 mmol·l-1 reported by Triplett et al. (19) shortly after a tournament wrestling match than the 14 mmol·l-1 averages found after the anaerobic test used by Horswill et al. (7). Thus, this performance test appeared to cause a similar metabolic stress as a wrestling match.
In conclusion, acute weight loss via energy restriction without dehydration impaired anaerobic performance in college wrestlers. Performance tended to return to initial levels when the athletes were fed a 75% carbohydrate diet over 5 h after weight loss but did not recover when the athletes ate a 47% carbohydrate diet. Thus, we recommend that wrestlers who lose weight using energy restriction to meet a weight class consume a high carbohydrate diet between weigh-in and the match.
1. Blom, P. C., A. T. Hostomark, O. Vaage, K. R. Kardel, and S. Maehlum. Effect of different post-exercise sugar diets on the rate of muscle glycogen synthesis. Med. Sci. Sports Exerc.
2. Fogelholm, G. M., R. Koskinenk, J. Laakso, T. Rankinen, and I. Ruokonen. Gradual and rapid weight loss: effects on nutrition and performance in male athletes. Med. Sci. Sports Exerc.
3. Greenhaff, P. L., M. Gleeson, and R. J. Maughan. Diet-induced acidosis and the performance of high intensity exercise in man.J. Appl. Physiol.
4. Grisdale, R. K., I. Jacobs, and E. Cafarelli. Relative effects of glycogen depletion and previous exercise on muscle force and endurance capacity. J. Appl. Physiol.
5. Guastella, P., J. Wygand, K. Davy, F. Pizza. The effects of rapid weight loss on anaerobic power in high school wrestlers. Med. Sci. Sports Exerc.
6. Hickner, R. C., C. A. Horswill, M. M. Welker, J. Scott, J. N. Roemmich, and D. L. Costill. Test development for the study of physical performance in wrestlers following weight loss. Int. J. Sports Med.
7. Horswill, C. A., R. C. Hickner, J. R. Scott, D. L. Costill, and D. Gould. Weight loss, dietary carbohydrate modifications, and high intensity, physical performance. Med. Sci. Sports Exerc.
8. Houston, M. E., D. A. Marrin, H. J. Green, and J. A. Thomson. The effect of rapid weight loss on physiological functions in wrestlers. Phys. Sportsmed.
9. Houston, M. E., M. T. Sharrett, and R. W. Bruce. Glycogen depletion and lactate responses in freestyle wrestling. Can. J. Appl. Sports Sci.
10. Jenkins, D. G., J. Palmer, and D. Spillman. The influence of dietary carbohydrate on performance of supramaximal intermittent exercise. Eur. J. Appl. Physiol.
11. Keizer, H. A., H. Kuipers, G. van Kranenburg, and P. Geurten. Influence of liquid and solid meals on muscle glycogen resynthesis, plasma fuel hormone response, and maximal physical working capacity.Int. J. Sport Med.
12. Lambert, C. P., M. G. Flynn, J. B. Boone, T. J. Michaud, and J. Rodriquez-Zayas. Effects of carbohydrate feeding on multiplebout resistance exercise. J. Appl. Sport Sci. Res.
13. Lohman, T. G. Skinfolds and body density and their relationship to body fatness: a review. Hum. Biol.
14. McMurray, R. G., C. R. Proctor, and W. L. Wilson. Effect of caloric deficit and dietary manipulation on aerobic and anaerobic exercise. Int. J. Sports Med.
15. Park, S. H., J. N. Roemmich, and C. A. Horswill. A season of wrestling and weight loss by adolescent wrestlers: effect on anaerobic arm power. J. Appl. Sport Sci. Res.
16. Pascoe, D. D., D. L. Costill, R. A. Robergs, J. A. Davis, W. J. Fink, and D. R. Pearson. Effects of exercise mode on muscle glycogen restorage during repeated days of exercise. Med. Sci. Sports Exerc.
17. Steen, S. and K. Brownell. Patterns of weight loss and regain in wrestlers: has the tradition changed? Med. Sci. Sports Exerc.
18. Steen, S. N. and S. McKinney. Nutrition assessment of college wrestlers. Physician Sportsmed.
19. Triplett, N. T., W. J. Kraemer, A. C. Fry, S. J. Fleck, L. P. Koziris, and J. M. Lynch. Hormonal responses to tournament wrestling.Med. Sci. Sports Exerc.
20. Walberg, J. L. Aerobic exercise and resistance weight-training during weight reduction: implications for obese persons and athletes. Sports Med.
21. Walberg, J. L., M. K. Leidy, D. J. Sturgill, D. E. Hinkel, S. J. Ritchey, and D. R. Sebolt. Macronutrient content of a hypoenergy diet affects nitrogen retention and muscle function in weight lifters.Int. J. Sports Med.
22. Webster, S., R. Rutt, and A. Weltman. Physiological effects of a weight loss regimen practiced by college wrestlers. Med. Sci. Sports Exerc.
23. Widerman, P. M. and R. D. Hagen. Bodyweight loss in a wrestler preparing for competition: a case report. Med. Sci. Sports Exerc.
24. Williams, J., W. Barnes, J. Signorile. A constant-load ergometer for measuring peak power and fatigue. J. Appl. Physiol.