The role of anaerobic ability in middle distance running performance

The purpose of this study was to assess the relationship between anaerobic ability and middle distance running performance. Ten runners of similar performance capacities (5 km times: 16.72, SE 0.2 min) were examined during 4 weeks of controlled training. The runners performed a battery of tests each week [maximum oxygen consumption (VO2max), vertical jump, and Margaria power run] and raced 5 km three times (weeks 1, 2, 4) on an indoor 200-m track (all subjects competing). Regression analysis revealed that the combination of time to exhaustion (TTE) during the VO2max test (r2 = 0.63) and measures from the Margaria power test (W.kg-1, r2 = 0.18; W, r2 = 0.05) accounted for 86% of the total variance in race times (P less than 0.05). Regression analysis demonstrated that TTE was influenced by both anaerobic ability [vertical jump, power (W.kg-1) and aerobic capacity (VO2max, ml.kg-1.min-1)]. These results indicate that the anaerobic systems influence middle distance performance in runners of similar abilities.     

Longitudinal associations between anaerobic threshold and distance running performance

Longitudinal alterations in anaerobic threshold (AT) and distance running performance were assessed three times within a 4-month period of intensive training, using 20 male, trained middle-distance runners (19-23 yr). A major effect of the high intensity regular intensive training together with 60- to 90-min AT level running training (2 d X wk-1) was a significant increase in the amount of O2 uptake corresponding to AT (VO2 AT; ml O2 X min-1 X kg-1) and in maximal oxygen uptake (VO2max; ml O2 X min-1 X kg-1). Both VO2 AT and VO2max showed significant correlations (r = -0.69 to -0.92 and r = -0.60 to -0.85, respectively) with the 10,000 m run time in every test. However, further analyses of the data indicate that increasing VO2 AT (r = -0.63, P less than 0.05) rather than VO2max (r = -0.15) could result in improving the 10,000 m race performance to a larger extent, and that the absolute amount of change (delta) in the 10,000 m run time is best accounted for by a combination of delta VO2 AT and delta 5,000 m run time. Our data suggest that, among runners not previously trained over long distances, training-induced alterations in AT in response to regular intensive training together with AT level running training may contribute significantly to the enhancement of AT and endurance running performance, probably together with an increase in muscle respiratory capacity.     

Metabolism of rats running up and down an incline

The purpose of these experiments was to determine oxygen consumption (VO2) in rats as a function of treadmill speed (10, 20, 30, 40, and 50 m . min-1) as they ran on the level and up and down a 16 degree (17.8%) incline. The slopes of the regression lines relating VO2 (ml O2 . kg-1 . min-1) to running speed (m . min-1) were linear for all three inclines. The regression slope for uphill runners (y = 1.25x + 47.7) was greater than the regression slopes for level (y = 0.88x + 41.2) (P less than 0.025) or downhill (y = 0.68x + 39.7) (P less than 0.005) runners, and the regression slope for level runners was greater than that for downhill runners (P less than 0.10). All VO2 measurements were submaximal. In conclusion, incline has a significant effect on the metabolism of rats running on a motor-driven treadmill.     

Blood glucose responses in humans mirror lactate responses for individual anaerobic threshold and for lactate minimum in track tests

The equilibrium point between blood lactate production and removal (La-(min)) and the individual anaerobic threshold (IAT) protocols have been used to evaluate exercise. During progressive exercise, blood lactate [La-]b, catecholamine and cortisol concentrations, show exponential increases at upper anaerobic threshold intensities. Since these hormones enhance blood glucose concentrations [Glc]b, this study investigated the [Glc] and [La-]b responses during incremental tests and the possibility of considering the individual glucose threshold (IGT) and glucose minimum (Glc(min)) in addition to IAT and La-(min) in evaluating exercise. A group of 15 male endurance runners ran in four tests on the track 3000 m run (v3km); IAT and IGT - 8 x 800 m runs at velocities between 84% and 102% of v3km; La-(min) and Glc(min) - after lactic acidosis induced by a 500-m sprint, the subjects ran 6 x 800 m at intensities between 87% and 97% of v3km; endurance test (ET) - 30 min at the velocity of IAT. Capillary blood (25 microl) was collected for [La-]b and [Glc]b measurements. The IAT and IGT were determined by [La-]b and [Glc]b kinetics during the second test. The La-(min) and Glc(min) were determined considering the lowest [La-] and [Glc]b during the third test. No differences were observed (P < 0.05) and high correlations were obtained between the velocities at IAT [283 (SD 19) and IGT 281 (SD 21) m. x min(-1); r = 0.096; P < 0.001] and between La-(min) [285 (SD 21)] and Glc(min) [287 (SD 20) m. x min(-1) r = 0.77; P < 0.05]. During ET, the [La-]b reached 5.0 (SD 1.1) and 5.3 (SD 1.0) mmol x l(-1) at 20 and 30 min, respectively (P > 0.05). We concluded that for these subjects it was possible to evaluate the aerobic capacity by IGT and Glc(min) as well as by IAT and La-(min).     

Specificity of physiological adaptation to endurance training in distance runners and competitive walkers

The present study was designed to evaluate the specificity of physiological adaptation to extra endurance training in five female competitive walkers and six female distance runners. The mean velocity (v) during training, corresponding to 4 mM blood lactate [onset of blood lactate accumulation (OBLA)] during treadmill incremental exercise (training v was 2.86 m.s-1, SD 0.21 in walkers and 4.02 m.s-1, SD 0.11 in runners) was added to their normal training programme and was performed for 20 min, 6 days a week for 8 weeks, and was called extra training. An additional six female distance runners performed only their normal training programme every day for about 120 min at an exercise intensity equivalent to their lactate threshold (LT) (i.e. a running v of about 3.33 m.s-1). After the extra training, there were statistically significant increases in blood lactate variables (i.e. oxygen uptake (VO2) at LT, v at LT, VO2 at OBLA, v at OBLA; P less than 0.05), and running v for 3,000 m (P less than 0.01) in the running training group. In the walking training group, there were significant increases in blood lactate variables (i.e., v at LT, v at OBLA; P less than 0.05), and walking economy. In contrast, there were no significant changes in lactate variables, running v and economy in the group of runners which carried out only the normal training programme. It is suggested that the changes in blood lactate variables such as LT and OBLA played a role in improving v of both the distance runners and the competitive walkers.(ABSTRACT TRUNCATED AT 250 WORDS)     

Calculation of record-time in the 800-meter run: predictive value of Margaria's equation

Margaria's equation (1976)--describing the relationship between the minimum time necessary to cover a distance equal or longer than 1,000 m (record-time TR) and the maximal oxygen consumption (VO2 max)--has been modified in order to be applied to the calculation of TR in the 800 m foot race. Fifteen subjects participated in this study (VO2 max = 63 +/- 3.5 ml O2 X kg-1 X min-1, measured TR = 131 +/- 10 seconds). It has been found the TR calculated from Margaria's equation (TRc) are underestimated (TRc = 104 +/- 10 seconds). By taking into account the actual energy cost of running (0.19 ml O2 X kg-1 X m-1) and the kinetics of VO2 at the onset of exercise, TRc averaged 133 +/- 8.5 seconds. Moreover, the relationship between TRc and measured TR (TRm) is highly significant (TRc = 50.4 + 0.65 TRm; r = 0.75; P less than 0.01). These results validate Margaria's equation modifications.     

Performance and metabolic responses to a high caffeine dose during prolonged exercise.

The present study examined whether a high caffeine dose improved running and cycling performance and altered substrate metabolism in well-trained runners. Seven trained competitive runners [maximal O2 uptake (VO2max) 72.6 +/- 1.5 ml.kg-1.min-1] completed four randomized and double-blind exercise trials at approximately 85% VO2max; two trials running to exhaustion and two trials cycling to exhaustion. Subjects ingested either placebo (PL, 9 mg/kg dextrose) or caffeine (CAF, 9 mg/kg) 1 h before exercise. Endurance times were increased (P less than 0.05) after CAF ingestion during running (PL 49.2 +/- 7.2 min, CAF 71.0 +/- 11.0 min) and cycling (PL 39.2 +/- 6.5 min, CAF 59.3 +/- 9.9 min). Plasma epinephrine concentration [EPI] was increased (P less than 0.05) with CAF before running (0.22 +/- 0.02 vs. 0.44 +/- 0.08 nM) and cycling (0.31 +/- 0.06 vs. 0.45 +/- 0.06 nM). CAF ingestion also increased [EPI] (P less than 0.05) during exercise; PL and CAF values at 15 min were 1.23 +/- 0.13 and 2.51 +/- 0.33 nM for running and 1.24 +/- 0.24 and 2.53 +/- 0.32 nM for cycling. Similar results were obtained at exhaustion. Plasma norepinephrine was unaffected by CAF at rest and during exercise. CAF ingestion also had no effect on respiratory exchange ratio or plasma free fatty acid data at rest or during exercise. Plasma glycerol was elevated (P less than 0.05) by CAF before exercise and at 15 min and exhaustion during running but only at exhaustion during cycling. Urinary [CAF] increased to 8.7 +/- 1.2 and 10.0 +/- 0.8 micrograms/ml after the running and cycling trials.(ABSTRACT TRUNCATED AT 250 WORDS)     

Comparison of mathematically determined blood lactate and heart rate “threshold” points and relationship with performance

The purpose of this study was to investigate the relationship between threshold points for heart rate (Thfc) and blood lactate (Thla) as determined by two objective mathematical models. The models used were the mono-segmental exponential (EXP) model of Hughson et al. and the log-log (LOG) model of Beaver et al. Inter-correlations of these threshold points and correlations with performance were also studied. Seventeen elite runners (mean, SD = 27.5, 6.5 years; 1.73, 0.05 m; 63.8, 7.3 kg; and maximum oxygen consumption of 67.8, 3.7 ml.kg-1.min-1) performed two maximal multistage running field tests on a 183.9-m indoor track with inclined turns. The initial speed of 9 km.h-1 (2.5 m.s-1) was increased by 0.5 km.h-1 (0.14 m.s-1) every lap for the fc test and by 1 km.h-1 (0.28 m.s-1) every 4 min for the la test. After fitting the la or the fc data to the two mathematical models, the threshold speed was assessed in the LOG model from the intersection of the two linear segments (LOG-la; LOG-fc) and in the EXP model from a tangent point (TI-la; TI-fc). Thla and Thfc speeds computed with the two models were significantly different (P less than 0.001) and poorly correlated (LOG-la vs LOG-fc: r = 0.36, TI-la vs TI-fc: r = 0.13). In general, Thfc were less well correlated with performance than Thla. With two different objective mathematical models, this study has shown significant differences and poor correlations between Thla and Thfc.(ABSTRACT TRUNCATED AT 250 WORDS)     

Energetics of locomotion in African pygmies

The energy cost of walking (Cw) and running (Cr), and the maximal O2 consumption (VO2max) were determined in a field study on 17 Pygmies (age 24 years, SD 6; height 160 cm, SD 5; body mass 57.2 kg, SD 4.8) living in the region of Bipindi, Cameroon. The Cw varied from 112 ml.kg-1.km-1, SD 25 [velocity (v), 4 km.h-1] to 143 ml.kg-1.km-1, SD 16 (v, 7 km.h-1). Optimal walking v was 5 km.h-1. The Cr was 156 ml.kg-1.km-1, SD 14 (v, 10 km.h-1) and was constant in the 8-11 km.h-1 speed range. The VO2max was 33.7 ml.kg-1.min-1, i.e. lower than in other African populations of the same age. The Cr and Cw were lower than in taller Caucasian endurance runners. These findings, which challenge the theory of physical similarity as applied to animal locomotion, may depend either on the mechanics of locomotion which in Pygmies may be different from that observed in Caucasians, or on a greater mechanical efficiency in Pygmies than in Caucasians. The low Cr values observed enable Pygmies to reach higher running speeds than would be expected on the basis of their VO2max.     

Effect of endurance training on excessive CO2 expiration due to lactate production in exercise

We attempted to determine the change in total excess volume of CO2 output (CO2 excess) due to bicarbonate buffering of lactic acid produced in exercise due to endurance training for approximately 2 months and to assess the relationship between the changes of CO2 excess and distance-running performance. Six male endurance runners, aged 19-22 years, were subjects. Maximal oxygen uptake (VO2max), oxygen uptake (VO2) at anaerobic threshold (AT), CO2 excess and blood lactate concentration were measured during incremental exercise on a cycle ergometer and 12-min exhausting running performance (12-min ERP) was also measured on the track before and after endurance training. The absolute magnitudes in the improvement due to training for CO2 excess per unit of body mass per unit of blood lactate accumulation (delta la-) in exercise (CO2 excess.mass-1.delta la-), 12-min ERP, VO2 at AT (AT-VO2) and VO2max on average were 0.8 ml.kg-1.l-1.mmol-1, 97.8 m, 4.4 ml.kg-1. min-1 and 7.3 ml.kg-1.min-1, respectively. The percentage change in CO2 excess.mass-1.delta la- (15.7%) was almost same as those of VO2max (13.7%) and AT-VO2 (13.2%). It was found to be a high correlation between the absolute amount of change in CO2 excess.mass-1.delta la-, and the absolute amount of change in AT-VO2 (r = 0.94, P less than 0.01).(ABSTRACT TRUNCATED AT 250 WORDS)     

The energetics of middle-distance running

In order to assess the relative contribution of aerobic processes to running velocity (v), 27 male athletes were selected on the basis of their middle-distance performances over 800, 1500, 3000 or 5000 m, during the 1987 track season. To be selected for study, the average running velocity (v) corresponding to their performances had to be superior to 90% of the best French v of the season. Maximum O2 consumption (VO2max) and energy cost of running (C) had been measured within the 2 months preceding the track season, which, together with oxygen consumption at rest (VO2rest) allowed us to calculate the maximal v that could be sustained under aerobic conditions: vamax = (VO2max - VO2rest) x C-1. The treadmill running v corresponding to a blood lactate of 4 mmol.l-1 (vla4), was also calculated. In the whole group, C was significantly related to height (r = -0.43; P less than 0.03). Neither C nor VO2max (with, in this case, the exception of the 3000 m athletes) were correlated to v. On the other hand, vamax was significantly correlated to v over distances longer than 800 m. These v were also correlated to vla4. However vla4 occurred at 87.5% SD 3.3% of vamax, this relationship was interpreted as being an expression of the correlation between vamax and v. Calculation of vamax provided a useful means of analysing the performances. At the level of achievement studied, v sustained over 3000 m corresponded to vamax.(ABSTRACT TRUNCATED AT 250 WORDS)     

Hyperpnoea and CO2 sensitivity of the respiratory centres during exercise

The aim of this study was to specify whether exercise hyperpnoea was related to the CO2 sensitivity of the respiratory centres measured during steady-state exercise of mild intensity. Thus, ventilation (VE), breathing pattern [tidal volume (VT), respiratory frequency (f), inspiratory time (TI), total time of the respiratory cycle (TTOT), VT/TI, TI/TTOT] and CO2 sensitivity of the respiratory centres determined by the rebreathing method were measured at rest (SCO2re) and during steady-state exercise (SCO2ex) of mild intensity [CO2 output (VCO2) = 20 ml.kg-1.min-1] in 11 sedentary male subjects (aged 20-34 years). The results showed that SCO2re and SCO2ex were not significantly different. During exercise, there was no correlation between VE and SCO2ex and, for the same VCO2, all subjects had very close VE values normalized for body mass (bm), regardless of their SCO2ex (VEbm0.75 = 1.44 l.min-1.kg-1 SD 0.10). A highly significant positive correlation between SCO2ex and VT (normalised for bm) (r = 0.80, P less than 0.01), TI (r = 0.77, P less than 0.01) and TTOT (r = 0.77, P less than 0.01) existed, as well as a highly significant negative correlation between SCO2ex and (normalised for bm-0.25) (r = -0.73, P less than 0.01). We conclude that the hyperpnoea during steady-state exercise of mild intensity is not related to the SCO2ex. The relationship between breathing pattern and SCO2ex suggests that the breathing pattern could influence the determination of the SCO2ex. This finding needs further investigation.     

Physiological correlates of middle-distance running performance. A comparative study between men and women.

To compare the relative contributions of their functional capacities to performance in relation to sex, two groups of middle-distance runners (24 men and 14 women) were selected on the basis of performances over 1500-m and 3000-m running races. To be selected for the study, the average running velocity (v) in relation to performances had to be superior to a percentage (90% for men and 88% for women) of the best French v achieved during the season by an athlete of the same sex. Maximal O2 consumption (VO2max) and energy cost of running (CR) were measured in the 2 months preceding the track season. This allowed us to calculate the maximal v that could be sustained under aerobic conditions, va,max. A v:va,max ratio derived from 1500-m to 3000-m races was used to calculate the maximal duration of a competitive race for which v = va,max (tva,max). In both groups va,max was correlated to v. The relationships calculated for each distance were similar in both sexes. The CR [0.179 (SD 0.010) ml.kg-1 x m-1 in the women versus 0.177 (SD 0.010) in the men] and tva,max [7.0 (SD 2.0) min versus 8.4 (SD 2.1)] also showed no difference. The relationships between VO2max and body mass (mb) calculated in the men and the women were different. At the same mb the women had a 10% lower CR than the men; their lower mb thus resulted in an identical CR.(ABSTRACT TRUNCATED AT 250 WORDS)     

Cardiovascular changes associated with decreased aerobic capacity and aging in long-distance runners

Fifty-five male runners aged between 30 to 80 years were examined to determine the relative roles of various cardiovascular parameters which may account for the decrease in maximal oxygen uptake (VO2max) with aging. All subjects had similar body fat composition and trained for a similar mileage each week. The parameters tested were VO2max, maximal heart rate (HRmax), cardiac output (Q), and arteriovenous difference in oxygen concentration (Ca-Cv)O2 during graded, maximal treadmill running. Average body fat and training mileage were roughly 12% and 50 km.week-1, respectively. The average 10-km run-time slowed significantly by 6.0%.decade-1 [( 10-km run-time (min) = 0.323 x age (years) + 24.4] (n = 49, r = 0.692, p less than 0.001]. A strong correlation was found between age and VO2max [( VO2max (ml.kg-1.min-1) = -0.439 x age + 76.5] (n = 55, r = -0.768, p less than 0.001]. Thus, VO2max decreased by 6.9%.decade-1 along with reductions of HRmax (3.2%.decade-1, p less than 0.001) and Q (5.8%.decade-1, p less than 0.001), while no significant change with age was observed in estimated (Ca-Cv)O2. It was concluded that the decline of VO2max with aging in runners was mainly explained by the central factors (represented by the decline of HR and Q in this study), rather than by the peripheral factor (represented by (Ca-Cv)O2).     

Vascular volumes and hematology in male and female runners and cyclists

To examine the hypothesis that foot-strike hemolysis alters vascular volumes and selected hematological properties is trained athletes, we have measured total blood volume (TBV), red cell volume (RCV) and plasma volume (PV) in cyclists (n = 21) and runners (n = 17) and compared them to those of untrained controls (n = 20). TBV (ml x kg(-1)) was calculated as the sum of RCV (ml x kg(-1)) and PV (ml x kg(-1)) obtained using 51Cr and 125I-labelled albumin, respectively. Hematological assessment was carried out using a Coulter counter. Peak aerobic power (VO2peak) was measured during progressive exercise to fatigue using both cycle and treadmill ergometry. RCV was 15% higher (P < 0.05) in male cyclists [35.4 (1.0), mean (SE); n = 12] and runners [35.3 (0.98); n = 9] compared to the controls [30.7 (0.92); n = 12]. Similar differences existed between the female cyclists [28.2 (2.1); n = 9] and runners [28.4 (1.0); n = 8] compared to the untrained controls [24.9 (1.4); n = 8]. For the male athletes, PV was between 19% (cyclists) and 28% (runners) higher (P < 0.05) in the trained athletes compared to the untrained controls. The differences in PV between the female groups were not significant. Although the males had a higher (P < 0.05) TBV, RCV and PV than the females, no differences between cyclists and runners were found for either gender. Mean cell volume was not different between the athletic groups. VO2peak (ml x kg(-1) x min(-1)) was higher (P < 0.05) in both male [68.4 (1.5)] and female [54.8 (2.1)] runners when compared to the untrained males [47.1 (1.0)] and females [40.5 (2.1)]. Although differences existed between the genders in VO2peak for both cyclists and runners, no differences were found between the athletic groups within a gender. Since the vascular volumes were not different between cyclists and runners for either the males or females, foot-strike hemolysis would not appear to have an effect on that parameter. The significant correlations (P < 0.05) found between VO2peak and RCV (r = 0.64 and 0.64) and TBV (r = 0.82 and 0.63) for the males and females, respectively, suggests a role for the vascular system in realizing a high aerobic power.     

Buffering capacity of deproteinized human vastus lateralis muscle

The in vitro deproteinized vastus lateralis muscle buffer capacity, carnosine, and histidine levels were examined in 20 men from 4 distinct populations (5 sprinters, 800-m runners; 5 rowers; 5 marathoners; 5 untrained). Needle biopsies were obtained at rest from the vastus lateralis muscle. The buffer capacity was determined in deproteinized homogenates by repeatedly titrating supernatant extracts over the pH range of 7.0-6.0 with 0.01 N HCl. Carnosine and histidine levels were determined on an amino acid AutoAnalyzer. Fast-twitch fiber percentage was determined by staining intensity of myosin adenosinetriphosphatase. High-intensity running performance was assessed on an inclined treadmill run to fatigue (20% incline; 3.5 m X s-1). Significantly (P less than 0.01) elevated buffer capacities, carnosine levels, and high-intensity running performances were demonstrated by the sprinters and rowers, but no significant differences existed between these variables for the marathoners vs. untrained subjects. Low but significant (P less than 0.05) interrelationships were demonstrated between buffer capacity, carnosine levels, and fast-twitch fiber composition. These findings indicate that the sprinters and rowers possess elevated buffering capabilities and carnosine levels compared with marathon runners and untrained subjects.     

Atrial natriuretic peptide response to endurance physical training in the rat

The effect of an endurance physical training programme on the plasma and atrial natriuretic peptides (ANP) and on renal glomerular ANP receptors was evaluated in male normotensive Wistar rats. Maximal O2 uptake was significantly greater in the endurance trained (117.1 ml O2.kg-1.min-1, SEM 6.18 versus the control rats 84.2 ml O2.kg-1.min-1, SEM 4.88, P less than 0.01. In addition, various muscle oxidative enzymes were also significantly higher in endurance trained animals. An increase in resting plasma [ANP] was observed after 11 weeks of physical training (40.02 pg.ml-1, SEM 7.07 vs 22.8 pg.ml-1, SEM 3.83, P less than 0.05). Glomerular ANP receptor density was lower in trained rats (272 fmol.mg-1 protein, SEM 3.1 vs 380 fmol.mg-1 protein, SEM 6.1, P less than 0.05), whereas atrial tissue [ANP] was not significantly different between controls and trained animals. However, in trained rats, circulating [ANP] was closely correlated with left atrial [ANP] (r = -0.92, P less than 0.05). Resting systolic blood pressure had not changed at the end of this physical training programme. It is considered that under physiological conditions ANP may be involved in long-term extracellular fluid volume homeostasis through the regulation of renal glomerular ANP receptors, and that the left atrium might play a significant role in this long term fluid volume control.     

Muscle metabolism in track athletes, using 31P magnetic resonance spectroscopy.

We tested whether preferred running event in track athletes would correlate with the initial rate of phosphocreatine (PCr) resynthesis following submaximal exercise. PCr recovery was measured in the calf muscles of 16 male track athletes and 7 male control subjects following 5 min of repeated plantar flexion against resistance. Pi, PCr, and pH were measured using phosphorus magnetic resonance spectroscopy (31P MRS) with an 8-cm surface coil in a 1.8-T magnet. During exercise, work levels were gradually increased to deplete PCr to 50-60% of the initial value. No drop in pH was seen in any of the subjects during this exercise. The areas of the PCr peaks following exercise were fit to monoexponential curves. Two or three tests were performed on each subject and the results averaged. Athletes were divided into three groups based on their primary event: sprinters running 400 m or less, middle-distance athletes running 400-1500 m, and long-distance athletes running farther than 1500 m. The maximal rates of PCr resynthesis (mmol.min-1.kg-1 muscle weight) were 64.8 +/- 8.6, for long-distance runners; 41.4 +/- 11, for middle-distance runners; 32.0 +/- 7.0, for sprinters; and 38.6 +/- 10, for controls (mean +/- SE). The faster PCr recovery rates seen in long-distance runners compared with sprinters indicate greater oxidative capacity, which is consistent with the known differences between athletes in these events.     

Carbohydrate supercompensation and muscle glycogen utilization during exhaustive running in highly trained athletes

Three female and three male highly trained endurance runners with mean maximal oxygen uptake (VO2max) values of 60.5 and 71.5 ml.kg-1.min-1, respectively, ran to exhaustion at 75%-80% of VO2max on two occasions after an overnight fast. One experiment was performed after a normal diet and training regimen (Norm), the other after a diet and training programme intended to increase muscle glycogen levels (Carb). Muscle glycogen concentration in the gastrocnemius muscle increased by 25% (P less than 0.05) from 581 mmol.kg-1 dry weight, SEM 50 to 722 mmol.kg-1 dry weight, SEM 34 after Carb. Running time to exhaustion, however, was not significantly different in Carb and Norm, 77 min, SEM 13 vs 70 min, SEM 8, respectively. The average glycogen concentration following exhaustive running was 553 mmol.kg-1 dry weight, SEM 70 in Carb and 434 mmol.kg-1 dry weight, SEM 57 in Norm, indicating that in both tests muscle glycogen stores were decreased by about 25%. Periodic acid-Schiff staining for semi-quantitative glycogen determination in individual fibres confirmed that none of the fibres appeared to be glycogen-empty after exhaustive running. The steady-state respiratory exchange ratio was higher in Carb than in Norm (0.92, SEM 0.01 vs 0.89, SEM 0.01; P less than 0.05). Since muscle glycogen utilization was identical in the two tests, the indication of higher utilization of total carbohydrate appears to be related to a higher utilization of liver glycogen. We have concluded that glycogen depletion of the gastrocnemius muscle is unlikely to be the cause of fatigue during exhaustive running at 75%-80% of VO2max in highly trained endurance runners. Furthermore, diet- and training-induced carbohydrate super-compensation does not appear to improve endurance capacity in such individuals.