L. L. SPRIET, D. A. MACLEAN, D. J. DYCK, E. HULTMAN, ... D. J. Dyck, E. Hultman,. G. Cederblad, and T. E. Graham. ... at 69% \io, max, and Erickson et al.
Caffeine ingestion and muscle metabolism during prolonged exercise in humans L. L. SPRIET, D. A. MACLEAN, D. J. DYCK, G. CEDERBLAD, AND T. E. GRAHAM
E. HULTMAN,
School of Human Biology, University of Guelph, Guelph, Ontario Nl G 2 WI, Canada; and Departments of Clinical Chemistry I and II, Huddinge University Hospital, Karolinska Institute, S-141 86 Huddinge, Sweden Spriet, L. L., D. A. MacLean, D. J. Dyck, E. Hultman, G. Cederblad, and T. E. Graham. Caffeine ingestion and musclemetabolismduring prolonged exercisein humansAm. J. Physiol. 262 (Endocrinol. Metab. 25): E891-E898, 1992.-We examinedthe effects of a high-caffeine doseon enduranceperformanceand muscleacetyl group metabolismduring prolonged exercise.Eight subjectscycled to exhaustion at -80% maximal oxygen uptake (VO 2max)1 h after ingestion of 9 mg/kg body wt dextrose (Pl) or caffeine (Caf). In the Pl trial, musclebiopsies weretaken at rest (1 h postingestion)and at 15min and exhaustion during exercise.The Caf trial followed the sameprotocol 1 wk later, with an additional biopsy at the time correspondingto Pl exhaustion. The subjectscycled significantly longer during the Caf trial (96.2 & 8.8 min) than in the Pl trial (75.8 t 4.8 min). Net glycogenolysisduring the initial 15min of cycling was reducedin the Caf vs. Pl trial (4.7 t 1.5 vs. 10.6t 1.3 mmol kg dry muscle-l min- l; P < 0.05). Muscle citrate concentration was increased at rest with Caf (0.59 t 0.07 vs. 0.37 t 0.05 mmol/kg dry muscle;P < 0.05) but increasedto similar values in both trials during cycling. Caf elevated the acetyl-CoA/ CoA-SH ratio at rest (0.316* 0.046vs. 0.201* 0.023;P < 0.05) but had no effect on the increasesin muscle acetyl-CoA and acetylcarnitine during exercise. The results indicate that Caf before exercise decreasedmuscleglycogenolysisby -55% over the first 15 min of exercise at -80% ir0, max. This “spared glycogen” was available late in exercise and coincided with a prolongedtime to exhaustion. Increasedutilization of intramuscular triacylglycerol and/or extramuscular free fatty acids after caffeine ingestion may inhibit carbohydrate useat rest and early during exercisevia elevations in musclecitrate and the acetylCoA/CoA-SH ratio. Muscle acetyl-CoA and acetylcarnitine were maintained above resting contents even at exhaustion when muscleglycogen was depleted. muscleglycogenolysis;epinephrine; acetyl-coenzyme A; acetylcarnitine; free carnitine; coenzyme A-SH; citrate; acetyl-coenzyme A/coenzyme A-SH; free fatty acids; intramuscular triacylglycerol l
l
AND CO-WORKERS (14, 27) were the first to report that caffeine ingestion produced an ergogenic effect during prolonged endurance exercise. They proposed that caffeine elevated plasma catecholamine concentrations which stimulated fat metabolism, either by increasing adipose tissue and/or muscle triacylglycerol lipolysis and consequently free fatty acid (FFA) oxidation. The enhanced fat oxidation may elevate muscle citrate content and subsequently reduce muscle glycogenolysis during exercise and may delay exhaustion in prolonged endurance exercise (18). However, much of the subsequent research has not demonstrated an ergogenie effect of caffeine nor has this research been able to provide supporting evidence for a metabolic explanation of caffeine’s ergogenic effect. A recent study from our laboratory demonstrated that ingestion of a high dose of caffeine (9 mg/kg body wt) COSTILL
0193-1849/92
$2.00
Copyright
produced a significant ergogenic effect in the same welltrained runners during both cycling and running at 85% of maximal oxygen uptake (Vo2 max) (22). Plasma epinephrine (Epi) concentration was elevated by caffeine at rest and during exercise, but plasma norepinephrine and FFA concentrations and respiratory exchange ratio (RER) were unaffected. This study confirmed the ergogenic and Epi effects of caffeine ingestion but failed to identify the metabolic cause. Although many studies have indirectly investigated the effects of caffeine on muscle metabolism, few have examined the effects of caffeine ingestion on muscle glycogen or fat utilization directly. Two studies measured pre- and postglycogen content during bouts of nonexhaustive exercise of moderate intensity and reported a glycogen-sparing effect after the ingestion of 5 mg/kg body wt caffeine. Essig et al. (18) reported a 42% reduction in glycogen use during 30 min of cycling at 69% \io, max, and Erickson et al. (17) found a 30% reduction in glycogen use during 90 min of cycling at 6570% \jozmax. Therefore, the first purpose of this study was to extend these findings and to examine the effects of a high dose of caffeine on the time course of glycogen utilization and endurance performance during intense cycling to exhaustion. The second purpose of the paper was to examine the potential mechanisms by which increased fat oxidation decreases carbohydrate utilization in human skeletal muscle. It has been suggested in heart muscle and rat skeletal muscle that increases in citrate content and acetyl-CoA/CoA-SH ratio decrease the respective activities of phosphofructokinase (PFK) and pyruvate dehydrogenase (PDH), key enzymes in carbohydrate metabolism (16, 34, 35, 42). Because little direct information exists in exercising human muscle, we measured muscle citrate, acetyl-CoA, acetylcarnitine, and the acetylCoA/CoA-SH ratio at rest and during exercise after placebo and caffeine ingestion. Although the present measurements will not prove the existence of these feedback mechanisms, they will determine whether measured changes are consistent with their proposed roles. METHODS Subjects. Seven males and one female volunteered for the experiment. All were recreational cyclists with a mean age of 27.8 k 2.5 (SE) yr and body weight of 74.5 t 3.4 kg. The experimental proceduresand potential risks of the study were explained to each subject both verbally and in writing. All subjects gave informed consent, and the experiment was approved by the University’s Ethics Committee. Preexperimental protocol. Each subject reported to the laboratory before the start of the experiment and performed an
0 1992 the American
Physiological
Society
E891
E892
CAFFEINE
AND
MUSCLE
METABOLISM
incremental ire, max test on a cycle ergometer to select the appropriate power output. After a 30-min rest, subjectscycled for lo-15 min at a power output designed to elicit -80% VO 2m8X.The mean\i0, max for the group was 4.06 t 0.23 l/min (54.7 & 2.6 ml. kg-l. min-l). Subsequently,subjectsreported to the laboratory at the sametime of the day on two occasions, separatedby 1 wk. During each trial they cycled to exhaustion at -80% VO, max after consuming either placebo or caffeine. Identical weights were placed on the pan balance of the bicycle in each trial and revolutions per minute were closely monitored to ensureidentical power outputs in both trials. The trials were administeredin single-blind fashion, with the first trial always placebo (Pl) and the secondtrial always caffeine (Caf; Fig. 1). The trials were ordered to obtain a biopsy during the caffeine trial that correspondedto exhaustion during the Pl trial, if the subject exercisedlonger in the Caf trial (Fig. 1). The subjectswere instructed to maintain their normal training programsduring the study and wereaskedto incorporate the weekly test into their training program asa hard workout. They were alsoaskedto eat in preparation for all trials asthey would before a competition, including a meal 2-4 h before each trial. All subjectsmaintained pretrial food diaries, which were analyzed for total energy and carbohydrate content (15). This analysis revealedthat all subjectsconsumeddiets high in carbohydrate content (59.6%) before each trial, ensuring that muscle, liver, and blood carbohydrate storeswere optimal. The subjects were also askedto refrain from caffeine consumption for 72 h before each trial. Five subjects were not caffeine users (~50 mg/day), two consumed180-250 mg/day, and one was a heavy caffeine user (-600 mg/day). Experimental protocol. Subjects reported to the laboratory and immediately gave a urine sample. A catheter was placed percutaneouslyinto a medial antecubital vein, and a salinedrip (loo-175 ml/h) was started to maintain catheter patency. A resting blood sample(referred to as-60 min) wasobtained, and subjectsthen consumedeither 9 mg/kg body wt placebo (dextrose) or caffeine in capsuleform along with water. After ingestion of the capsules (1 h), a second resting blood sample (referred to as0 min) wasobtained, and a resting musclebiopsy wastaken from the vastus lateralis (4). Subjects then cycled to volitional exhaustion at -80% VO 2max(inability to maintain pedal cadence), and blood and expired gas samples were obtained every 15 min and at exhaustion. All blood sampleswere taken while the subject was exercising. Muscle biopsies were also taken at 15 min and exhaustion in the Pl trial and at 15 min, the time corresponding to Pl exhaustion, and exhaustion in the Caf trial (Fig. 1). The subjectswere unaware of the cycling duration during the trials and were given constant encouragementto cycle to exhaustion. A secondurine samplewas obtained within 15 min of completion of exercise. The results of the experiment were not dis-
i
pL
B
B
a
a
b rest
time
CAF
-60 rest b=>b---
exercise 0
15
EXH
EXH
exercise Bym+-+l
0 0 B B Fig. 1. Schematic representation of experimental Caf, caffeine; Exh, exhaustion; B, biopsy.
ti t B B design. Pl, placebo;
DURING
PROLONGED
EXERCISE
closeduntil completion of both trials. Analyses. Expired gassampleswere analyzed for fractions of 0, and CO, with an Applied Electrochemical S-3A 0, analyzer and SensorMedics LB-2 CO, detector, respectively. The analyzers were calibrated with gasesof known concentrations, previously determined by micro-Scholander technique. Expired volume was determined with a Parkinson-Cowan volumeter, after calibration with a Tissot spirometer. The musclebiopsy sampleswere immediately frozen in liquid N2, removed from the needle,and stored at -8OOC.The entire biopsy was then freeze-dried and powdered to dissect out all nonmuscularelements.Two small aliquots of powderedmuscle (each 2-3 mg) were used for the enzymatic determination of glycogen (24). An additional aliquot (3 mg) was extracted with 0.5 M HClO, (1.0 mM EDTA), neutralized with 2.2 M KHC03, and analyzed enzymatically for musclecitrate (3). The remaining muscle(- lo-15 mg) wasextracted as describedabove and analyzed for total carnitine, acetylcarnitine, free carnitine, total CoA, and acetyl-CoA using the radiolabeledtechniquesof Cederblad et al. (11). The total CoA determination wasthe sum of CoA-SH and acetyl-CoA (additional CoA esters were not determined). CoA-SH wascalculatedby subtracting acetyl-CoA from total CoA. The inter- and intra-assay variabilities were 13.6 and 3.8% for acetyl-CoA and 9.8 and 8.7% for citrate, respectively. Muscle wet-to-dry weight ratios were similar for Pl and Caf resting biopsies(4.07 t 0.07; 4.02 t 0.04), and the increasein muscle water content with exercise was also similar between trials (4-g%). A portion of the above extract was also usedfor the determination of total creatine on each biopsy (19). All musclemetabolite contents were corrected to the highest total creatine content for each individual’s biopsies. The average total creatine content was 135.3t 2.7 mmol/kg dry muscle.All muscledata were expressedper kilogram dry muscle. Blood sampleswere immediately divided into two aliquots; 3 ml were transferred to a nontreated tube for serum, and 7 ml were transferred into a sodium-heparinized tube. Hematocrit wasimmediately measuredin triplicate with high-speedcentrifugation usingblood from the heparinized tube. An aliquot of 100 ~1heparinized blood wasremoved and addedto 500 ~1of 0.3 M HC104. A solution containing 120 ~1of 0.24 M ethylene glycolbis(P-aminoethyl ether)-N,N,N’,N’-tetraacetic acid (EGTA) and reducedglutathione wasthen addedto the remaining heparinized whole blood, and the samplewas centrifuged to separate the plasmafraction. The EGTA- and glutathione-treated plasmawas analyzed in duplicate for Epi and norepinephrine concentration usinghighperformance liquid chromatography (Waters) as describedby Weiker et al. (41). The whole blood acid extracts were analyzed enzymatically in triplicate for lactate and glucose(3), Serum glycerol was analyzed enzymatically (21), and serumFFA was measured with an enzymatic calorimetric technique (Wako NEFA C kit; Wako Chemicals,Dallas, TX). Statistics. Cycle times to exhaustion after caffeine and placebo ingestion were comparedwith a paired t test. Muscle resting and 15-min and Pl exhaustion data were analyzed by twoway analysesof variance (time and treatment) for repeated measures.Duncan’s multiple-range tests were usedto compare meanswhen a significant F ratio was obtained. Pl exhaustion and Caf exhaustion muscledata were compared with paired t tests. The plasmaand expired gasdata were complicatedby the fact that each subject exercisedfor a different duration. However, completedata setswere obtained at -60 and 0 min at rest and 15, 30, 45, and 60 min and at Pl and Caf exhaustion during exercise.Resting data were analyzed with two-way analysesof variance as above (time and treatment). Becausetime differenceswere not of interest in the exerciseplasmaand expired gas
CAFFEINE
AND MUSCLE
METABOLISM
variables, paired t tests were used to determine if significant differencesexisted between Pl and Caf at 15,30,45, and 60 min and at Pl and Caf exhaustion. Significance wasacceptedat P 5 0.05, and all data are meanst SE. RESULTS
Cycling time to exhaustion at -80% 00, max increased significantly after caffeine ingestion. The individual performance data demonstrate that seven of eight subjects cycled longer with caffeine while one subject did not change (Fig. 2). This subject (subject 2) stopped during the Caf trial due to stomach discomfort, secondary to the caffeine ingestion. The mean cycling times during the Pl and Caf trials were 75.8 t 4.8 and 96.2 t 8.8 min, respectively, with subject 2 and 78.9 t 4.7 and 102.6 t 7.9 min without subject 2. The mean VO, during the Pl and Caf trials was 78.2 t 1.9 and 84.6 t 2.7% \io, max. The Vo2 during the Caf trial was significantly higher than Pl at 15, 30, and 45 min (Table 1). RER tended to be higher during the Caf trial and was significantly higher at 15 and 30 min (Table 1). Ventilation was also significantly higher at all time points except 60 min during the Caf trial (Table 1). Resting plasma Epi concentration ([Epi]) was unaffected by Pl ingestion but increased significantly after caffeine ingestion (Fig. 3). The exercise-induced increase in [Epi] was also significantly greater in the Caf trial at 15 and 30 min and at exhaustion (Fig. 3). Resting norepinephrine concentration was unaffected by caffeine ingestion but did increase significantly during the 60 min of rest in the Pl trial (Fig. 3). Caffeine had no effect on the large increase in norepinephrine concentration during exercise. Caffeine ingestion increased the plasma FFA concentration ([FFA]) from 0.24 t 0.03 to 0.46 t 0.14 mM during 60 min at rest, but the difference did not reach statistical significance (Fig. 4). Pl ingestion had no effect on plasma FFA during rest (-60 min, 0.29 t 0.05 mM; 0 min, 0.26 t 0.05 mM). Plasma [FFA] decreased after 15 min of exercise with Caf (0.22 t 0.02 mM) and was not different from Pl. Beyond 15 min in both trials, plasma 140
cl Ea
$20 .-5 WlOO z i= 0W 80 5 z 0
60
kt
40
E z
20
2
3
PLACEBO CAFFEINE
4
5
6
7
8
SUBJECTS Fig. 2. Individual performance times of subjects cycling to exhaustion at -80% maximal O2 uptake (VO 2 max) after placebo or caffeine ingestion. Subject 5 was a caffeine user (600mg/day), subjects 4 and 6 were mild users (180-250 mg/day), and the others were not caffeine users (~50 mg/day)~
DURING
PROLONGED
E893
EXERCISE
[FFA] levels gradually increased and were not significantly different until exhaustion, where the Caf value was higher (Fig. 4). Plasma glycerol concentration ( [glycerol]) was significantly increased by caffeine ingestion at rest, whereas Pl ingestion had no effect (Fig. 4). Caffeine had no effect on the gradual increase in plasma [glycerol] during exercise, except at exhaustion, where Caf glycerol was higher. Caffeine ingestion had no effect on resting blood concentrations of lactate and glucose (Table 2). However, significantly higher glucose and lactate concentrations occurred at 15, 30, and 45 min of exercise after Caf ingestion. Plasma citrate concentrations were unaffected by caffeine both at rest and during exercise (Table 2). Muscle glycogen content decreased from 463 t 28 mmol glucosyl units/kg dry muscle at rest to 304 t 27 and 80 -+ 13 mmol/kg dry muscle at 15 min and exhaustion (76 min), respectively, during the Pl trial (Fig. 5). In the Caf trial, glycogen decreased from 485 t 36 mmol/kg dry muscle at rest to 414 t 41 at 15 min, 137 t 31 at 76 min, and 39 t 13 mmol/kg dry muscle at exhaustion. Resting and exhaustion glycogen contents were not different in the two trials. However, glycogen content was significantly higher at 15 min and at the Pl exhaustion point (76 min) in the Caf trial (Fig. 5). The rate of glycogen utilization during the initial 15 min of cycling was also significantly reduced in the Caf trial (Pl vs. Caf: 10.6 t 1.3 vs. 4.7 t 1.5 mmol kg dry muscle-l min-l) and represented a sparing of 55%. The rates of glycogen utilization from 15 min to exhaustion were similar in the Pl and Caf trials (3.8 t 0.4,4.1 t 0.2 mmolkg dry muscle-l emi&). Muscle citrate levels were significantly elevated 60 min after Caf ingestion (0.59 t 0.07 mmol/kg dry muscle) as compared with Pl (0.37 t 0.05 mmol/kg dry muscle). Cycling increased citrate concentration to 0.89 t 0.08 and 0.73 t 0.07 mmol/k g d ry muscle at 15 min and exhaustion (76 min) in the Pl trial (Fig. 6). Citrate values at 15 and 76 min (0.87 t 0.09 and 0.66 t 0.08 mmol/kg dry muscle) and at exhaustion (0.65 t 0.06 mmol/kg dry muscle) were similar in the Caf trial. Muscle acetylcarnitine content was unaffected by caffeine ingestion at all time points (Table 3). In both trials, acetylcarnitine increased significantly above rest after 15 min of cycling and remained elevated throughout the exercise period. The changes in free carnitine were reciprocal to the acetylcarnitine changes such that measured total carnitine was unchanged at all time points. The addition of acetylcarnitine and free carnitine closely matched the measured values for total carnitine. Caffeine ingestion had no effect on resting or exercise acetyl-CoA contents (Table 3). Acetyl-CoA content was significantly increased over rest at all time points during cycling in both trials. However, at Pl exhaustion in both trials and at exhaustion in the Caf trial, acetyl-CoA content was significantly lower than the 15-min values. Total muscle CoA content was unaffected by caffeine or exercise except for a decreased content at exhaustion in the Caf trial. Consequently, the changes in calculated CoA-SH content were reciprocal to the changes in acetyl-CoA. Caffeine ingestion had no effect on the l
l
E894
CAFFEINE
AND MUSCLE
METABOLISM
DURING
PROLONGED
EXERCISE
Table 1. Oxygen uptake, respiratory exchange ratio, and ventilatory data during cycling to exhaustion after placebo or caffeine ingestion Time,
min Exhaustion
15 Orovo2
30
45
60
max
Pl 74.3k5.3 Caf 82.0&4.3* RER Pl 0.81t0.01 Caf 0.85t0.01* VE, l/min Pl 63.9t6.2 Caf 77.9t3.3* Data are means k SE; n = 8 subjects. Pl, placebo; ventilation, STPD. * Significantly different from Pl.
r -12
74.2k2.6 83.8t2.8*
78.5k1.6 84.3t2.5*
83.2t2.8 82.4t2.9
84.1t2.1
0.81t0.01 0.83t0.01*
0.81t0.01 0.83t0.01
0.82kO.01 0.84kO.01
0.82t0.01 0.84kO.01
63.6k4.4
70.3k3.7 80.4k3.0”
80.6k4.5 84.6k3.5
84.3t3.5 99.3t4.6”
78.1k2.9” Caf, caffeine;
\iO,
86.1t2.9
max, maximal 0, uptake; RER, respiratory exchange ratio;
expired
7jE,
1
14
-
A
-
-
PL ,ACEBO
CAFFEINE
-
PLACEBO
CAFFEINE
0.8
I
ZlO z E a2 9 E w
4-
-60
-45
-30
-15
0
15
30
45
60
75
90
-60
105
-45
-30
-15
0
15
TIME
TIME (mid
30
45
60
75
90
105
(mid
0.7
25 -
PLACEBO
-
-CAFFEINE
-
PLACEBO
CAFFEINE
0.6 a g
Oa5
0”
0.4
e 0 3 a
P
0.3 0.2 0.1
I
I
I
-60
-45
-30
I
-15
I
I
0
15
I
I
I
30
45
60
I
75
I
90
I
0
I
-60
105
I
-45
I
-30
I
-15
I
I
0
15
I
30
I
45
I
I
60
75
I
90
I
105
TIME (min) Fig. 3. Plasma epinephrine and norepinephrine responses during cycling to exhaustion after placebo or caffeine ingestion. *Significantly different from placebo. +Significantly different from -60 min.
TIME (mid Fig. 4. Plasma free fatty acid (FFA) and glycerol concentrations during cycnng to exhaustion after placebo or caffeine ingestion. *Significantly different from placebo. +Significantly different from -60 min.
increase in acetyl-CoA/CoA-SH ratio during exercise but did elevate the ratio significantly at rest.
caffeine ingestion produced a glycogen sparing of 42% during 30 min of cycling (69% VO, max) in active males. Erickson et al. (17) reported that caffeine ingestion reduced glycogen utilization from 91.4 t 10.0 to 63.1 t 7.9 wet muscle during 90 min of 6570% . mmol/kg vo 2 max cycling in trained subjects. The present study demonstrated that the glycogen-sparing effect of caffeine also occurred at a more intense power output (-80% . vo 2 max) and was confined to the initial 15 min of cycling. The delayed utilization of glycogen early in exercise enabled the subjects to cycle longer before experiencing glycogen depletion and exhaustion. The findings are consistent with the often stated conclusion that depletion of muscle glycogen stores limits prolonged exercise at power outputs between 65 and 85% VO, max(26, 39). Bergstrom and Hultman (5) and Hermansen et al. (25)
DISCUSSION
A significant finding of this study was the large glycogen-sparing effect of caffeine ingestion within the initial 15 min of cycling at -80% vo2 IIlaX. Beyond 15 min of cycling the rates of glycogen utilization were similar in both trials. Therefore, in the Caf trial, muscle glycogen was not depleted at the time point corresponding to exhaustion in the Pl trial, and the remaining glycogen coincided with a prolonged cycle time to exhaustion. The present results extend the findings of two previous studies that examined glycogen utilization during nonexhaustive and moderate-intensity cycling after the ingestion of 5 mg/kg caffeine. Essig et al. (18) reported that
1.
CAFFEINE
AND MUSCLE
METABOLISM
DURING
PROLONGED
E895
EXERCISE
Table 2. Blood lactate, glucose, and plasma citrate concentrations during cycling to exhaustion after placebo or caffeine ingestion Time,
min Exhaustion
-60
0
Lactate, mM Pl 0.90t0.09 0.95kO.03 Caf 0.94&O. 13 1.14t0.21 Glucose, mM Pl 4.0520.28 3.29kO.27 Caf 3.54t0.32 3.04t0.34 Citrate, PM Pl 91.1k7.2 84.4t11.9 Caf 90.3t7.2 96.4t13.6 Data are means k SE; n = 8 subjects. -60, 60 min
%600
30
15
3.20zko.29 4.74&0.75*
2.87t0.42 4.73kO.87"
3.05t0.63 3.49t0.74
3.34kO.45 3.8420.57
3.08t0.22 3.42*0.18*
3.14t0.21 3.81t0.25*
3.21t0.20 4.05&0.34*
3.15t0.11 3.36k0.18
2.77k0.22 3.09kO.37
80.3k4.8 81 .Ot9.3
89.3k6.7 91.9t9.6
before exercise. * Significantly
1;
-15
c3
0
15
30
PLACEBO
45
60
-
75
CAFFEINE
90
105
TIME (mid
Fig. 5. Muscle glycogen contents during cycling to exhaustion after placebo or caffeine ingestion. *Significantly different from placebo. dm, dry muscle. -
PLACEBO
-
CAFFEINE
5 2 0.8
\ 5
E 0.6
-5
E 0.4 2 6 0.2 Of-1 -15
I 0
15
45
30
I -----I-L-.I 60
75
90
105
(mid Fig. 6. Muscle citrate content during cycling to exhaustion after placebo or caffeine ingestion. *Significantly different from placebo. TIME
were the first to study the rates of muscle glycogenolysis in subjects cycling to exhaustion at -80% VO, max. Serial muscle biopsies taken at 15 to 20-min intervals throughout the exercise (69-80 min) revealed that the glycogenolytic rate was exponential, with the greatest amount of glycogen used in the initial 15- to 20-min interval. The glycogen data in the Pl trial of this study were consistent with these findings, as the respective glycogenolytic rates were 10.6 and 3.8 mmol kg dry muscle-l min-l in the 0- to 15 and 15- to 76-min intervals, respectively. Interestingly, the entire glycogen-sparing effect of caffeine was limited to the initial 15 min of exercise, when glycogenolysis is the greatest. This is the most efficient time to spare glycogen, since the 50% reduction in glycogenolysis in the Caf trial amounted to 88 mmol/kg dry muscle in l
60
3.06&O. 19 4.25t0.44*
r -
45
112.7t7.2 109.3tlO.O
108.4t9.9 123.7t19.2
108.5t7.2 126.3t13.2
different from Pl.
absolute terms or -20% of the resting muscle glycogen content. Several mechanisms have been proposed to explain the glycogen-sparing effect of caffeine. Inherent in these discussions is the assumption that fat metabolism increased after caffeine ingestion to maintain a constant supply of ATP during intense exercise, although this has been difficult to prove. It is believed that caffeine enhances lipolysis from adipose tissue, via the increase in plasma [Epi], resulting in enhanced FFA oxidation by the muscle. However, this study and many others have not been able to demonstrate a significantly elevated plasma [FFA] before exercise after caffeine ingestion (14, 17, 22, 27). Unfortunately, most studies measured forearm venous [FFA], and this is not optimal for estimating the energy contribution of circulating FFA to exercising muscle. Given the energy density and rapid turnover of FFA in plasma, turnover studies will be required to accurately determine the metabolic significance of exogenous FFA after caffeine ingestion. Enhanced fat oxidation after caffeine ingestion may also be due to increased utilization of intramuscular triacylglycerol during the initial minutes of exercise. Little is known regarding the lipase responsible for the breakdown of triacylglycerol, but it is thought to be sensitive to Epi and is mediated through the second messenger adenosine 3’,5’-cyclic monophosphate (CAMP) system (33). If caffeine crosses the muscle membrane in large amounts, it may also inhibit phosphodiesterase, the enzyme responsible for the breakdown of CAMP (7). However, it is unlikely that phosphodiesterase activity is affected in this study since the plasma caffeine concentration would be ~75 pM (ES),a value much lower than used in the in vitro work (7). Caffeine may also antagonize the inhibition of adenylate cyclase activity by adenosine at the level of the muscle membrane adenosine receptor. This would produce elevated CAMP concentration ([CAMP]) and could possibly stimulate lipolysis. Essig et al. (18) provided the only direct support for the role of enhanced triacylglycerol utilization after caffeine ingestion. Preexercise triacylglycerol concentration was elevated by 18% 1 h after the ingestion of 5 mg/kg caffeine, and triacylglycerol utilization was 150% greater during 30 min of exercise at 69% \iO, max as compared with the Pl trial. It is difficult to reconcile why Caf-induced increases in [Epi] and [CAMP] and any effects on phosphodiesterase would affect glycogen and triacylglycerol metabolism
ES96
CAFFEINE
AND MUSCLE
METABOLISM
DURING
PROLONGED
EXERCISE
Table 3. Muscle carnitine and CoA compounds during cycling to exhaustion after placebo or caffeine ingestion Placebo Rest
15 min
Caffeine Exhaustion
Rest
15 min
Pl-exhaustion
Exhaustion
14.221.5” Acetylcarnitine, mmol/kg 3.2t0.6 17.lt0.7* 4.9H.9 16.9tl.O* 15.0&l .6* 12.4*1.4* Free carnitine, mmol/kg 20.3~~0.6 7.7t0.8” lO.ltl.4* 18.4tl.O 7.8kl.l” 8.9t1.2” 10.5t1.4” Total carnitine, mmol/kg 23.4t0.9 23.8tl.O 23.5t0.9 22.4t1.5 23.6kl.O 22.4k1.5 23.6kl.O Acetyl-CoA, pmol/kg 10.5t1.3 34.Ok4.0” 24.6&3.9*? 14.6k3.0 32.6t3.1* 23.9+3.4*? 19.6k2.6” CoA-SH, pmol/kg 54.7k5.2 33.8*3.8* 34.4&4.3* 45.3k2.6 37.2k5.3 32.Ok2.3” 27.1t2.1” 65.2t5.5 67.8t4.0 59.0t4.4 59.9k5.3 69.8t6.0 55.9t4.6 46.7t2.6” Total CoA, pmol/kg Acetyl-CoA/CoA-SH 0.2OlkO.023 1.214&0.304* 0.845&0.193* 0.316&0.046$ 1.006&0.169* 0.758&0.111* 0.770&0.121* Data are means & SE and are expressed per kg dry muscle; n = 8 subjects except for Caf at rest and Pl-exhaustion in which n = 4 subjects and Caf exhaustion in which n = 7 subjects. CoA-SH calculated from total CoA minus acetyl-CoA. * Significantly different from rest. JTSignificantly different from 15 min. $ Significantly different from Pl.
differently. One possibility involves the potential of caffeine to directly inhibit phosphorylase a, the more active form of the enzyme responsible for glycogen breakdown. The support for this possibility comes from an in vitro study that reported a 60% reduction in rabbit muscle phosphorylase a activity after the addition of 1,000 PM caffeine (a nonphysiological level) in the absence of adenine nucleotides (29). However, when physiological concentrations of AMP and ATP were added, the inhibitory effect was abolished. Even if this mechanism did exist, it is known that the initial conversion of phosphorylase to the a form at the onset of exercise is quickly converted back to the b form with sustained exercise (12, 40). Therefore, it is questionable whether initial inhibition of the phosphorylase a activity by caffeine could account for the measured reduction in glycogenolysis. A second possibility involves inhibition of glycolysis at the PFK step after caffeine ingestion as discussed below. If glucose 6-phosphate (G-6-P) content increases as a consequence of the PFK inhibition, it may inhibit phosphorylase b activity, as suggested by in vitro work (1). We attempted to determine whether enhanced fat metabolism downregulates carbohydrate metabolism in skeletal muscle, as originally proposed by Randle et al. (34) in the glucose-fatty acid cycle for heart muscle. Increases in skeletal muscle citrate and acetyl-CoA contents due to increased fat metabolism may inhibit muscle glycolysis at the reactions catalyzed by PFK and PDH, respectively (16,35,42). Inhibition at the PFK step may also lead to an accumulation of G-6-P, inhibition of hexokinase, and ultimately reduced glucose uptake. The caffeine-induced alterations in substrate utilization in the present study had no significant effect on muscle acetylCoA and CoA-SH contents at rest and during exercise, although resting acetyl-CoA was 14.6 t 2.6 pmol/kg dry muscle with caffeine and 10.5 t 1.2 pmol/kg dry muscle without caffeine. Despite these responses, the acetylCoA/CoA-SH ratio was significantly higher at rest after caffeine ingestion. The citrate content was also elevated before exercise in the Caf trial, suggesting that inhibition of PDH and PFK may have occurred at rest and possibly early in exercise. Although the glycolytic rate is low at rest, caffeine ingestion may decrease it further and/or prime the system to inhibit glycolysis at the onset of exercise when the glycogenolytic and glycolytic rates are high. By 15 min of exercise, Pl-Caf differences in the acetyl-CoA/CoA-SH ratio and citrate content were not
present. This is consistent with the lack of glycogen sparing beyond 15 min of exercise. Unfortunately, we have no muscle acetyl-CoA, CoA-SH, and citrate data for the time period between rest and 15 min of exercise. Repeated measurements during this time period will be required to test this postulation. Central to the above discussion is the assumption that fat metabolism is enhanced after caffeine ingestion. However, the RER data in this study are not consistent with this assumption. There are several possible reasons for this inconsistency. At a power output of -80% VO, m8x, ventilation, acid-base balance, and CO2 stores may not be in steady state. Furthermore, these conditions are not constant across treatments, since ventilation (Table 1) and blood lactate (a reflection of acidosis; Table 2) are higher in the Caf trial. In addition, we did not measure RER within the period that glycogen sparing occurred. Thus we do not feel that the RER data can be employed for metabolic interpretations in this study. Plasma [Epi] was twofold greater after 15 min of cycling in the Caf trial while muscle glycogenolysis was reduced by 50%. Numerous human and rat studies have demonstrated that increasing [Epi] enhances muscle glycogenolysis during exercise and electrical stimulation (2, 28, 36, 40). However, many studies used experimental [Epi] levels, which are higher than those experienced during prolonged exercise (28, 36), or infused large amounts of Epi and did not quantify [Epi] (40). For example, Jansson et al. (28) increased the Epi concentration in the femoral artery of one leg from a normal value of 3.87 to -12.7 nM and reported enhanced muscle glycogen utilization during 45 min of cycling at 50% VO, max. Equivalent [Epi] do not occur during intense aerobic exercise at -85% \jozmax (22, 30) and are only achieved during short-term supramaximal activity (6, 31). Although @-adrenergic blockade studies suggest a role for circulating Epi in the utilization of muscle glycogen (20), it has been difficult to obtain direct experimental evidence in humans to prove that Epi is an important regulator of muscle glycogenolysis during aerobic exercise at the intensity used in this study. We suggest that Epi may be more important in the regulation of fat metabolism and that its potential role in muscle glycogenolysis is overridden by other intracellular changes. It is suggested that inhibition at PFK (and possibly PDH) increases G-6-P content, which not only decreases glucose phosphorylation and subsequently
CAFFEINE
AND MUSCLE
METABOLISM
uptake but also inhibits phosphorylase b activity (1). As stated above, since little of the phosphorylase is in the a form during prolonged exercise (12,40), regulation of the less active b form may be most important in this situation. In other words, there are a number of redundant regulatory systems in muscle, and, in this situation, these negate the potentially positive effect of Epi on glycogenolysis. Nevertheless, the transient downregulation of glycogenolysis at the onset of exercise must be the result of a powerful stimulus in light of its potential for upregulation. It is well established that acetyl groups formed in the PDH reaction can be transformed from acetyl-CoA to acetylcarnitine via the carnitine acetyltransferase enzyme. Large increases in acetylcarnitine with equimolar decreases in free carnitine have been observed during intense exercise in horse (9) and human (10, 13, 23, 37, 38) muscle. Most authors concluded that carnitine appears to regulate the acetyl-CoA/CoA-SH ratio during exercise by buffering excess production of acetyl units. However, no studies have examined the acetylcarnitine/ acetyl-CoA relationship during prolonged intense exercise. In the present study (~80% \io, max), acetyl-CoA and acetylcarnitine contents increased significantly above rest at 15 min. The Aacetylcarnitine-to-Aacetyl-CoA ratio between rest and 15 min was 592 and 667 in the Pl and Caf trials, respectively (using mean data, Table 3). However, at exhaustion in both trials, acetyl-CoA and acetylcarnitine contents decreased below the 15min values. The decreases from 15 min to exhaustion occurred to different degrees such that the Aacetylcarnitineto-Aacetyl-CoA ratios were reduced to 309 and 346 in the Pl and Caf trials, respectively (ratio decreased in 7 subjects and did not change in 1). As fatigue approached and production of carbohydrate-derived acetyl-CoA decreased, the acetyl-CoA content decreased at a more rapid rate than the decrease in acetylcarnitine. A similar decrease in the Aacetylcarnitine-to-Aacetyl-CoA ratio has been reported in human skeletal muscle when comparing intense exercise to 10 min postexercise (13). These results suggest that the transport of acetyl groups out of the mitochondrion in the form of acetylcarnitine is a more facilitated process than transport in the reverse direction. The significance of this lag in the return of acetylcarnitine into the mitochondria is not readily apparent, since acetyl-CoA was maintained well above the resting values at exhaustion in both trials (1 possibility is discussed below). It should also be noted that the change in the Aacetylcarnitine-to-Aacetyl-CoA ratio in the present study demonstrates that muscle acetylcarnitine is a poor indicator of acetyl-CoA content during prolonged exercise. Therefore, using acetylcarnitine to predict acetyl-CoA (38) is not warranted in this situation. The results of the present study suggest that total depletion of acetyl-CoA is not the immediate cause of fatigue during intense cycling to exhaustion. Whether the decrease in acetyl-CoA concentration ([acetyl-CoA]) late in exercise contributes to fatigue is dependent on the in vivo relationship between acetyl-CoA and citrate synthase. If citrate synthase is always flux generating
DURING
PROLONGED
ES97
EXERCISE
during this type of exercise and therefore saturated with its substrate acetyl-CoA (32), this step would not contribute to fatigue. Although the Michaelis constant of citrate synthase for acetyl-CoA has been determined in vitro, the significance of these measurements to the in vivo situation in the mitochondrion is unknown. If substrate regulation of citrate synthase does contribute to fatigue, the above-mentioned lag in transport of acetyl groups back into the mitochondria may be important. Future work will require intramitochondrial measurements to assess the relationships between [acetyl-CoA], citrate synthase activity (tricarboxylic acid cycle), and fatigue during prolonged exercise. In summary, the ingestion of a high-caffeine dose 1 h before prolonged exercise decreases muscle glycogenolysis bY . m 55% in the initial 15 min of exercise at -80% vo 2 1TlaX.The spared glycogen is available during the later stages of exercise, resulting in a prolonged time to exhaustion. Increased utilization of intramuscular triacylglycerol and/or extramuscular FFA after caffeine ingestion may inhibit glycolysis at rest and early during exercise via elevations in muscle acetyl-CoA/CoA-SH ratio and citrate content. Muscle acetyl-CoA and acetylcarnitine contents increase dramatically at the onset of intense aerobic exercise. When the exercise is prolonged, acetyl-CoA and acetylcarnitine contents decrease in different proportions but are maintained above resting levels at exhaustion when muscle glycogen is depleted. It is not known whether the decreasing acetyl-CoA content contibutes to fatigue in this type of exercise. We thank Premila Sathasivam, Sandra Peters, and Agneta Laveskog for excellent technical assistance. This research was supported by grants from Sport Canada, the Natural Science and Engineering Research Council of Canada, the Swedish Medical Research Council (Grant no. 02647), the Swedish Sports Research Council, and the Karolinska Institute. Address for reprint requests: L. L. Spriet, School of Human Biology, University of Guelph, Guelph, Ontario NlG 2W1, Canada. Received 24 September 1991; accepted in final form 3 February 1992. REFERENCES 1. Aragon, J. J., J. K. Tornheim, and J. M. Lowenstein. On a possible role of IMP in the regulation of phosphorylase activity in skeletal muscle. FEBS Lett. 117, Suppl.: K56-K64, 1980. 2. Arnall, D. A., J. C. Marker, R. K. Conlee, and W. W. Winder. Effect of infusing epinephrine on liver and muscle glycogenolysis during exercise in rats. Am. J. Physiol. 250 Metab.
(Endocrinol.
13): E641-E649,
1986.
Bergmeyer, H. U. Methods of Enzymatic Analysis. New York: Academic, 1974. 4. Bergstrom, J. Percutaneous needle biopsy of skeletal muscle in physiological and clinical research. Stand. J. CLin. Lab. Invest. 35: 3.
609-616,
1975.
5.
Bergstrom, J., and E. Hultman. A study of the glycogen metabolism during exercise in man. Scar-d. J. Clin. Lab. Invest. 19:
6.
Brooks, S., M. E. Nevill, L. Meleagros, H. K. A. Lakomy, G. M. Hall, S. R. Bloom, and C. Williams. The hormonal response to repetitive brief maximal exercise in humans. Eur. J.
218-228,
Appl.
1967.
Physiol.
60: 144-148,
1990.
7.
Butcher, R. W., and E. W. Sutherland. Adenosine 3’,5’-phosphate in biological materials. I. Purification and properties of cyclic 3’,5’-nucleotide phoshodiesterase and use of this enzyme to characterize adenosine 3’,5’-phosphate in human urine. J. Biol.
8.
Cadarette, B. S., L. Levine, C. L. Berube, B. M. Posner, and W. J. Evans. Effects of varied dosages of caffeine on endur-
Chem.
237: 1244-1250,
1962.
ES98
9.
10, 11
12. 13.
14. 15. 16.
17.
CAFFEINE
AND MUSCLE
METABOLISM
ante exercise to fatigue. In: Biochemistry of Exercise, edited by H. G. Knuttgen, J. A. Vogel, and J. Poortmans. Champaign, IL: Human Kinetics, 1983, p. 871-876. Carlin, J. I., R. C. Harris, G. Cederblad, D. ConstantinTeodosiu, D. H. Snow, and E. Hultman. Association between muscle acetyl-CoA and acetylcarnitine in the exercising horse. J. Appl. Physiol. 69: 42-45, 1990. Carlin, J. I., W. G. Reddan, M. Sanjak, and R. Hodach. Carnitine metabolism during prolonged exercise and recovery in humans. J. Appl. Physiol. 61: 1275-1278, 1986. Cederblad, G., J. I. Carlin, D. Constantin-Teodosiu, P. Harper, and E. Hultman. Radioisotopic assays of CoASH and carnitine and their acetylated forms in human skeletal muscle. Anal. Biochem. 185: 274-278, 1990. Chasiotis, D., K. Sahlin, and E. Hultman. Regulation of glycogenolysis in human skeletal muscle at rest and during exercise. J. Appl. Physiol. 53: 708-715, 1982. Constantin-Teodosiu, D., J. I. Carlin, G. Cederblad, R. C. Harris, and E. Hultman. Acetyl group accumulation and pyruvate dehydrogenase activity in human muscle during incremental exercise. Acta Physiol. &and. 143: 367-372, 1991. Costill, D. L., G. Dalasky, and W. Fink. Effects of caffeine ingestion on metabolism and exercise performance. Med. Sci. sports 10: 155-158, 1978. Dibblee, L., and T. E. Graham. A longitudinal study of changes in aerobic fitness, body composition and energy intake in primigravid patients. Am. J. Obstet. Gynecol. 147: 908-914, 1983. Dobson, G. P., E. Yamamoto, and P. W. Hochachka. Phosphofructokinase control in muscle: nature and reversal of pH-dependent ATP inhibtion. Am. J. Physiol. 250 (Regulatory Integrative Comp. Physiol. 19): R71-R76, 1986. Erickson, M. A., R. J. Schwarzkopf, and R. D. McKenzie. Effects of caffeine, fructose and glucose ingestion on muscle glycogen utilization during exercise. Med. Sci. Sports Exercise 19: 579-583,
1987.
18. Essig, D., D. L. Costill, and P. J. Van Handel. Effects of caffeine ingestion on utilization of muscle glycogen and lipid during leg ergometer cycling. Int. J. Sports Med. 1: 86-90, 1980. 19. Forsberg, A. M., E. Nilsson, J. Wernernan, J. Bergstrom, and E. Hultman. Muscle composition in relation to age and sex. Clin. Sci. Lond. 81: 249-256, 1991. 20. Galbo, H., J. J. Holst, N. J. Christensen, and J. Hilsted. Glucagon and plasma catecholamines during P-receptor blockade in exercising man. J. Appl. Physiol. 40: 855-863, 1976. 21. Garland, P. B., and P. J. Randle. A rapid enzymatic assay for glycerol. Nature Lond. 196: 987-988, 1962. 22. Graham, T. E., and L. L. Spriet. Performance and metabolic responses to a high caffeine dose during prolonged exercise. J. Appl. Physiol. 71: 2292-2298, 1991. 23. Harris, R. C., C. V. L. Foster, and E. Hultman. Acetylcarnitine formation during intense muscular contraction in humans. J. Appl. Physiol. 63: 440-442, 1987. 24. Harris, R. C., E. Hultman, and L.-O. Nordesjo. Glycogen, glycolytic intermediates and high-energy phosphates determined in muscle biopsy samples of musculus quadriceps femoris of man at rest. Methods and variance of values. Stand. J. Clin. Lab. Invest. 33: 109-119, 1974.
DURING 25.
26.
27.
28.
29. 30.
31. 32. 33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
PROLONGED
EXERCISE
Hermansen, L., E. Hultman, and B. Saltin. Muscle glycogen during prolonged severe work. Acta Physiol. Stand. 71: 129-139, 1967. Hultman, E., and L. L. Spriet. Dietary intake prior to and during exercise. In: Exercise, Nutrition and Energy Metabolism, edited by E. S. Horton and R. L. Terjung. Toronto: MacMillan, 1988, p. 132-149. Ivy, J. L., D. L. Costill, W. J. Fink, and R. W. Lower. Influence of caffeine and carbohydrate feedings on endurance performance. Med. Sci. Sports 11: 6-11, 1979. Jannson, E., P. Hjemdahl, and L. Kaijser. Epinephrine-induced changes in muscle carbohydrate metabolism during exercise in male subjects. J. Appl. Physiol. 60: 1466-1470, 1986. Kavinsky, P. J., S. Shechosky, and R. J. Fetterick. Synergistic regulation of phosphorylase a by glucose and caffeine. J. Biol. Chem. 253: 9102-9106, 1978. Kjaer, M., J. Bangsbo, G. Lortie, and H. Galbo. Hormonal response to exercise in humans: influence of hypoxia and physical training. Am. J. Physiol. 254 (Regulatory Integrative Comp. Physiol. 23): R197-R203, 1988. Kjaer, M., P. A. Farrell, N. J. Christensen, and H. Galbo. Increased epinephrine response and inaccurate glucoregulation in exercising athletes. J. Appl. Physiol. 61: 1693-1700, 1986. Newsholme, E. A., and A. R. Leech. Biochemistry for the Medical Sciences. Toronto: Wiley, 1983. Oscai, L. B., D. A. Essig, and W. K. Palmer. Lipase regulation of muscle triglyceride hydolysis. J. Appl. Physiol. 69: 15711577, 1990. Randle, P. J., P. B. Garland, C. N. Hales, and E. A. Newsholme. The glucose-fatty acid cycle: its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet 1: 785-789, 1963. Rennie, M. J., and J. 0. Holloszy. Inhibition of glucose uptake and glycogenolysis by availability of oleate in well-oxygenated perfused skeletal muscle. Biochem. J. 168: 161-170, 1977. Richter, E. A., N. B. Ruderman, H. Gavras, E. R. Belur, and H. Galbo. Muscle glycogenolysis during exercise: dual control by epinephrine and contractions. Am. J. Physiol. 242 (Endocrinol. Metab. 5): E25-E32, 1982. Sahlin, K. Muscle carnitine metabolism during incremental dynamic exercise in humans. Acta Physiol. Stand. 138: 259-262, 1990. Sahlin, K., A. Katz, and S. Broberg. Tricarboxylic acid cycle intermediates in human muscle during prolonged exercise. Am. J. Physiol. 259 (Cell Physiol. 28): C834-C841, 1990. Saltin, B., and J. Karlsson. Muscle glycogen utilization during work of different intensities. In: Muscle Metabolism During Exercise, edited by B. Pernow and B. Saltin. New York: Plenum, 1971, p. 289-299. Spriet, L. L., J. M. Ren, and E. Hultman. Epinephrine infusion enhances muscle glycogenolysis during prolonged electrical stimulation. J. Appl. Physiol. 64: 1439-1444, 1988. Weiker, H., M. Feraudi, H. Hagele, and R. Pluto. Electrochemical determination of catecholamines in urine and plasma separations with HPLC. Clin. Chem. Acta 141: 17-25, 1984. Wu, T.-F. L., and E. J. Davis. Regulation of glycoytic flux in an energetically controlled cell-free system: the effects of adenine nucleotide ratios, inorganic phosphate, pH, and citrate. Arch. Biochem. Biophys. 209: 85-99, 1981.
