152

High Efficiency of Type I Muscle Fibers ImprovesPerformance

J F Horowitz, L. S. Sidossis, E. F CoyleHuman Perfonnance Laboratory, Department of Kinesiology and Health Education, The University of Texas at Austin

Abstract

J F Horowitz, L. S. Sidossis and E. F Coyle,High Efficiency of Type I Muscle Fibers Improves Perform-ance. Int. 1. Sports Med., Vol. 15, No.3, pp. 152-157, 1994.

Accepted after revision: September 28, 1993

We have recently demonstrated that peoplewith a high percentage of Type I muscle fibers display a rela-tively high muscular efficiency when cycling. These individu-als generate a relatively high muscular power output at a givensteady-state level of oxygen consumption and caloric expen-diture. The purpose of this study was to directly determine theextent to which differences in muscle fiber composition andefficiency influence endurance performance in competitivecyclists. The percentage of Type I and II muscle fibers wasdetermined from several biopsies from the vastus lateraliswhich were histochemically stained for ATPase activity.During a laboratory performance test, 14 endurance trainedcyclists (mean:!:SE; V02max, 5.2:!:0.11/min; body weight,74:!: 1kg) cycled an ergometer for 1h at the highest work ratethey could tolerate. V02 and RER were simultaneouslymeasured using open circuit spirometry for calculating caloricexpenditure. Subjects were divided into two groups of seven

according to their muscle fiber type composition: High %Type I Group (> 56 % Type I fibers); Normal % Type I Group(38-55 % Type I fibers). Each subject from High % Type IGroup was paired with a subject from the Normal % Type IGroup according to their similarity in V02max, blood lactatethreshold and average V02 maintained during the I h perform-ance test. Both groups averaged 4.5:!:O.1l/min during the 1hperformance test (i.e., 86-88 % V02max). However, theHigh % Type I Group, which possessed an average of72:t 3 % Type I fibers, was able to maintain a 9 % higherpower output (i.e., 342:!:9 vs 315:!:11 watts; p<O.OOI) thanthe Normal % Type I Group which possessed an average of48:t 2 % Type I fibers. Gross efficiency was thus signifi-cantly higher in the High % Type I Group compared to theNormal % Type I Group (i.e., 21.9:!:0.3 % vs. 20.4:!:0.3 %;p<O.OOI). We conclude that a high percentage of Type Imuscle fibers improves endurance performance ability by sig-nificantly increasing the power output generated for a givenrate of oxygen consumption and energy expenditure.

Key words

Endurance, bicycling, myosin type, muscularcontraction, physical exertion

Introduction

It is generally recognized that people can varygreatly in cycling efficiency (8) and running economy (4,9,23).Obviously, superior economy or mechanical efficiency haspotential to improve athletic performance. Indeed, running andwalking economy have been correlated with endurance per-formance ability (4,9,15). The factors responsible for the heter-ogeneity in economy and mechanical efficiency during exercisehave not been identified (19,20).

We have recently reported that both gross anddelta mechanical efficiency while cycling at a cadence of 80rpm is highly correlated (i.e., r= 0.75-0.85) with the percent-age of Type I muscle fibers within the vastus lateralis of well-trained cyclists (8). This observation in people indicates thatduring relatively slow velocity muscular contractions, Type I

Int. 1. Sports Med. 15 (1994) 152-157

@ Georg Thieme Verlag Stuttgart. New York

muscle fibers (i.e., slow twitch) appear more efficient at con-verting chemical energy into mechanical work compared toType II muscle fibers (i.e., fast twitch). This observation isconsistent with the previous work performed in vitro usingisolated muscle preparations (12,17).

The knowledge that exercise performance isgreatly influenced by economy (4,9,15), together with our re-cent observations that mechanical efficiency is related tomuscle fiber composition of the exercising musculature (8),leads to the obvious assumption that muscle fiber compositionmay directly influence exercise performance by improving me-chanical efficiency. However, no study to date has directly de-termined that improved mechanical efficiency as a result of arelatively high percentage of Type I muscle fibers indeed im-proves endurance performance. This was the primary purposeof this investigation.

PerfOrmance and Muscle Fiber Type Int. J. Sports Med. 15 (1994) 153

Another purpose of this investigation was toidentify the functional advantage of possessing a high percent-age of Type I muscle fibers during a cycling performance testlasting one hour. It has been suggested that a high percentageof Type I fibers allows for superior endurance running perform-ance (10). However, to our knowledge, the reason for this as-sumed superior performance has not been well studied. Im-proved mechanical efficiency is just one aspect which maypredispose individuals with a high percentage of Type I fibersto excel in endurance activities. Theoretically, Type I musclefibers could possess other functional and metabolic abilitieswhich make them superior for endurance performance, such assuperior oxidative ability and reduced muscle glycogenolysis(3). Therefore, this study also sought to determine if cyclingperformance is enhanced in individuals with a relatively highpercentage of Type I fibers, more than can be expected simplyas a result of superior mechanical efficiency.

Methods

Subjects and general design

Fourteen endurance-trained male cyclists wererecruited for this study. Subjects were informed of the experi-mental procedures and possible risks involved and each subjectsigned a consent form, approved by the Internal Review Boardof the University of Texas at Austin. All subjects performedpreliminary testing to determine maximal oxygen consumption(V02max), blood lactate threshold and muscle fiber type com-position of the vastus lateralis. The subjects were grouped ac-cording to their muscle fiber type composition (i.e., above orbelow the median of 56 % Type I muscle fibers). Subsequently,their endurance performance ability was evaluated during aone-hour laboratory test. Furthermore, based upon a similaraverage V02 during the one-hour laboratory test, each subjectfrom one group was paired with an individual from the othergroup, as described below.

Preliminary testing

V02max was determined while cycling(Monark ergometer model 819) using a continuous incrementalprotocol, lasting 7-10 minutes. V02max was defined as therate of oxygen consumption at which an increase in work rateno longer elicited an increase in V02. Additionally, this crite-rion coincided with an RER value of greater than 1.1.

V02 at the blood lactate threshold (V02 at LT)was assessed using a continuous cycling protocol. The bloodlactate threshold was determined by graphing venous bloodlactate concentrations (13), measured after five minutes of cy-cling at each offive intensities, corresponding to approximately50, 60, 70, 80 and 90 % of their predetermined V02max. Bloodlactate threshold was defined as the point at which blood lactateconcentration is elevated I mM above baseline values (5).

Muscle fiber type determination

Muscle samples were obtained from the vastuslateralis using the needle biopsy technique. Incisions in the skinwere made 12-20cm above the patella. To improve the ac-curacy ofthe muscle fiber type assessment, 2 to 4 biopsies wereperformed at different locations in the vastus lateralis of bothlegs, as previously described (8), in an attempt to obtain at least1,000 fibers for typing from each subject. Muscle samples were

prepared for histochemical analysis by first transverselysectioning the samples (10 /lm) using a Cryostat microtome.The samples were then pre-incubated at pH 4.3, 4.55 and 10.3and stained for ATPase activity (2). Fibers were classified asType I (i.e., slow twitch) or Type II (i.e., fast twitch) with noattempt to subclassify TypeII fibers because almost all Type IIfibers were Type IIa. Muscle fiber area was determined on atleast 100 fibers from each biopsy using computer based digitiz-ing pad and the % area Type I was determined.

One-hour performance test

Performance was assessed by measuring theminute by minute and ultimately the average power producedwhile the subjects cycled for one hour on a Monark cycleergometer (Model 819) equipped with racing handle bars andseat as well as pedals for cleated shoes. Power was measured bymonitoring pedal revolutions per minute (rpm) from an electriccounter (:t 1RPM) and resistance from a calibrated potentiome-ter attached to the pendulum (:t 1NM). The subjects were in-structed to generate the highest work rate possible throughoutthe one hour cycling bout. During the initial eight minutes ofthe exercise bout, the work rate was preset, based on a perform-ance prediction from the results ofthe preliminary blood lactatethreshold test (6,7). Following the initial eight minutes, thesubjects were allowed to vary both pedal cadence and resistanceto achieve the goal of producing the highest work rate possiblefor one hour. The subjects were supplied with visual feedbackof pedaling cadence, power output, heart rate, and elapsed time.V02 was measured during the initial eight minutes of the exer-cise protocol, as well as whenever the work rate was changed.V02 measurements were made for at least three minutes follow-ing the attainment of steady state. In the event that the subjectdid not request for the work rate to be altered, less than sevenminutes elapsed between steady state V02 measurements. Theaverage V02 during the 1h test was calculated from the steady-state values and the time at each work rate.

Selection of subject pairs

Subjects were divided into two groups accord-ing to their percentage of Type I muscle fibers. Each subjectwith a relatively high percentage of Type I muscle fibers (High% Type I) (i.e., greater than 56 % Type I) was paired with anindividual with a relatively average or normal distribution ofType I muscle fibers (Normal % Type 1) (i.e., less than 56 %Type I, with a range of38-55 % Type I). The Normal % TypeI group displayed a fiber type distribution typical of healthyyoung people (23). The pairing of subjects was based upon thefollowing criteria: 1) less than 2 % difference in VO2max,2) less than 4 % difference in V02 at lactate threshold,3) muscle fiber composition which differed by more than 10%Type I fibers of the total, and 4) less than 0.1 liter/min differ-ence in mean VO2 during the performance test.

Measurement of gas exchange

All measurements of oxygen consumption(V02) and respiratory exchange ratio (RER) were performedusing indirect calorimetry. While the subjects inhaled through aDaniels valve, the volume of inspired air was measured using aParkinson-Cowan CD4 dry gas meter. The expired gases werecontinuously sampled from a mixing chamber and analyzed for02 (Applied Electrochemistry SA3) and CO2 (Beckman LB-2).

~

154 Int. J. SportsMed. 15 (1994) J. F Horowitz, L. S. Sidossi, E. F Coyle

23These instruments were interfaced with an Apple lIe computerfor calculations of \7'02 and RER. Gross efficiency was calcu-lated as the ratio of the energy produced by the cyclist on theergometer (i.e., power output) to the r~te of caloric expenditure(i.e., indirect calorimetry) as previously described (8).

Statistics

Statistical differences were determined be-tween the matched subject pairs using Student's t-test for pairedcomparisons. Additionally, the subjects were ranked accordingto their individual gross efficiency, divided into two equalgroups, and the % Type I fibers of the groups were comparedusing Student's t-test for unpaired comparisons. The Pearsonproduct moment formula was used to calculate the correlationcoefficient. Statistically significant differences are reported atp<0.05.

Results

Subject pairs according to differencesin muscle fiber type composition

The population of subjects was fairly homo-geneous and possessed the following mean (I SE) charac-teristics: age, 25I I yr; weight, 74I I kg; V'Ozmax, 5.2I 0.111min; and \70z at LT, 3.97IO.16I1min. However, these subjectsexhibited a high degree of heterogeneity in muscle fiber typecomposition of the vastus lateralis. The percent of Type I musclefibers ranged from 38 % to 83 % (Table 1). As shown inTable 1, the subjects were divided into two groups according totheir percentage of Type I fibers (i.e., above and below 56 %Type I fibers) and then paired (i.e., 1 + 8, 2 + 9, 3 + 10, etc.).The High % Type I group (i.e., subjects 1- 7, Table 1)possessed an average of 73 I3 % Type I fibers whereas theNormal % Type I group (i.e., subjects 8-14, Table 1) possessedan average of 48 I 2 % Type I fibers. The percentage of musclefiber area composed of Type I fibers reflected a similar pattern(i.e., 72I3 % Type I area and 45I4 % Type I area for the Highand Normal groups, respectively). However, the respective Highand Normal % Type I groups were very similar regarding\70zmax (i.e., 5.2IO.l vs 5.l:t0.21/min), V'Oz at LT (i.e.,4.0IO.l vs 3.9IO.2l1min) and years of bicycle training (i.e.,6Il vs4Ilyrs).

..,-,

~ 22.......I:~

'<3

is 21r.o:I

r=0.75.'"'"Q...~ 20

. ...

1930 40 50 60 70

%Type I Fibers

80 90

Fig. 1 Correlation between the percent of Type I muscle fibersand gross efficiency (%) while cycling (r = 0.75; P < 0.001; n = 14).

One-hour performance bout

Table1 indicatesthat the averageV'Ozfor eachpair of subjects during the one-hour performance bout wasindeed within O.1l1min of each other and thus the High a1J.dNormal % Type I groups were identical regarding their averagerate of energy expenditure. This represented approximately 86-88 % of V'Ozmax. However, despite the similarity of energyexpenditure during the one-hour performance bout, the averagepower output of the High % Type I group was 9% (p<0.05)higher than the Normal % Type I group (i.e., 342I9 vs 315I11 watts; Table 1). When comparing each pair of subjects, in allcases the subject belonging to the High % Type I group wasmore powerful (Table 1). As shown in Fig. 1, gross cyclingefficiency during the one-hour performance bout was positivelycorrelated with % Type I muscle fibers (i.e., r = 0.75; p<O.OOI).The average pedalling cadences during the I h performancebout were nearly identical for the two groups (i.e., 91 I I vs89 I 1RPM for the Normal and High % Type I groups, respec-tively). These pedalling rates are typical for endurance cyclingevents (14).

High % Type I Group (subjects 1 - 7) possessed more than 56 % Type I fibers whereas the Normal % Type I Grouppossessed 38-55 % Type I fibers (subjects 8-14). The subjects were paired according to the criteria in methods section.'Significantly greater than Normal % Type I Group (p < 0.002)(By design, the groups differed in % Type I muscle fibers.)

Table 1 Comparison ofMuscle Fiber Composition Performance 'il02 Performance Power the % Type I muscle(% Type I) (I/min) (Watts) fiber composition, aver-

Subject High % Normal % High % Normal % High % Normal % age oxygen consump-Pair Type I Type I Type I Type I Type I Type I tion (\702) and average

power (watts) during the1 +8 83 46 4.62 4.67 357 335 1 h performance bout for2+9 77 54 4.39 4.38 336 293 the individual subjects3 + 10 76 49 4.59 4.54 363 336 and pairs.4 + 11 76 55 4.12 4.06 325 2875 + 12 70 38 4.61 4.66 3.24 3146 + 13 64 45 4.00 4.01 313 2877 + 14 62 52 5.00 4.91 376 359

Mean:l:SE 73:1:3* 48:1:2 4.48:1:0.13 4.46:1:0.13 342:1:9* 315:!: 11

, ' Performance and Muscle Fiber Type

,

Subjects grouped accordingto gross efficiency

In order to examine these data devoid of poten-tial investigator bias when selecting subject pairs, subjects wereranked based upon their gross efficiency during the 1h per-formance bout and then were formed into two equal groups ofseven subjects each. When this was done the gross efficiency ofone group was 22.10:t0.21 % (i.e, High Efficiency) whereasthe gross efficiency of the other group was 20.26:tO.20% (i.e.,Low Efficiency). These two groups displayed a significant(p <0.05) difference in muscle fiber type composition with theHigh Efficiency group possessing 70:t 4 % Type I fibers com-pared to 51:t4% Type I fibers for the Low Efficiency group.These two groups maintained an identical average VOl duringthe one hour performance bout (i.e., 4.47:t0.131/min), whilethe High Efficiency group generated 10% more power than theLow Efficiency group (i.e., 344:t8 vs 314:t 10 watts).

Discussion

The observation that mechanical efficiencywhile cycling is highly correlated with the percentage of Type Imuscle fibers within the vastus lateralis muscle (i.e., r = 0.75,p<O.OOI; Fig. 1) is in agreement with the data from our previ-ous work (8). The present investigation was specifically de-signed to examine how this relationship influences cycling per-formance. Clearly, superior mechanical efficiency while cyclingwould be expected to provide a performance advantage. How-ever, no study has directly determined that high mechanicalefficiency, associated with having a relatively high percentageof Type I muscle fibers, results in greater cycling performance.

Having access to a population of competitivecyclists with homogeneous characteristics and training, we wereable to create two groups that were identical regarding VOzmaxand VOl at LT, yet who differed greatly in their muscle fibercomposition within their vastus lateralis. Interestingly, duringthe one-hour laboratory performance test, the subject pairs, andtherefore the High and'Normal % Type I groups, maintained avery similar average VOl and thus similar rates of energy ex-penditure. Therefore, the 9 % greater average power outputduring the performance test displayed by the High % Type Igroup, compared to the Normal % Type I group, was due togreater gross efficiency when cycling. It should be realized thata 9 % mean difference in performance power among homo-geneously trained groups of competitive cyclists is quite large,with potential for much impact on their athletic standing. Evenmore remarkable, some of the subject pairs differed by as muchas 13-15 % in one-hour power performance despite comparablerates of oxygen consumption over the one hour period (i.e.,Table 1, subject 2 vs 9; subject 4 vs 11).

These data can be grouped and analyzed inseveral different ways, all with the same basic findings. Forexample, to avoid grouping subjects according to their musclefiber type, subjects were ranked and grouped according to theirgross efficiency during the performance test. In this case, theHigh Efficiency and Low Efficiency groups were significantlydifferent regarding their % Type I muscle fibers (i.e., 70:t4%vs 51 :t 4 % Type I fibers; p < 0.05). This indicates that the directpositive relationship between % Type I fibers and performanceability is robust in this population and not an artifact of themethod used for data analysis and presentation.

Int. J. Sports Med. 15 (1994) 155

The relationship between muscular efficiencyand muscle fiber type is dependent upon the velocity of con-traction (12, 17). Peak muscular efficiency occurs when the con-traction velocity is approximately one-third of the maximalvelocity of contraction (Vmax) of the specific muscle fiber(17). Since the Vmax of the Type II muscle fibers in humans isthree to five times higher than the Vmax of a Type I fiber (11),the absolute velocity at peak efficiency is also three to fivetimes higher in a Type II muscle fiber when compared to a TypeI muscle fiber. In mammalian muscle, the velocity of contrac-tion at peak efficiency for Type I and Type II muscle fibers havebeen reported to be about 1 and 5 fiber lengths/sec, respec-tively (11,12).

We have recently reported that while cycling atsimilar pedalling rates as employed in the present study (i.e.,80- 90 rpm), the vastus lateralis shortens at a rate between I and1.5 fiber lengths/ sec (8, 11). Clearly, this contraction velocity iscloser to the velocity of peak efficiency of a Type I fiber than aType II fiber. It is likely that the subjects in the High % Type Igroup were more efficient during the 1h performance bout thanthose in the Normal % Type I group because while cycling, agreater percentage of their active musculature was contractingat a velocity close to the velocity of peak efficiency of Type Imuscle fibers. .

A high percentage of Type I muscle fibers hasbeen associated with superior 10km running performance (10).In general, this observation has been assumed to indicate thatpossessing a relatively high percentage of Type I muscle fiberspredisposes those individuals to have a greater aerobic capacity,possibly as a consequence of Type I muscle fibers having moremitochondria and thus higher oxidative ability than Type IIfibers. Indeed, in untrained individuals, mitochondrial enzymeactivity in Type I muscle fibers is roughly two-fold greater thanthat of Type II muscle fibers (3,18). However, this is not thecase in well-trained endurance atWetes who have been trainingintensely for several years (3,6,16), such as the subjects of thepresent study. Endurance athletes performing high intensitytraining for prolonged periods display almost equal mito-chondrial activity of the Krebs cycle enzymes in Type I andType II fibers (3,18), although enzymes related to fat oxidationand glycolysis are respectively higher and lower in Type I com-pared to Type II fibers (3).

In agreement with the idea that Type I and TypeII fibers in well-trained endurance atWetes do not differ inoverall oxidative ability (i.e., to oxidize pyruvate), we foundthat the Normal and High % Type I groups were similar in bothVOzmax and the percentage of VOzmax maintained for the 1hperformance bout (i.e., 86-88 % VOzmax). Therefore, we havefound no compelling evidence or strong support for the conceptthat Type I fibers in well-trained endurance athletes improveperformance by allowing higher rates of energy expenditure. Itshould be realized, however, that the present study was notoptimally designed to detect functional differences in the oxi-dative ability of Type I and Type II fibers of endurance athletesbecause the present subjects were paired based upon a similarVOzmax and VOl during the 1h performance bout. These ob-servations serve only to indicate that differences between TypeI and Type II fibers in well-trained endurance athletes do notendow these athletes with markedly different oxidative functionas reflected in VOzmax, blood lactate threshold or the percent-age of VOzmax maintained for 1 hour. Therefore, well-trained

156 Int. J. Sports Med. 15 (1994) J. E Horowitz, L. S. Sidossi, E. E Coyle

endurance athletes with a greater percentage of Type I musclefibers are able to generate more power during high intensityendurance performance lasting about one hour compared towell-trained endurance athletes with a normal percentage ofType I muscle fibers, not because of a greater ability to resyn-thesize ATP aerobically, but apparently due to superior effi-ciency in converting the chemical energy from ATP hydrolysisduring contraction, into mechanical work.

Regarding athletic performance, gross effi-ciency is the most practical and functional measure because itreflects whole body oxygen consumption. Gross efficiency re-flects energy expenditure from all bodily processes. It has beenrecently shown that during intense exercise the vast majority ofoxygen is consumed by the exercising muscles for power pro-duction and that V02 measurements at the mouth provide avalid reflection of changes within the exercising leg muscles(21,22).

The potential importance of mechanical andmuscular efficiency to endurance performance is well recog-nized, especially in sports such as running, racewalking, andswimming. Since calculation of the actual physical work ac-complished during these activities is not practical, it is commonto simply determine the V02 required for movement at a givenvelocity, and term this "economy". For a long time we haverecognized that running and racewalking performance is signif-icantly related to the economy of motion (9, 15). Using a similarapproach to our present study, Conley and Krahenbuhl (4) as-sembled a group of competitive runners who were very similarin training and V02max. In this homogeneous population ofrunners they found that most of the variation in 10kIDrunningperformance was accounted for by differences in runningeconomy. This is similar to our present findings that cyclingperformance is largely determined by gross cycling efficiencyin our group of homogeneous cyclists.

However, despite extensive research, investiga-tors have failed to identify the biomechanical or physiologicalfactors that are primarily responsible for determining runningeconomy (19,20). We think it is quite possible that muscle fibertype also influences running economy, although we recognizethat the mechanics of running are more complex than cyclingand thus prone to the variable influence of other factors. Boscoet al. (1) reported a significant positive correlation betweenenergy cost while running and the percentage of Type II (i.e.,fast twitch) fibers. However, these investigators suggested thatthe apparent enhanced economy of those individuals with agreater percentage of Type I muscle fibers is a result of thesuperior elastic energy storage capacity of these fibers at thespecific slow velocities of muscle contraction while running(1). Williams and Cavanagh (24), failed to detect a significantdifference in muscle fiber type among three groups of runners(n= 31) who exhibited high, medium and low runningeconomy, although there was a trend for the groups with thehighest economy to possess more Type I muscle fibers.

In conclusion, two groups of competitive en-durance-trained cyclists with large differences in muscle fibercomposition within the vastus lateralis muscle (i.e., 73:!::3 vs48:!::2% Type I muscle fibers) were identical in V02max andV02 at the blood lactate threshold as well as the average V02and rate of energy expenditure that could be maintained duringa 1h performance test. However, gross efficiency while cycling

was highly related to % Type I muscle fibers (r = 0.75,p<O.OOI). Therefore, although energy expenditure during the 1h performance test was similar, the group of cyclists with ahigher than normal percentage of Type I muscle fibers were ableto generate 9 % (p < 0.002) more power throughout the 1h per-formance bout. In competitive endurance-trained cyclists, itseems clear that gross efficiency is positively related to %Type I muscle fibers and that cyclists with a relatively highpercentage of Type I muscle fibers display at least a 9 % greaterperformance ability than equally well-trained cyclists possess-ing a normal muscle fiber type composition.

References

I Bosco C., Montanari G., Ribacchi R., Giovenali P., Latteri E, lachelliG., Faina M., Colli R., Dal Monte A., La Rosa M., Cortili G.,Saibene E: Relationship between the efficiency of muscular workduring jumping and the energetics of running. EurJ Appl Physiol56:138-143, 1987.

2 Brooke M. H., Kaiser K. K.: Three "myosin ATPase" systems: thenature of their pH lability and sulfhydryl dependence. J HistochemCytochem 18: 647-672, 1970. .

3 Chi M. M.-Y., Hintz C. S., Coyle E. E, Martin W H. I., Ivy J. L.,Nemeth P. M., Holloszy 1. 0., Lowry O. H.: Effects of detraining onenzymes of energy metabolism in individual human muscle fibers.Am J PhysioI244: C276-C287, 1983.

4 Conley D. L., Krahenbuhl G. S.: Running economy and distancerunning performance of highly trained athletes. Med Sci SportsExerc 12: 357-360, 1980.

5 Coyle E. E, Martin W H., Ehsani A. A., Hagberg 1. M., BloomfieldS. A., Sinacore D. R., Holloszy 1.0.: Blood lactate threshold in somewell-trained ischemic heart disease patients. J Appl Physiol 54: 18-23, 1983.

6 Coyle E. E, Coggan A. R., Hopper M. K., Walters T. 1.:Determinantsof endurance in well trained cyclists. J Appl Physiol64: 2622-2630,1988.

7 Coyle E. F., Feltner M. E., Kautz S. A., Hamilton M. T., Montain S.1., Baylor A. M., Abraham L. D., Petrek G. W: Physiological andbiomechanical factors associated with elite endurance cycling per-formance. Med Sci Sports Exerc 23: 93-107, 1991.

8 Coyle E. E, Sodissis L. S., Horowitz 1. E, Beltz 1. D.: Cyclingefficiency is related to the percentage of Type I muscle fibers. MedSci Sports Exerc 24: 782-788, 1992.

9 Farrell P. A., Wilmore 1. H., Coyle E. E, Billing 1. E., Costill D. L.:Plasma lactate accumulation and distance running performance. MedSci Sports Exerc II: 338-344, 1979.

10 Fink W 1., Costill D. L., Pollock M. 1.: Submaximal and maximalworking capacity of elite distance runners. Part II. Muscle fibercomposition and enzyme activity. Ann N Y Acad Sci 301: 323-327,1977.

11 Fitts R. H., Costill D. L., Gardetto P. R.: Effect of swim exercisetraining on human muscle fiber function. J Appl Physiol 66: 465-475, 1989.

12 Goldspink G.: Energy turnover during contraction of different typesof muscle. In: Biomechanics VI-A. E. Asmussen, K. Jorgensen(Ed.), Baltimore: University Park Press, 1978, pp 27-39.

13 Gutman 1., Wahlefeld A. W: L-(+)-lactate. Determination with lac-tate dehydrogenase and NAD. In: Methods of Enzymatic Analysis(2nd ed.). H. U. Bergmeyer (Ed.), New York: Academic, 1974, pp1464-1468.

14 Hagberg 1. M., Mullin 1. P., Giese M. D., Spitznagel E.: Effect ofpedalling rate on submaximal exercise responses of competitive cy-clists. J Appl Physiol Respir Environ Exerc Physiol 51: 447-451,1981.

15 Hagberg 1. M., Coyle E. E: Physiological determinants of enduranceperformance as studied in competitive racewalkers. Med Sci SportsExerc 15: 287-289, 1983.

16 Jansson E., Kaijser L.: Muscle adaptation to extreme endurancetraining in man. Acta Physiol Scand 100: 315-324, 1977.

Performance and Muscle Fiber 1j;pe Int. J. Sports Med. 15 (1994) 157

[7 Kushmerick M. J., Davies R. E.: The chemical energetics of muscle

contraction II. The chemistry, efficiency, and power of maximallyworking sartorius muscle. Proc R Soc Lond Ser B 1174: 315-353,1969.

18 Lowry C. v., Kimmey 1. S., Felder S., Chi M. M.- Y., Kaiser K K,Passonneau P. N., Kirk K. A., Lowry O. H.: Enzyme patterns insingle human muscle fibers. J Bioi Chern 253: 8269-8277, 1978.

19 Martin P. E., Morgan D. w: Biomechanical considerations for eco-nomical walking and running. Med Sci Sports Exerc 24: 467-474,1992.

20 Morgan D. W, Craib M.: Physiological aspects of running economy.

Med Sci Sports Exerc 24: 456-461, 1992.21 Poole D. c., Schaffartzik W, Knight D. R., Derion T., Kennedy B.,

Guy H. J., predi1etto R, Wagner P. D.: Contribution of exercisinglegs to the slow component of oxygen uptake kinetics in humans. JAppl Physiol7: 1245-1253, 1991.

22 Poole D. C., Gaesser G. A., Hogan M. C., Knight D. R., Wagner P.D.: Pulmonary and leg V02 during submaximal exercise: implica-

tion for muscular efficiency. J Appl Physiol72: 805-810, 1992.23 Saltin B., Gollnick P. D.: Skeletal muscle adaptability: significance

for metabolism and performance, in: Peachy L. (ed): Handbook ofPhysiology, Section 10: Skeletal Muscle, Bethesda, MD: AmericanPhysiological Society, 1983, pp 555-631.

24 Williams K. R., Cavanagh P. R.: Relationship between distance run-ning mechanics, running economy, and performance. J Appl Physiol63: 1236-1245, 1987.

E. F Coyle, Ph. D.

Human Performance Laboratory

The University of Texas at AustinAustinTexas 78712U.S.A.