Re Sports Medicine 7: 310-330 (1989) 01 12-1642/89/0005-0310($ 10. 50/0 C> ADIS Press Limited All rights reserved.

Factors Affecting Running Economy Don W. Morgan. Philip E. Martin and Gary S. Krahenbuhl Department of Physical Education. Sport. and Leisure Studies. Washington State University. Pullman. Washington. and Department of Health and Physical Education. Arizona State University. Tempe. Arizona. USA

Co"te"ts

Summary ......... ..... ....... ..... .. ................ ........ ........ .. .. ..................... ........... ..... .... ..... ... .. ............. .. I. Measurement of Running Economy .............. .......................... ...... .......... ......... ............ .... 2. Relationship Between Running Economy and Performance ......... ... ......... ....... .............. 3. Physiological and Environmental Factors Affecting Economy ........................... ............ 3. 1 Intraindividual Variability in Running Economy .................... ....... ...... ... ...... ....... .... 3.2 Gender ....................... ......... .. ............................. ........ ....... ..... .................... .. ... ... .... ... .. .. 3.3 Age ............ ........ ..... ................. ....................... ........... ... ........... ............................ ........... 3.4 Treadmill and O verground Running ................ ........ .. .... ........... ......... ...................... .. 3.5 Temperature .. .......... ............................... ........... ......... .... ... ................ ....... ......... ...... ...... 3.6 Fatigue ..................................... ...... ... ..................................................... .... .. .................. 3.7 Training .................................. .. .... ..... .. ............................................... ... ...... ........ ........... 4. Biomechanical Considerations for Economical Running ........................... ..................... 4.1 Relationship Between Body Structure and Economy...... .......... ... .. ...... .................... 4.1 . 1 Body Mass .................................. ...... ......... ....................... ................... .... ............. 4.1.2 Body and Segment Mass Distribution ................. .............................................. 4.2 Relationship Between Running Kinematics and Economy .......... ........... .... ............. 4.2.1 Running Speed ... .................. ... ........................ ....... ............................... .............. 4.2.2 Stride Length ....... ............. ....... ..... .......................... ............... ........... ...... .. .. ... .... .. 4.2. 3 Other Kinematic Variables .................. ............... ........................ ... .. .. ...... .. ..... .... 4.3 Relationship Between Running Kinetics and Economy ............. .... ....... ........ .......... 4.3. 1 Ground Reaction Forces .............................................................. .... .. ............. .... 4.3.2 Mechanical Power 5. Future Research Directions .............................................. ........... ..................................... .

SIImmary

Running economy. defined as the steady·stalE VOl for a given running velocity. has been shown to account for a large and significant proportion of variation in distance· running performance among runners roughly comparable in V02 max· Despite this rec· ognition. relatively lillie is known regarding the potpourri of physiological. environmental. structural and mechanical faclOrs potentially associated with a lower aerobic demand of running. Early allempts at quantifying the energy expenditure of exhaustive runs incorporated measurements of oxygen consumption before. during. and after exercise. The validity of this approach has been questioned. however. since recent evidence has demonstrated that only a moderate relationship exists between postexercise V02 and anaerobic metabolism. The energy demands for submaximal running (i.e. running economy) can be quantified

310 312 313 315 315 316 317 317 318 318 319 320 320 320 321 322 322 322 323 323 323 324 325

Factors Affecting Running Economy

311

by calculating the steady-state V02. expressed with respect 10 body mass and time. for a standardised. submaximal running speed. Since this l'ariable represents the aerobic demand of running. the generation of energy must derive wholly from cell respiration and not from substantial protein catabolism. Research has indicated that at low to moderate work rates. the steady-state energy condition is allained in about 3 minutes. Trained individuals reach steady-state sooner that unfit subjects. While limited by methodological constraints. the existence of a steady-state has also been verified by the lack of blood lactate accumulation and the presence of a respiralOry exchange ratio of less than 1.00. The ability of economy. either singly or in combination with V02 max. to account for a substantial portion of performance variation among trained distance runners and untrained subjects of comparable ability and fitness level has been demonstrated in recent cross-sectional studies. Limited data from short and long term longitudinal research also suggests that endurance running success is linked 10 training and growth-related improvements in economy. Intraindividual variation in economy has been shown to vary between 2% and 11% for a given speed. Most of this variation can probably be allributed to biological error. While the majority of evidence does not support a gender difference in running economy. data from some studies suggest that males may be more economical than women. Prepubescent children are less economical than older children and adults. whereas older adults exhibit the same trend when compared to younger counterparts. Because of air and wind resistance. the aerobic demands of indoor treadmill running significantly underestimate the cost of overground running. especially at higher speeds. As body temperature rises during exercise. V02 increases as a result of the 'QIO effect.' While conflicting data exist with respect 10 the e.ffect offatigue on the aerobic demand of running. recent work incorporating physiological and biomechanical measures demonstrated that a 30-minute maximal run produced little or no change in the metabolic and biomechanical profiles of trained runners. No consensus exists regarding the effects of different types and intensities of training on running economy. Substantial variation in economy among long distance runners who compete in the same event suggests non-training faclOrs may also influence economy. A number of anthropometric and biomechanical factors have been considered for their ability to account for some of the interindividual variability in running economy commonly observed. Despite assumptions 10 the contrary. it seems appropriate to conclude that when the confounding influence of speed is negated. few biomechanical variables have been shown consistently to account for a substantial portion of variation in economy. It has been suggested that at present it is not possible to distinguish whether mechanical l'ariables describing the running pallern of an uneconomical runner contribute 10 making the runner uneconomical. or whether the pal/ern reflects the means by which the individual has optimised his or her own anatomical and physiological features.

Determinants of successful distance running performance include maximal aerobic power (Y02 max) [Boileau et al. 1982; Brandon & Boileau 1987; Hagan et al. 1981; Saltin & Astrand 1967], fractional utilisation of YO:! max (Conley et al. 1981 b, 1984; CostilI et al. 1973; Leger et al. 1986; SjOdin & Svendenhag 1985), lactate threshold (AlIen et al. 1985; Farrell et al. 1979; Heck et al. 1985; Jacobs 1986; SjOdin & Jacobs 1981), muscle respiratory capacity and skeletal muscle fibre type (Costill et al. 1976a,b; Ivy et al. 1980a,b), relative

leanness (Pollock et al. 1977; Wilmore et al. 1977), and fuel supply (Bergstrom et al. 1967; Sherman & Lamb 1988). A number of investigators have also demonstrated that running economy, the aerobic demand (Y02) for a given submaximal running speed, is an important correlate of endurance running performance among individuals roughly comparable in Y02 max (Bransford & Howley 1977; Conley & Krahenbuhl 1980; Conley et aI. 1981a,b, 1984; Daniels 1985; Daniels et al. 1977, 1978a; Krahenbuhl et al. 1989; Morgan et al. 1989).

Factors Affecting Running Economy

The recent emergence of running economy as an attribute of distance running success belies the relative paucity of information currently available regarding the myriad of variables potentially linked to the metabolic demands of locomotion. The main objective of this review, therefore, is to discuss the measurement of running economy and to synthesise research findings with respect to: (a) the relationship between running economy and endurance running performance; (b) physiological and environmental factors affecting running economy; and (c) inter-relationships among running economy, running mechanics, and structural characteristics. The review will conclude by addressing future research questions related to the study of running economy.

1. Measurement 0/ Running Economy Initial attempts to quantify the total oxygen requirement during running incorporated ~02 measurements before, during, and after running bouts performed at various speeds (Adams & Bernauer 1968; Furusawa et al. 1924; Hill & Lupton 1922, 1923; Robinson et al. 1958; Sargent 1926). The conceptual basis for this assessment of energy expenditure was that aerobic and anaerobic metabolism contributed to the overall energy cost of running, particularly during high intensity exercise. In light of a contemporary understanding of the relationship between recovery oxygen consumption and oxygen deficiency during exercise (Gaesser & Brooks 1984), postexercise oxygen consumption, or the oxygen debt, has recently been termed an inadequate measure of energy metabolism during intense exercise (Brooks & Fahey 1984). Additionally, estimation of the anaerobic contribution during high intensity exercise is tenuous and dependent upon knowledge of the concentrations of adenosine triphosphate (A TP), creatine phosphate, muscle glycogen and lactate, the total water pool in the body available for lactate uptake, the distribution of lactate between extra- and intracellular water, and the amount of exercising muscle mass (Astrand & Rodahl 1986). In contrast to the problems inherent in quan-

312

tifying the energy cost of maximal or near-maximal running, the aerobic demand of submaximal running, i.e. running economy, can be obtained by measuring the steady-state ~02 for a standardised running speed. The use of indirect calorimetry to accurately reflect metabolic rate during exercise is based on 2 assumptions. The first assumption is that the ATP requi~ement derives wholly from cell respiration, and not from phosphagen breakdown or the anaerobic degradation of carbohydrates (Brooks & Fahey 1984). Several researchers (Margaria et al. 1963b, 1969; Whipp et al. 1970) have supported the legitimacy of this notion during submaximal exercise. At non-steady-state exercise rates, the contribution of anaerobic metabolism constitutes a portion of the total energy expenditure of active muscle. Consequently, the aerobic demand of running at near-maximal speeds may underestimate the true energy cost (Bransford & Howley 1977). Use of indirect calorimetry to assess the energy demands of running also assumes that the contribution of protein and amino acid degradation to the active energy requirement is insignificant (Brooks & Fahey 1984). With regard to this point, it has been demonstrated that protein catabolism increases: (a) with the severity of exercise; (b) during prolonged, exhaustive exercise; and (c) in glycogen-depleted subjects (Ahlborg et al. 1974; Felig & Wahren 1971; Lemon & Mullin 1980; White & Brooks 1981). Given the short duration (6 to 10 minutes) and submaxima1 nature of running economy tests, the validity of this assumption is probably not in question, and any error associated with its use is most likely trivial. Attainment of a steady-state ~02 condition is dependent upon ~02 kinetics. A number of studies (Cerretelli et al. 1966; DiPrampero et al. 1970; Henry 1951; Margaria et al. 1965) have shown that the half-time ~02 response is about 30 seconds, while more recent investigations (Hagberg et al. 1978; Whipp & Wasserman 1972) have indicated that ~02 kinetics vary with work intensity and fitness level. At low and moderate work rates, Whipp and Wasserman (1972) found that a steady-state condition was achieved within 3 minutes. These authors also reported that while attainment of a

313

Factors Affecting Running Economy

steady-state condition was delayed at higher work rates. the VO~ at any point prior to reaching steadystate was higher in a well-trained subject. This suggests that the contribution of non-aerobic energy sources during submaximal running is a function of fitness status. Other criteria which have been used to verify the existence of a steady-state energy condition include determination of the anaerobic threshold and the existence ofa respiratory exchange ratio (RER) less than unity. Various methods have been used to locate the an aerobic threshold. defined as the V02 or running velocity below which the rate of muscle lactate production does not exceed lactate uptake or removal (MacDougall 1977: Tanaka et al. 1983). One approach has been to utilise changes in ventilatory parameters (e.g. non-linear increases in ventilation and the volume of expired C02. increases in end-tidal 02 with no commensurate drop in end-tidal C02. and an abrupt increase in the RER) as indicative of greater reliance upon anaerobic energy sources. While the advantage of this approach is that it does not require direct blood or muscle lactate measurements. it sutTers from drawbacks such as the lack of sensitivity in locating the 'breakaway' point at which disproportionate increases in gas exchange indices occur relative to V02 (Green et al. 1983: Orr et al. 1981) and the asynchronous manifestation of blood lactate and ventilatory thresholds (Hagberg et al. 1981: Walsh & Banister 1988). As a result of these methodological considerations. many researchers have chosen to determine the anaerobic threshold from measurements of blood lactate. As with ventilatory determination of this variable. a number of methods have been employed to identify the exercise intensity beyond which blood lactate accumulates disproportionately . These include \isual inspection of the lactate-V02 plot to determine the point at which lactate concentrations increase abo \'e a esting r level (Henritze et al. 1985: Hughes et al. 1982: h'Y et al. 1980b). the generation of linear regression equations above and below the curvature of the lactate plot (Farrell et al. 1979: Senay & Kok 1977). the repetition of prolonged tests to locate the intensity

at which lactate entry into the blood is equivalent to its maximal elimination from blood and muscle compartments (Heck et at. 1985: Stegmann & Kindermann 1982). and objectively defining the threshold as the velocity associated with a 4 mmol/ L blood lactate concentration (Kohrt et at. 1987: Mader et at. 1976: Morgan et at. 1989: Sjodin & Jacobs 1981). While each of these approaches offers specific advantages. they also sutTer from limitations such as the subjectivity inherent in visually determining the lactate threshold. the need to collect a large number of data points above and below the breakpoint. the extensive amount oftesting required to locate a particular individual's anaerobic threshold. and interindividual variability. An RER (V'C02/V02) of tess than 1.00 has also been used to support the presence of a steady-state metabolic condition (Bransford & Howley 1977: Conley & Krahenbuhl 1980: Conley et at. 1981 b). During low to moderate intensity exercise. metabolic C02 production and low blood lactate concentrations generate V'C02 values which do not exceed the VO:! requirements of muscle. At higher work rates. increased C02 concentrations and greater blood lactate accumulation result in a disproportionate increase in ventilation and V'C02 relative to YO:!. thus producing a rise in the RER above unity. Since this approach incorporates ventilatory parameters to reflect changes in blood lactate concentration. the use of this measure as evidence of a steady-state level of energy expenditure is probably limited.

2. Relationship Between Running Economy and Performance Until recently. intcrindividual ditTerences in the aerobic demand of running were considered of little use in discriminating performance variability among runners ( ..... por et al. 1980: Costill et at. 1973). In the last 20 years. however. investigators have focused greater attention on the predictive and diagnostic value of running economy. Initial work

Factors Affecting Running Economy

314

70

B

co

...r::

&

~

~

230

250

270

290

310

VV02 max (m/min)

A

"

330

350

FIg. 1. Relationship between Vo2 and predicted velocity at

V02 ""'. (VVo2 mo.) In 2 well· trained subjects (A and B) with dlf· ferent IOkm run times. Reproduced with permission of Morgan et II. (1989).

by Costill and Winrow (1970) suggested that performance variability in 2 middle-aged ultramarathon runners with similar ~02 max values (a = 1.4 ml/kg/min) could be attributed to individual differences in economy. In a study of 12 elite distance runners of comparable ability and fitness level conducted several years later, Conley and Krahenbuhl (1980) confirmed the results of Costill and Winrow and obtained significant correlations between running economy and IOkm run time ranging from 0.79 to 0.83. The ability of economy to account for nearly identical 2-mile run times (10:31 and 10:35) among 2 champion male runners with different ~02 max values (a> 10 ml/kg/min) was also documented in an early study conducted by Daniels (1974). More recently, the interrelationships among running performance, ~02 max, and running economy among subjects homogeneous in V02 max have been examined in cross-sectional work by Morgan et at. (1989). In their study, the predicted running velocity at ~02 max (VV02 max) for 10 well-trained runners was derived by combining the relative contributions of ~02 max and economy (Daniels 1985). Results from the study demonstrated that the strength of the relationship between IOkm run time and VV02 max (r = -0.87) exceeded that associated with either running econ-

omy (r = 0.64) or V02 max (r = -0.45). and was similar to that explained by treadmill velocity at a 4 mmol/L blood lactate concentration (r = -0.82) [Jacobs 1986; SjOdin & Jacobs 1981]. A graphic depiction of the utility of VV02 max to account for performance differences in this subject cohort is shown in figure I. Two subjects who displayed similar V02 max values (a = 1.2 ml/kg/min) and dissimilar IOkm run times (a = 3.52 min). for instance, exhibited different VV02 max values consistent with individual variations in economy (~ = 13.4 ml/kg/km). Conversely, data presented in figure 2 for 2 additional subjects indicated that different combinations of V02 max and economy produced nearly identical 10km run times and VV02 max values. Illustrations of the interplay between V02 max and submaximal V02 are provided in short term longitudinal studies of several champion runners (Conley et at. 1981a, 1984). During a 6-month period of training, American mile record holder Steve Scott improved his V02 max to 77.2 ml/kg/min from an off-season low of 74.4 ml/kg/min. During the same period, his aerobic demand running at 268 m/mindecreased to 45.3 ml/kg/min from an off-season high of 48.5 ml/kg/min. The combined improvement in maximal aerobic power and running economy reduced the relative intensity of run-

70 _60

.1:::

...r::

~

s.

~50

0

~ ~

230

250

270

290

310

vV0 2 rna,(m/min)

330

350

Fig. 2. Relationship between V02 and predicted velocity at V02 m.. (VV02 m..) in 2 well-trained subjects (e and 0) with similar 10km run times. Reproduced wi1h permission of Morgan at al. (1989).

Factors Affecting Running Economy

ning at 268 m/min from 65.1 % to 58.6% of V02 max (Conley et al. 1984). A more intensive longitudinal study of an Arizona roadracing champion provided further clarification of this phenomenon. In this study. Conley et al. (1981 a) tested the subject weekly during 18 weeks of training. During this time, the subject's V0 2 max increased from 70.2 to 75.1 ml/kg/min. In the same time period. his economy at 295 m/min improved from a pretraining high of 58.7 to 53.5 mljkg/min at the end of training. As an indication of the combined etTect of these changes, the individual would have covered an additional 960 metres in a 30-minute run at his race pace intensity of 93% V02 max. While limited in number. long term longitudinal studies have lent further credence to the importance of economy in distance-running performance. Daniels et al. (1978a) studied 20 young boys (10 to 13 years) who had engaged in middle and long distance running training for 2 to 5 years. While no change in relative V02 max was observed. significant reductions in 1- and 2-mile run times were accompanied by lower aerobic demands of running. Extending Daniels' work, Krahenbuhl et at. (1989) studied active, non-run trained boys over a 7-year period, starting when the subjects were 10 years of age. Results from their investigation showed that relative V02 max remained stable, but 9-minute run distance increased by 29%. Improvements in running performance were coincident with a 13% reduction in submaximal V0 2 expressed relative to distance travelled (mljkg/km). They also noted that the mean estimated percentage of V0 2 max incurred during a 9-minute endurance run at 17 years of age was 16% higher than the value recorded at 10 years of age. Figure 3 depicts the improvement in children's running velocity (the pace maintained for an endurance run of a fixed time) which would result from both an economical running style and the ability to generate a higher relative percentage of V02 max' Seasonal variations in economy and distancerunning performance have also been Quantified in elite adult distance runners (Svendenhag & Sjodin 1985). Over a mean period of 22 months. signifi-

315

110 100

(\o{\\~

90

ho 070 .>

ef-o

cri~O\\~(\\

~

eCo(\rfi0

p.,oo\e~

i 60

A

B

c

240 200 220 160 180 Treadmill running pace (m/mln)

Fig. 3. Variations in treadmill running pace associated with differences in running economy and the percentage of V02 max Incurred during a 9-minute run in younger (age 10) and older (age 17) subjects. A = treadmill pace in a younger subject running at 85% V02 max; B = treadmill pace in a younger subject running at 99% V02 max; C = treadmill pace In an older subject running at 99% V02 max. Reproduced with permission of Krahenbuhl et al. (1989).

cant reductions in the aerobic demands of running at IS and 20 km/ h were observed in 16 male runners who underwent alternating periods of slow distance. uphill, and interval training. Faster 5000m run times were also observed during the course of the athletes' training. In discussing their findings. the authors suggested that the enhancement in running performance which occurred after V02 max plateaued may have been associated with a slow, steady improvement in economy.

3. Physiological and Environmental Factors Affecting Economy 3.1 Intraindividual Variability in Running Economy The magnitude of normal intraindividual variation in running economy has received scant attention. This is unfortunate. since knowledge of the daily stability of running economy is essential if the efficacy of a particular intervention aimed at modifying economy is to be evaluated. Small sample sizes and the limited number of test sessions typically employed in variability studies restrict the

Factors Affecting Running Economy

degree to which meaningful conclusions can be drawn regarding running economy stability. Daniels et al. (1984b) measured running economy in lOwell-trained male subjects who performed 15 treadmill tests (4 equally-spaced testing periods composed of 3 or 6 runs each) at 268 ml min during a 7-month period. Results from their investigation showed that when running speed, learning, footwear, and test equipment were controlled, individual stability in running economy varied by as much as II % within a particular test period. In related work involving 10 elite male runners who performed four 6-minute treadmill tests at 230, 248, 268, and 293 mlmin on 3 separate days, Morgan et al. (1987) found that daily variation in running economy ranged from 3% to 5% of the mean ~02 at each speed. They also reported that the largest within-subject variation in economy represented 9% of the mean ~02 values. Because circadian variation, training activity, and length of treadmill accommodation were not strictly controlled in the previously-cited studies, it was not possible to apportion within-subject economy variation into biological and non-biological components. In an attempt to resolve this question, Morgan et al. (1988) quantified daily stability in running economy among runners with similar fitness and IOkm performance backgrounds. Following two 3D-minute treadmill accommodation runs, 16 male subjects completed two IO-minute economy tests at 200 mlmin in the same pair of shoes and at the same time of day within a 4-day period. Each subject also refrained from road racing during the testing period and reduced workout intensity and duration. Expressed as a percentage of the initial test value, intraindividual economy variation was 1.6% (range = 0.4% to 3.4%) and the mean intraclass economy correlation between the two testing periods was 0.97. When compared to intraindividual variation in economy reported in other studies (3% to II %) [Daniels et al. 1984b; Morgan et al. 1987), results from the Morgan et al. (1988) investigation suggest that when treadmill running experience, footwear, time of day, and training activity are controlled, stable economy values can be obtained in trained male runners.

316

Daily stability in running economy has also been quantified by isolating biological and technological variation. Using this approach in a design in which 4 ~02 measures were obtained per subject. Armstrong and Costill (1985) demonstrated that 90% of the total day-to-day economy variation exhibited by IO healthy males running at submaximal speeds (170, 200, and 230 m/min) could be ascribed to biological error. Technological error, or variation associated with instrument fluctuation, represented only 10% of the total within-subject variation in economy. 3.2 Gender Much of the available evidence suggests that the aerobic demand of submaximal running is not significantly different between trained males and females when expressed relative to total body mass (Daniels 1985; Daniels et al. 1977; Davies & Thompson 1979; Hagan et al. 1980; Maughan & Leiper 1983). Some investigators, though, have reported gender differences in running economy. Bransford and Howley (1977), for example, found that trained and untrained male subjects exhibited significantly lower aerobic demands, relative to body mass, when compared to trained and untrained females, respectively. Similarly, Cureton and Sparling (1980) observed that males had a higher ~02 max relative to body mass and were more economical than females when sex-specific variation in body fat was taken into account. Based on these results, they concluded that women would be at a distinct disadvantage in long distance running events, since they would be unable to maintain as fast a running speed. Caloric costs (kcal/kg) of running a mile or kilometre have also been shown to be significantly lower in men (Bhambani & Singh 1985; Howley & Glover 1974). Hypotheses which have been advanced to account for gender variations in running economy include differences in vertical displacement of the body and training experience and intensity (Bransford & Howley 1977; Howley & Glover 1974). Bhambani and Singh (1985) reported no difference in vertical lift per stride or vertical lift per kilo-

317

Factors Affecting Running Econom~

metre travelled among active male and female subjects who demonstrated significant differences in economy. These investigators speculated, however. that the higher stride frequency and greater oxygen debt exhibited by the females may have contributed to the higher overall energy cost of running. The preceding discussion illustrates the need to identify appropriate performance-based criteria to assess gender variation in economy. In an attempt to address this issue, Pate et al. (1985) used recorded times for a 24.2km road race to match male and female adult distance runners and found no difference in the aerobic demand of running expressed relative to body mass. 3.3 Age Cross-sectional research (Astrand 1952; Krahenbuhl et al. 1985; MacDougall et al. 1983) demonstrates that younger children are less economical than older children or adults. A recent review of pertinent studies (Leger & Mercier 1984) indicates that the gross energy cost of running increases 2% per year from 115 to 8 years of age. Longitudinal studies by other investigators (Daniels & Oldridge 1971: Daniels et al. 1978a; Krahenbuhl et al. 1989) have also confirmed that run-trained and non runtrained prepubescent males improve their running economy as they grow older. Because V02 max expressed relative to body mass remains constant throughout childhood in males and decreases among females (Astrand 1952: Bar-Or 1983: Krahenbuhl et al. 1985. 1989). younger children operate at a disadvantage when competing in endurance running events because they utilise a higher percentage of their VO:> max at any given speed. Although little research has been conducted in this area, it has been suggested that differences in leg length. stride length. basal metabolic rate. body surface area to body mass ratio. reduced glycolytic capacity. training. and growth-related factors may partially account for the observed variability in economy between younger and older children and between children and adults (Bar-Or 1983: Daniels & Oldridge 1971: Daniels et al. 1978a: Krahenbuhl

et al. 1989; MacDougall et al. 1983; Rowland & Green 1988). At the other end of the age continuum, little is known regarding the extent and magnitude of economy variation in older adults. Limited research in this area suggests that older individuals are less economical than their younger counterparts during walking (Larish et al. 1987: Sidney & Shephard 1977: Waters et al. 1983). Possible reasons for this reduction in economy include decreased hip flexibility. reduced antagonistic muscle relaxation. increased body fat mass. and increased cardiac and respiratory demands (Larish et al. 1987; Sidney & Shephard 1977). It has also been suggested that older adults. by adopting a shortened stride length. may be economising on force production that the ageing musculoskeletal system must generate and endure (Larish et al. 1987). 3.4 Treadmill and Overground Running Due to the difficulty of obtaining metabolic data in field situations. economy measurements have typically been made indoors on treadmills. Since air and wind resistance are not factors during indoor testing, caution must be used in applying treadmill data to overground conditions (Daniels 1985: Daniels et al. I 986a). Early work by Pugh (1970, 1971) demonstrated that the extra VO:> associated with treadmill running at 4.42 mlsec increased as a function of the square of opposing wind velocity. Additionally, when wind and running velocity were equivalent, as would occur during overground running in calm air. the added oxygen cost increased with the cube of wind velocity . Based on his data. Pugh (1970) estimated that 8% of the total energy cost of middle-distance (SOOOm) track running would be expended overcoming air resistance. Later work by Davies (1980a) essentially supported Pugh's original observations. although differences in the cost of overcoming air resistance during outdoor running were somewhat lower (middle-distance running 4%. marathon running 2%). Insight into the aerobic demands of treadmill and overground running under various wind con-

Factors Affecting Running Economy

ditions has recently been provided by Daniels et al. (1986a). Using 6 elite male runners, it was found that at 268 and 322 mlmin, the cost of overground running in calm air was 7. 1% greater than treadmill running. It was also observed that when tailwind velocity equalled running velocity. overground ~02 was equivalent to treadmill ~02 . These investigators concluded that as wind velocity increased, the deleterious effects of a headwind increasingly outweighed the benefits of a tailwind. While most studies have indicated that overground running is more costly than indoor treadmill running, some investigators have reported no significant differences between the two conditions. McMiken and Daniels (1976) found no difference in economy between track and treadmill running at speeds ranging from 180 and 260 m/min. Within the speed range of 136 and 286 mlmin, Bassett et al. (1985) also observed no significant difference in the mean ~02 values for level treadmill running vs level overground running and for graded treadmill running (5.7%) vs graded overground running. 3.5 Temperature A number of studies have documented the effect of increased core temperature (the 'QIO' effect) on ~02. Saltin and Stenberg (1964) reported a 5% rise in ~02 during 3 hours of constant-load exercise under normal conditions. MacDougall et al. (1974) also observed that ~02 was significantly higher in subjects who exercised at 70% ~02 max under hyperthermic conditions compared to normal or hypothermic conditions. The authors suggested a variety of reasons for this rise in ~02 , including an increased energy requirement for peripheral circulation. increased sweat gland activity. hyperventilation and a decreased efficiency of energy metabolism. Evidence regarding the genesis of the oxygen debt provides additional evidence for the link between temperature and oxygen consumption. Brooks and his colleagues (Brooks et al. 1970, 1971; Gaesser & Brooks 1984) have shown that when rat skeletal muscle and liver mitochondria are incubated at high physiological temperatures, respiratory control is disrupted. Conse-

318

quently, more oxygen is required to synthesise a given amount of ATP. Although the majority of research has demonstrated that ~02 rises with an increase in core temperature, Rowell et al. (1969) reported no significant alteration in submaximal or maximal oxygen consumption during hyperthermic exercise. These authors suggested that an increased mechanical efficiency of muscle may have explained the lack of change in ~02. Data from other studies (Dill 1965; Maron et al. 1976), which have shown a reduction in ~02 during the latter portion of a prolonged run, also support the possibility of increased muscular efficiency with elevated muscle temperatures. 3.6 Fatigue The association between running economy and fatigue is not well understood. Early research in this area focused on determining the metabolic costs of various racing strategies employed during exhaustive runs lasting It!ss than 5 minutes (Adams & Bernauer 1968; Ariyoshi et al. 1979; Robinson et al. 1958). Results from these studies produced no consensus on this issue, but instead. led to the adoption of divergent strategies (e.g. aconservative start, a fast start. and a steady pace) to achieve minimal oxygen costs during short term. maximal runs. More recent data obtained on elite and trained endurance male runners performing longer runs have produced conflicting results, with I study reporting higher aerobic demands following a competitive distance race (Cavanagh et al. 1985) and the other demonstrating no change in economy I day after a hard training workout (Martin et al. 1987). In the Martin et al. investigation. the aerobic demand of running remained constant despite a decrease in the respiratory exchange ratio and an increase in free fatty acid concentration. In explaining these results, it was speculated that subtle modifications in the gait pattern nullified the expected rise in ~02 associated with increased fat metabolism. Methodological constraints inherent in this study, however, limited the extent to which

Factors Affecting Running Econom y

substantive conclusions could be drawn regarding the perturbability of economy. In an attempt to resolve this question, Morgan (1988) replicated the Martin et al. (1987) study using 16 male distance runners and an expanded experimental design. Results from the more investigation study revealed no changes in submaximal V02 or heart rate 1,2 or 4 days after an exhaustive 30-minute level treadmill run performed at 85% of each subject's VV02 max (~ 89% V02 max). While respiratory exchange ratio values wert significantly lower I and 2 days after the exhaustive run (~ = 0.02), the estimated percentage of kilocalories derived from fat was increased by only 6.6%. Additionally, biomechanical analyses revealed little variation in 21 temporaL kinematic, and kinetic gait descriptors previously linked to variation in economy (Williams & Cavanagh 1987). From a practical standpoint, these findings suggest that an intense 30-minute training run or a competitive 10km race would not raise the aerobic demand of running by increasing dependence on fat metabolism or disrupting the gait pattern in subsequent submaximal runs over the short term. Viewed from a theoretical perspective, these results demonstrate the imperturbability of the metabolic and biomechanical profiles of trained runners following a prolonged (30-minute) maximal run. 3.7 Training Little consensus exists regarding the effects of training on running economy. This is due partly to the relative lack of longitudinal studies in this area and to limitations in experimental design such as: (a) the use of small sample sizes: (b) the lack of multiple economy measures to account for normal intraindividual variation: and (c) failure to control factors which influence or may potentially influence economy (e.g. fatigue leveL state of training, circadian variation, training accommodation, footwear mass and design). The absence of research examining the interactive effects of various types and thresholds of training with subject fitness levels has also hampered a better understanding of this topic (Daniels 1985).

319

Despite these limitations, various training modes have been associated with better economy. Work by Daniels and associates (Daniels & Oldridge 1971; Daniels et al. 1978a) has shown that running economy was improved in 20 pre- and postadolescent males who engaged in middle and long distance running programmes over a 2- to 5-year period. These researchers concluded that growth-related factors and training were likely causes for the enhancement in running economy. Patton and Vogel (1977) reported that a 6-month conditioning programme consisting of long distance running at moderate intensities (2- and 4-mile runs at 8- to 9minute per mile paces) significantly improved economy in 60 untrained and trained military personnel. Since VOl max was increased only in the untrained group, it was suggested that a combination of training, improved mechanical efficiency, and treadmill habituation ma y have been associated with the observed reduction in submaximal V02. Longitudinal studies have also demonstrated that interval training or a combination of interval and long distance training improves running economy (Conley et al. 1981 a, 1984). Since both of these investigations were case studies involving a single elite male runner, the representativeness of these data is unknown. As mentioned earlier. Svendenhag and Sjodin (1985) observed significant improvements in running economy over a mean period of 22 months in 16 elite male distance runners who performed long distance, uphill, and interval running. Likewise, supplementing regular training with I weekly 20-minute run performed at the velocity eliciting a 4 mmol/L blood lactate concentration has been shown to improve the economy of middle and long distance male runners (Sjodin et al. 1982). In discussing their findings, the authors suggested that alterations in running style and intracellular oxidative capacity may have been responsible for the lower oxygen demand. Other investigations have shown no change in economy with training. Wilcox and Bulbulian (1984), for instance, reported no significant difference in economy over an 8-week cross-country training session in 7 collegiate runners who aver-

320

Factors Affecting Running Economy

aged 60 to 70 miles of running and 2 high-intensity workouts per week. Likewise, Daniels et al. (1978b) found no change in running economy in 15 welltrained recreational runners after an 8-week period of controlled long distance and interval training in which total running mileage was increased from 20 to 30 km/week to 50 to 70 km/week. Attempts to improve running economy in children using various combinations of run training and running instruction conducted over an II-week period have also proved unsuccessful (Petray & Krahenbuhl 1985). While acknowledging the value of longitudinal research in quantifying the influence of training on running economy, cross-sectional differences in economy among subject groups varying in training status have been demonstrated in many studies (Bransford & Howley 1977; Daniels 1985; Dolgener 1982; Krahenbuhl & Pangrazi 1983; Mayers & Gutin 1979; Pollock et al. 1980). The preponderance of these data indicate that trained subjects are more economical than their untrained or less trained counterparts. It has been speculated that untrained subjects may be less economical due to a lack of training, a reduced predisposition for success in distance running, a diminished efficiency of mechanical movement or a decreased efficiency of oxidative energy supply (Bransford & Howley 1977). Unfortunately, few studies have been conducted to confirm or refute these hypotheses. The relationship between training and the aerobic demand of running has further been examined by comparing the economy of athletes engaged in different running events. Evidence from a number of sources (Daniels 1985; Daniels et al. I 986b; Pollock 1977; Pollock et al. 1980) suggests that long distance and elite marathon runners are more economical than middle distance runners. This difference in economy may be due to a lower vertical displacement of the body or to other metabolic and neuromuscular factors linked to slow distance training (Svendenhag & Sjodin 1984). An alternative, but equally plausible, hypothesis is that long distance runners gravitate naturally towards endurance events because the attribute of economy does not have the opportunity to manifest itself

during short term races. In other words, differences in the aerobic demand of running among various types of runners might reflect differences in the expression of a genetic predisposition towards worse or better economy. While cross-sectional research has demonstrated inter group variability in economy among athletes, some investigators have reported no significant differences in economy between long distance and shorter distance runners (Boileau et al. 1982; Dolgener 1982; Svendenhag & SjOdin 1984). These latter findings advance the notion that a particular type of run training exerts a negligible effect on economy. Support for this alternative view can be found in studies showing substantial variation in economy among long distance runners who compete in similar running events (Conley & Krahenbuhl 1980; Costill et al. 1973; Daniels 1974; Daniels et al. 1977; Daniels et al. 1984a; Farrell et al. 1979; Morgan et al. 1989).

4. Biomechanical Considerations for Economical Running While many investigations have considered the relationships between various physiological attributes and running economy, far less research is available providing insight into how descriptors of running mechanics affect economy. It is frequently suggested that biomechanical factors may account for a substantial portion of interindividual variations in running economy. For example, in a comprehensive analysis of 31 recreational runners, Williams and Cavanagh (1987) reported that selected biomechanical factors were associated with lower aerobic demands of running. Despite their conclusion, however, the base of support for this notion is rather limited because of the paucity of research that has been conducted on this topic. 4.1 Relationship Between Body Structure and Economy

4.1.1 Body Mass Because running economy is usually normalised to body mass, it is generally assumed that economy is independent of body mass and thereby accounts

Factors Affecting Running Economy

for none of the interindividual variability in economy. Davies (1980b) noted that previous work (Davies & Thompson 1979) showed lightweight men to be no more or less economical than their heavier counterparts. Skinner et al. (1973) also reported nearly identical economy values during treadmill walking for lean . obese. and weighted lean subjects. Other researchers. though. have reported that body mass can influence economy even when economy is expressed relative to body mass. In a study of 11- to 13-year-old children running on a treadmill under unloaded and loaded conditions (either 5 or 10% of bodyweight carried on the trunk). Davies (l980b) found a significant difference in the slopes of the speed-economy relationship for the unloaded and 5% loaded conditions such that lower aerobic demands were observed for the loaded condition at higher running speeds (e.g. 14 to 16 km/h). They noted that the addition of weights had little effect on economy at low running speeds (e.g. 9 km/h). and that increasing the load from 5% to 10% of bodyweight had no additional effect on economy. Taylor and colleagues (Taylor et al. 1982: Taylor 1986) also contend that body mass is a significant determinant of economy. They concluded that the aerobic demand for humans and most terrestrial vertebrates to travel a given distance is dependent upon the animal's weight. but independent of the speed. gait (i.e. walk. run. trot. gallop). and mode of locomotion (i.e. bipedal. quadrupedaL etc). Using their allometric equation relating economy to body mass (Mb): V02

= 0.533

Mb -0.3\6.

Vg

+ 0.300 Mb- 0303

where V02 is expressed in mljkg/sec. Mb is body mass in kg. and Vg is velocity in m/sec. it can be shown that predicted aerobic demands for running at 4.0 m/sec would range from 42.7 mljkg/min for a 50 kg individual to 36.8 ml/kg/min for an 80kg individual. From a study of 14 elite female distance runners. Williams et al. (1987) found support for a modest inverse relationship (r = -0.52) between body mass or weight and economy and a slightly

321

stronger relationship between maximal thigh circumference and economy (r = -0.58). indicating that heavier than average runners exhibited better economy than did lighter runners. They further noted that the relationship between bodyweight and economy for the female runners was consistent with that observed for a group of male runners (r = -0.39) [Williams & Cavanagh 1986]. For the elite male runners. they noted that anthropometric variables. rather than those describing running mechanics. correlated most highly with economy. In particular. those reflecting linear dimensions of the body (e.g. leg length. pelvic width and foot length) showed the strongest links to V02, with correlations ranging from -0.55 to -0.68.

4.1.2 Body and Segment A1ass Dislribution Considering mass distribution within the body. Cavanagh and Kram (1985) suggested that a potential source for individual differences in economy is variation in the distribution of mass among limb segments. Similarly. Myers and Steudel (1985) noted that functional morphologists have long regarded limb morphology as a significant influence on the energetic cost of locomotion in terrestrial animals. Assuming that all other factors are unchanged (e.g. speed, body mass, running style). these authors hypothesised that a runner with a proportionately smaller amount of body mass concentrated in the extremities, particularly the legs, would perform less work in moving the body segments during runing than one with a greater proportion of body mass concentrated in the extremities. Indirect support for this hypothesis comes from numerous loading studies (e.g. Catlin & Dressendorfer 1979: Cureton et al. 1978; Hettinger & Muller 1952: Inman et al. 1981: Jones et al. 1984: Keren et al. 1981: Martin 1985: Myers & Steudel 1985). In general. results of these studies indicate that the aerobic demand of carrying a given load on the distal aspect of the lower extremity is approximately an order of magnitude higher than that for carrying the same load on the trunk. For example. the increase in aerobic demand for load carried on the trunk is approximately 0.1 % per 100g of load. while that for loading of the foot is slightly below

Factors Affecting Running Economy

1.0% per lOOg of load. As expected, the increase in aerobic demand is less as the additional load is p0sitioned more proximally on the extremities (Martin 1985; Myers & Steudel 1985). Despite this indirect support. results of Taylor et al. (1974) contradict this notion. They reported nearly identical aerobic demands across a considerable range of running speeds in 3 species of animals (cheetah. gazelle and goat) with similar body masses and limb lengths. but large differences in both limb mass and mass distribution of the limbs. While they found the hypothetical relationship between body mass distribution and economy to be intuitively appealing. they concluded that the effect of differences in mass distribution on economy was too small to measure. 4.2 Relationship Between Running Kinematics and Economy

4.2.1 Running Speed Perhaps the most basic descriptor of running mechanics is the speed or velocity of running. As noted by Daniels (1985), research since 1950 generally supports the concept of a linear relationship between running speed and economy. For example, Margaria and colleagues (Margaria 1963; Margaria et al. 1963a) noted that the energy cost of running expressed relative to distance travelled (i.e. kcal/kgjkm) is essentially constant. Daniels (1985) further observed that the 'concept of a linear relationship between velocity and V02 seems to hold up during submaximal running, where energy demands are met aerobically and where the range of running speeds is rather limited' (p. 333). Nevertheless, Daniels et al. (1977) showed that the slope of the relationship between economy and running speed can vary for a group of subjects depending upon the speeds chosen for analysis. At slow running speeds, slopes tended to be somewhat lower or flatter than those calculated when only higher speeds were examined (Daniels 1985). 4.2.2 Stride Length When considering the effect of selected descriptors of running mechanics on economy under controlled running speeds, stride length is one of the

322

few variables that has been shown by direct experimental evidence to affect economy. Results from a number of studies (Cavanagh & Williams 1982; Hogberg 1952; Kaneko et al. 1987; Knuttgen 1961; Powers et al. 1982) have indicated that the aerobic demand of running at a given speed tends to increase curvilinearly as stride length is either lengthened or shortened from that which is freelychosen by the runner. This basic curvilinear relationship between stride length and economy has also been shown for racewalking (Morgan & Martin 1986). Early work involving stride length manipulation (Hogberg 1952) indicated that a well-trained subject at 14 and 16 km/h was most economical at the self-selected stride length versus running at stride lengths shorter and longer than the self-selected value. A comparison of the aerobic demands associated with these various stride length conditions also revealed that V02 while running with a stride length 13.3% longer than the optimal stride length was 11.9% higher than that for the optimal stride length condition. In contrast, a nearly equal decrease in stride length from the optimal condition (i.e. 11.9% shorter than the optimal stride length) resulted in only a 6.0% increase in V02. In a more thorough evaluation of the relationship between economy and stride length, Cavanagh and Williams (1982) evaluated 7 stride length conditions at a single running speed (13.8 km/h) for lOwell-trained recreational runners. Their results indicated that a curvilinear relationship existed between stride length and V02 as stride length was varied from the self-selected value by as much as 20% of leg length. In accordance with Hogberg's (1952) results, Cavanagh and Williams found that V02 was lowest at stride lengths close to the selfselected condition. In contrast with the results of Hogberg, however, group trends demonstrated that increases in V02 were nearly identical for similar increases and decreases in stride length from the self-selected value. Responses of individual subjects indicated that some subjects showed greater increases in V02 as stride length was lengthened from the self-selected condition. while others dis-

Factors Affecting Running Economy

played greater increases with decreasing stride lengths. Based on these results. Cavanagh and Williams (1982) concluded that there is little need for a coach to dictate a particular stride length profile in most athletes since they tend to display nearly optimal stride lengths. They suggested that this phenomenon might be due to 2 mechanisms. The first states that runners may gravitate naturally toward an optimal stride length/stride rate combination over time through an iterative process based on perceived exertion. A second possibility is that runners may adapt physiologically through repeated training at a particular combination of stride length and stride frequency for a given running speed. Unfortunately. neither possibility has been evaluated by subsequent research. With respect to this latter point. of particular interest is whether lasting changes in the running style of an uneconomical runner can be made such that a more economical pattern is ultimately achieved. Kaneko et al. (1987) provided further insight into the link between stride length and economy by quantifying the mechanical power output for several stride frequency /stride length conditions. Their results. which were based on 4 subjects. demonstrated the expected curvilinear response between economy and stride frequency and a similar response. although less dramatic in nature. between economy and mechanical power. They speculated that the economy response may be associated with muscle fibre recruitment. At lower frequencies. the muscles need to develop relatively high external power to achieve longer stride lengths, while at high stride frequencies, the mechanical power associated with moving the limbs increases. They indicated that these extreme conditions may require a greater reliance on less economical fast twitch fibres than more intermediate stride length/ stride frequency combinations.

4.2.3 Other Kinematic "ariables While stride length and stride frequency have been associated with economy from direct experimental manipulation, other discrete kinematic factors have demonstrated a relationship with econ-

323

omy. Williams and colleagues have reported results from 3 similar and reasonably comprehensive studies of distance runners, including analyses of 13 elite male distance runners (Williams & Cavanagh 1986), 14 elite female distance runners (Williams et al. 1987), and 31 recreational distance runners (Williams & Cavanagh 1987). In general, correlations between economy and biomechanical descriptors were moderate to low (Williams & Cavanagh 1986). Table I serves to summarise the relationships reported between economy and selected kinematic factors. Although a number of significant trends were observed in these studies, it is interesting to note that there are few consistent trends among the studies. Not only are different variables significantly associated with economy in the 3 studies, but also for 2 variables (maximum thigh extension and maximum knee extension at or near toe oft), opposing results were observed. Williams et al. (1987) could provide no explanation for these contradictory trends, but suggested that these data should be considered preliminary and noted that the identification of definitive trends awaits the development of much larger data bases. 4.3 Relationship Between Running Kinetics and Economy

4.3.1 Ground Reaction Forces Like the kinematic descriptors just discussed, several descriptors of the ground reaction force have been associated with running economy. Williams and Cavanagh (1986) found that both support time and the peak medial force correlated positively with economy (r = 0.49 and 0.50. respectively). In addition, Williams and Cavanagh (1987) reported that more economical runners had significantly lower first peaks in the vertical component of the ground reaction force and tended to have smaller anteroposterior and vertical peak forces and more of a rear foot striking pattern. Based on these results and those noted previously for kinematic descriptors. they suggested that differences in approach kinematics (i .e. just preceding foot contact) may affect muscular demands both before and during

324

Factors Affecting Running Economy

TMIe I. Kinematic fectonl shown to have a stIItisticaI association wi\tI running economy from research by Wiliams Ind colleagues (Wllliami & Cavanagh 1986. 1987; Williams et II. 1987) Kinematic vlrlable

Correlation 13 elite males·

Change in AP velocity in training shoes Incr.... in horizontal velocity in propulsion Vertical OICiHation of eM Trunk lean Wrilt excursion MIX. thigh extension Max. thigh extenllon velocity MIX. knee extenllon at 0( near toe off Max. knee flexion during support Max. knee flexion velocity Minimum linelr knee velocity Shlnk Ingle It foot strike Max. dorsiflexion Ingle Max. dorsiftexlon velocity MIX. plantar flexion angle Max. plantar flexion velocity Horizontal heel velocity at foot contact a b

14 elite females·

31 recreational malesb

-0.62 0.53 Trend Trend 0.53 0.58

-0.41 -0.51 -0.55 Trend 0.59

-0.60 -0.59 0.46 -0.64

Correlational lnaJyseS in which positive correlations reflect a tendency fO( higher values fO( the reported variable to be aSSOCiated with higher Ii'Ch vllues. Based on comparisons of good, intermediate, and poor economy subgroups. • denotes a statistically significant difference between groups, while trend indicates a conSistent trend but no statistically significant difference between groups.

support, thereby affecting economy. They also suggested that the need to provide cushioning with foot contact may also be associated with economy. Specifically, those individuals who strike the ground more forward on the foot may have to rely on the musculature to assist with cushioning to a greater extent than rear foot strikers who can perhaps rely on footwear and skeletal structures to take more of the load. 4.3.2 Mechanical Power Because economy is considered to be a global indicator of the physiological demand of running, it is logical to assume that a global mechanical descriptor of the output of the neuromuscular system would be more closely associated with economy than individual descriptors of discrete instants or events of the running pattern. Many attempts to explain the aerobic demand oflocomotion have assumed that the major energetic cost occurs when

muscles shorten and perform mechanical work (Taylor et aI. 1980). Beckett and Chang (1968) went further in saying that 'in certain well-learned tasks such as walking and running, it seems reasonable to expect that the movement of the body components would be made in such a way as to minimize the amount of mechanical work that is done' (p. 147). Even though mechanical work or power is often assumed to be closely associated with economy, surprisingly few studies have directly examined the ability of variations in mechanical power output to explain interindividual variations in economy. Of the research that has been conducted, little definitive support is evident for this view. Numerous investigations (e.g. Burdett et al. 1983; Cavagoa et al. 1964; Fukunaga ct al. 1977; Fukunaga & Matsuo 1980; Heglund et al. 1982; Kaneko et al. 1981; Luhtanen & Komi 1978; Shorten et al. 1981; Taylor 1986) have shown that as the speed of locomotion increases, the mechan-

Factors Affecting Running Economy

ical work done per step and the average mechanical power increases. For example, Shorten et al. (1981) found running economy to be closely associated with several expressions of average mechanical power (r > 0.86) in their analysis of 4 runners at 6 speeds. Similarly, Burdett et al. (1983) reported high correlations between mechanical power and walking economy (r > 0.79) for 6 subjects studied under 5 walking speeds. Because these studies employed multiple speeds, and since both metabolic and mechanical power are speed-dependent, they do not provide a rigorous test of whether mechanical power variations contribute to the large variations in economy often observed between individuals. Taylor (1986) suggested that mechanical work or power does not satisfactorily explain economy variations. He noted that the ' mechanical cost of locomotion' (p. 409) expressed in J/kg/m can be predicted from the speed of locomotion, but is independent of the body mass of the individual (Heglund et al. 1979; Taylor 1986). ]n contrast, metabolic energy cost of locomotion, as expressed in the equation above. is independent of speed but dependent on body mass (Taylor et al. 1970; Taylor 1977). From this, Taylor (1986) concluded that ' it seems clear that one must look beyond a 'mechanical work' explanation for the general relationship between energy cost of locomotion and body size' (p. 410). As an alternative explanation, Taylor (1985) suggests 'that it is the time course of force development during locomotion , rather than the mechanical work that the muscles perform, that determines the metabolic cost of locomotion' (p. 253). Of particular importance to this hypothesis are the elastic characteristics of the muscle-tendon units. In elaborating upon this concept. he indicated that the speeds and stride frequencies that animals. and presumably humans as well. select within gaits (e.g. walk, run , trot. gallop) are those where storage and recovery of elastic energy are maximised. Providing modest support for the notion that mechanical power may be linked to economy, Williams and Cavanagh (1987) reported that one of 3 significant predictors of econom y for a single run-

325

ning speed in a multiple linear regression model was net positive power. They also noted that the least economical runners showed significantly less mechanical energy transfer between the legs and the trunk than their more economical counterparts. Furthermore, while the more economical runners tended to display lower net positive power, lower total mechanical power, and greater between-segment energy transfer than the less economical runners, these differences were not statistically significant.

5. Future Research Directions While a growing body of knowledge is emerging with respect to the study of economy, many unanswered questions remain. Some issues which deserve further attention include the following: I. What is the daily stability of running economy over the long term? While a few studies have attempted to assess within-subject variation in V02 over the short term (3 to 4 sessions), the degree of long term stability inherent in this variable is unknown. Knowledge of the true magnitude of normal running economy variation within trained and untrained subjects is crucial in separating experimental treatment effects from biological instability. 2. Is there a gender difference in economy, and if so, why? Some studies have shown that when expressed relative to total body mass, males are more economical than females . Other studies, though , have reported no gender differences in this measure. In resolving this question, greater attention should be devoted to recruiting large sample sizes of men and women who display similar actitivy, training, and performance backgrounds. Additionally, econom y measures should be complemented by biomechanical and body structure analyses in an attempt to account for observed differences in V0 2. 3. Wh y are younger children less economical than older children and adults and why are older adults less economical than their younger counterparts" Cross-disciplinary studies of children incorporating metabolic. biomechanical, structural,

Factors Affecting Running Economy

strength, flexibility, motor control, and physical activity measures are needed to better understand the time course of economy improvements from early childhood to postadolescence and the mechanisms which underlie these changes. At the other end of the age spectrum, descriptive and longitudinal studies involving older adults are almost non-existent. From a practical standpoint, the implementation of specific training programmes designed to reduce the energy demands of locomotion could enhance the quality of life for older adults. From a more theoretical perspective, the study of economy in the aged could yield insight into the factors which dictate the selection of gait patterns generated by an ageing musculoskeletal system. 4. How does economy vary across a wide spectrum of speeds? Little is known regarding the energy demands of running over a wide range of speeds. Expressed per unit of distance travelled, research in this area might allow the determination of particular speeds of running most economical for trained and untrained subjects. 5. What are the effects of intense, long-duration runs and overtraining on running economy? While previous work has shown that a 30-minute maximal run did not perturb ,,"02 or running style over a 4-day period (Morgan 1988), it is possible that intense, longer duration runs (e.g. half-marathon or marathon distances) might produce significant changes in economy and related aspects of the gait pattern. Indirect support for this view can be found in work by Buckalew et al. (1985), who observed that most alterations in gait mechanics in a group of elite marathoners occurred between the 20- and 24-mile marks. Related to this question is the possibility that overtraining, defined as a serious and chronic maladaptation to training (Wells & Pate 1988), may be associated with a higher aerobic demand of running. Few studies have examined the effects of multiple bouts of hard training or competition on running economy. Longitudinal and cross-disciplinary studies of activities in which overtraining has been shown to be a common problem (e.g. long distance running and swimming) are needed to identify interrelationships among metabolic, bio-

326

mechanical, and psychological markers of training and performance. 6. What combination of type, intensity, duration, and frequency of training best improves economy in trained and untrained subjects? Future efforts in this area should include comprehensive biomechanical and structural analyses in order to provide a data-based rationale for potential metabolic differences observed among different subject groups exposed to various training modalities. 7. Can uneconomical running styles be altered by biomechanical training? The degree to which running mechanics can be modified in uneconomical runners has received little attention. The development and implementation of specific techniques aimed at optimising the mechanical profiles of subjects displaying uneconomical running styles might reduce the aerobic demand of locomotion in this cohort and promote a better understanding of the association between metabolic and biomechanical correlates of economy and the extent to which gait pattern optimisation persists in the absence of training.

Rt/trtIKts Adams W. Bernauer E. The effect of selected pace variations on the oxygen requirement of running a 4:37 mile. Research Quarterly 39: 837·846. 1968 Ahlborg G. Felig P. Hagenfeldt L. Hendler R. Wahren J. Substrate turnover during prolonged exercise in man: splanchnic and leg metabolism of glucose. free fatt y acids. and amino acids. Journal of Clinical Investigation 53: 1080-1090. 1974 Allen W. Seals D, Hurley B. Ehsani A. Hagberg J. Lactate threshold and distance running performance in young and older endurance athletes. Journal of Applied Physiology 58: 1281·1284. 1985 Apor p. Fekete G, Kostre W. Data on aerobic efficiency of running. Acta Physiologica Academiae Scientiarum Hungaricae 3: 275-280. 19110 Ariyosh i M. Yamaji K. Shephard R. Influence of running pace upon performance: effects upon treadmill endurance time and oxygen cost. European Journal of Applied Physiology 41: 8391 . 1979 Armstrong L. Costill D. Variability of respiration and metabolism: responses to submaximal cycling and running. Research Quarterl y for Exercise and Sport 56: 93-96. 1985 Astrand P-O. Experimental studies of physical working capacit y in relation to sex and age. Ejnar M unksgaard. Copenhagen. 1952 Astrand p-O. Rodahl K. Textbook of work physi ology. McGrawHill . New York . 1986

Factors Affecting Running Economy

Bar-Or O. Pedlatnc spons med,c,ne. Springer-Verlag. New York. 1983 Bassett D. G,ese M. !\agle F. Ward A. Raab D. Aerobic requirements of overground versus treadmill running. Medicine and Science In Sports and ExerCIse 17: 477-481. 1985 Beckett R. Chang " ..-\n evaluation of the kinematics of gait by minimum ent'rg~. J()urnal ()f Biomechanics I: 147-15'>. 1'>6H Bergstrom J. Hermansen L. Saltin B. Diet. muscle glycogen and phYSIcal p.cta Ph~Slologica Scandinavica 71: 140150. 1967 Bhambani Y. SIngh M. Metabolic and CInematographic analysis of walking and running In men and women. Med,Cine and Science In Sports and ExerCISe 17: 131-13 7. 1985 Boileau R. Mavhew J. Relller W. Lussie r L. Ph~"ologlCal characteristIC s "I' elite middle and long dIStance runners. Ca nadian Journal of >.pphed Sports Sciences 7: 16 7 -172. 1982 Brandon L. Boileau R. The contribution of selected variables to middle and long distance running performance . Journal of Sports MedIcine and Physical Fitness 27: 157-164. 1987 Bransford D. Howley E. Oxygen cost of running in traIned and untraIned men and women. Medicine and SCIence In Sports 9: 41-44.1977 Brooks G. Fahey T. Exercise physiology. John Wiley & Sons. Ne w York. 1984 Brooks G. Hittleman K. Faulkner J. Beyer R. Temp
327

CoslIII D. FInk W. Pollock M. Muscle fiber and enzyme activitIes of elite distance runners. Medicine and Science in Sports 8: 96-100. 1976b Costill D. Thomason H. Roberts E. Fractional utilization of the aerobic capacity during distance running. Medicine and Science in Sports 5: 248-252. 1973 Costill D. Winruw E. A companson of two middle-aged ultramarathon runners. Research Quarterly 41: 135-139. 1970 Cureton K. Sparling P. Distance running performance and metabolic responses to runlllng in men and women with excess weight exp<,rimental alterations in excess weight on aerobic capacity and distance running performance. Medicine and Scie nce in Sports 10: 194-199. 1978 Daniels J. Ph YSIOlogical characterIStics o f cha mpIOn male athletes. Research Quarterly 45: 342-348. 1974 Daniels J. A phYSIologist's VICW of running economy. Medicine and Science in Sports and Exercise 17: 332-338. 1985 Daniels J. Krahenbuhl G. Fost er C. Gilbert J. Dan iels S. AerobIC responses of female distance runners to submaximal and maximal exercise. Annals of the New York Academv of Sciences . 301: 72ti-7H 1977 Daniels J. Oldridge N. Changes in ox ygen consumption of young boys during growth and running training. Medicine and Science in Sports 3: 161-165. 1971 Daniels J. Old ridge N. Nagle F. White B. Differences and changes in ~02 among yo ung runners 10 to 18 years of age. Medicine and Science in Sports 10: 200-203. 1978a Daniels J. S,'ardlna N. Foley P. ~02 submax durmg five modes of exercise . In Bachl et al. (Eds) Proceedings of the World Congress on Spons Medicine. pp 604-615. Urban & Schwartzenberg. Vienna. 1984a Dallleis J. Scardina N. Haves J. Folev P. VariatIons in VOo submax during treadmill r';nning ..-\bstract. Medicine and sCience in Sports and E.vercise 16: 108. 1984b Damels J. Scardina N. Hayes J. Foley P. Elite and sub-elite female middle- and long-distance runners. In Landers (Ed.) The 1984 Olympic Scientific Congress Proceedings. Vol. 3. Sport and elite performers. pp. 57-72. Human Kinetics. Champaign. 1986b Dallleis J. Yarbrough R. Foster C. Changes in ~02 max and runIlIng performance with traming. European Journal of Applied Physiology 39: 249-254. 1978b Daniels N. Daniels J. Baldwin C. Bradley P. The effect of wind on the aerobic demand of running. Abstract. Presented at the 1986 National Meeting of the American College of Sports Medicine, Indianapolis. 1986a Davies C. Effects of wind assistance and resistance on the forward motIOn of a runner. Journal of Applied Physiology 48: 702709. 1980a Davies C. Metabolic cost of exerCIse and phySIcal performance In children WIth some observations on external loading. European Journal of>.pphed Phy sio log~ 45: 95-102. 1980b DaVIes C. Th o mpson M. AerobIC performance of female marathon and male ultramarathon athletes. European Journal of .>.pphed Physiolog~ 41: 233-245 . 1979 Dill D. Marathoner DeMar: physiological stud,es. Journal of the National Cancer Inslltute 35: 185-191. 1965 DiPrampero P. Davies C. Cerretelli P. Margaria R . An analysis of 02 debt contracted in submaximal exerCIse. Journal of .-\pplied Physiology 29: 547-551. 1970 Dolgener F. Oxygen cost of walking and running in untrained . sprint trained. and endurance trained females. Journal of Sports MedIcine and Physical Fitness 22: 60-65. 1982 Farrell P. Wilmore J. Coyle E. Billings J. Costill D. Plasma lactate accumulation and distance running performance. MediCine and Sc,ence In Sports II : 338-344. 1979

Factors Affecting Running Economy

Felig P. Wahren J. Amino acid metabolism in exercising man. Journal of Clinical Investigation 50: 2703. 1971 Fukunaga T. Matsuo A. Effect of running velocity on external mechanical power output. Ergonomics 23: 123-136. 1980 Fukunaga T. Matsuo A. Yuase K. Fukimatsu H. Asahina K. Mechanical power output in running. In Asmussen & Jorgensen (Eds) Biomechanics VI-B. pp. 17-22. University Park Press. Baltimore. 1977 Furusawa K. Hill A. long C. Lupton H. Muscular exercise and oxygen requirement. Proceedings of the Royal Society of london (Biology) 97: 167-176. 1924 Gaesser G, Brooks G. Metabolic bases of excess postoCxercise oxygen con~umption: a review. Medicine and Science in Sports and Exercise 4 16: 29-43. 1984 Green H. Hughson R. Orr G. Ranney D. Anaerobic threshold. blood lactate. and muscle metabolites in progressive exercise. Journal of Applied Physiology 54: 1032-1038. 1983 Hagan R. Smith M. Gettman L. Marathon performance in relation to maximal aerobic power and training indices. Medicine and Science in Sports and Exercise 13: 185-189. 1981 Hagan R. Strathman L. Gettman L. Oxygen uptake and energy expenditure during horizontal treadmill running. Journal of Applied Physiology 49: 571-575. 1980 Hagberg J. Coyle E. Carroll J. Miller J. Martin W. et al. Exercise hyperventilation in patients with McArdle's disease. Journal of Applied Physiology 52: 991-994, 1982 Hagberg J. Nagle F. Carlson J. Transient 02 uptake response at the onset of exercise. Journal of Applied Physiology 44: 90-92. 1978 Heck H. Mader A. Hess G. Mucke S, Muller R. et al. Justification of the 4-mmol/1 lactate threshold. International Journal of Sports !'.Iedicine 6: 117-130. 1985 Heglund NC, Cavagna GA. Fedak MA. Taylor CR. Muscle efficiency during locomotion: how does it vary with body size and speed? Federation Proceedings 38: 1443. 1979 He81und NC, Fedak MA. Taylor CR. Cavagna GA. Energetics and mechanics of terrestrial locomotion. IV. Total mechanical energy changes as a function of speed and body size in birds and mammals. Journal of Experimental Biology 97: 57-66, 1982 Henritze J. Weltman A, Schurrer R. Barlow K. Effects of training at and above the lactate threshold on the lactate threshold and maximal oxygen uptake. European Journal of Applied Physiology 54: 84-88. 1985 Henry F. Aerobic oxygen consumption and alactic debt in muscular work. Journal of Applied Physiology 3: 427-438, 1951 Hettinger T, Muller EA. Der Einflub des Schuhgewichtes auf den Energieumsatz beim Gehen und Lastentragen (The influence of shoe weight on the energy cost of walking and load carrying). Arbeitsphysiologie 14: 437-441. 1952 Hill A, Lupton H. The oxygen consumption during running. Journal of Physiology (london) 56: xxxii-xxxiii, 1922 Hill A, Lupton H. Muscular exercise, lactic acid, and the supply and utilization of oxygen. Quarterly Journal of Medicine 16: 135-171. 1923 HOgberg P. How do stride length and stride frequency influence the energy output during running. Arbeitsphysiologie 14: 437441. 1952 Howley E. Glover M. The caloric costs of running and walking one mile for men and women. Medicine and Science in Sports 6: 235-237, 1974 Hughes E, Turner S. Brooks G. Effects of glycogen depletion and pedaling speed on 'anaerobic threshold'. Journal of Applied Physiology 52: 1598-1607. 1982 Inman VT, Ralston HJ, Todd B. Human walking, pp. 62-77, Williams and Wilkins, Baltimore, 1981 Ivy J, Costill D. Maxwell B. Skeletal muscle determinants of maximum aerobic power in man. European Journal of Applied Physiology 44: 1-8. 1980a Ivy J, Withers R. Van Handel P. Elger D, Costill D. Muscle res-

328

piratory capacity and fiber type as determinants of the lactate threshold. Journal of Applied Physiology 48: 523-527. 1980b Jacobs I. Blood lactate: implications for traiRlng and sports performance. Sports Medicine 3: 10-25. 1986 Jones BM. Toner M, Daniels W, Knapik J. The energ) cost and heart rate response of trained and untrained subjects walking and running in shoes and boots. Ergonomics 27: 895-902. 1984 Kaneko M. Ito A. Fuchimoto T, Toyooka J. Mechanical work and efficiency of young distance runners during level running. In Morecki et al. (Eds) Biomechanics VII-B. pp. 234-240. U niversity Park Press, Baltimore, 1981 Kaneko M. Matsumoto M. Ito A. Fuchimoto T. Optimum step frequency in constant speed running. In Jonsson (Ed.) Biomechanics X-B. pp. 803-807. Human Kinetics. Champaign. 1987 Keren G, Epstein Y. Magazanik A, Sohar E. The energy cost of walking and running with and without a backpack. European Journal of Applied Physiology 46: 317-324. 1981 Knuttgen HG. Oxygen uptake and pulse rate while running with undetermined and determined stride lengths at different speeds. Acta Physiologica Scandinavica 52: 366-371. 1961 Kohrt WM , Morgan DW, Bates B. Skinner JS. Physiological responses of triathletes to maximal swimming, cycling and running. Medicine and Science in Sports and Exercise 19: 51-55. 1987 Krahenbuhl G, Morgan D. Pangrazi R. longitudinal changes in distance-running performance of young males. International Journal of Sports Medicine. in press, 1989 Krahenbuhl G, Pangrazi R. Characteristics associated with running performance in young boys. Medicine and Science in Sports and Exercise 15: 486-490, 1983 Krahenbuhl G. Skinner J, Kohrt W. Developmental aspects of maximal aerobic power in children. In Terjung (Ed.) Exercise and sport sciences review. Vol. 13. pp. 503-538. Macmillan. New York, 1985 Larish D, Martin P, Mungiole M. Characteristic patterns of gait in the healthy old. Annals of the New York Academy of Sciences 515: 18-32. 1987 Leger L, Mercier D. Gross energy cost of horizontal treadmill and track running. Sports Medicine I: 270-277. 1984 Leger L. Mercier D. Gauvin L. The relationship between % ~02 max and running performance time. In Landers (Ed.) The 1984 Olympic Scientific Congress Proceedings. Vol. 13. Sport and elite performers. Human Kinetics. pp. 113-119. Champaign IL. 1986 Lemon P. Mullin J. The effect of initial muscle glycogen levels on protein catabolism during exercise. Journal of Applied Physiology 48: 624-629. 1980 Luhtanen p. Komi PV. Mechanical energy state during running. European Journal of Applied Physiology 38: 41-48. 1978 MacDougall JD. The anaerobic threshold: its significance for the endurance athlete. Canadian Journal of Applied Sports Sciences 2: 137-140. 1977 MacDougall J. Reddan W, Layton C. Dempsey J. Effects of metabolic hyperthermia on performance during heavy prolonged exercise. Journal of Applied Physiology 36: 538-544. 1974 MacDougall J. Roche P. Bar-Or O. Moroz J. Maximal aerobic capacity of Canadian school children: prediction based on agerelated oxygen cost of running. International Journal of Sports Medicine 4: 194-198. 1983 Mader A. Liesen H. Heck H. Phillippi H. Schurch P. Zur Beurteilung der sportartspezifischen Ausdauerleistungsfahigkeit. Sportarzt und Sportmedizin 27: 80-88. 109-112. 1976 Margaria R. Biochemistry of muscular contraction and recovery. Journal of Sports Medicine and Physical Fitness 3: 145-156. 1963 Margaria R. Cerretelli P. Aghemo P. Sassi J. Energy cost of running. Journal of Applied Physiology 18: 367-370. 1963a Margaria R. Cerretelli P. DiPrampero P. Massari C. Torelli G.

Factors Affecting Running Economy

Kinetics and mechani sm of oxygen debt contract ion in man . Journal of -\pplled Ph \s iology 18: 371-377. 1963b Marga ria R. Mangili F. C ulliea F. Cerretdli P. The kinetics of oxygen consumpt io n at the on set of muscula r exercise in man. Ergonomics 8: 44 -54. 1965 Margaria R. Oliva R. DIPrampero P. Cerretelli P. Energy utilizat io n in intl' rmith.'tlt "'\l' rl' i ~c of supermaximai intensity ,

J ou rnal of -\pplied Ph \ Slolog\ 26: 752- 756. 1969 Maron M. Hor\ath S. Wilkerson J. (iliner J . Oxygen uptake measurements during co mpetitl\ e marathon running. Journal of-\pplied Ph\slolog\ 40: 8 36-838. In 6 Martin P. \h'c hanlcal and ph"iologl ca l responses to lower extremity loading during running. MediC ine and Science In Spons an d Exermc 17: 42 7-433. I QS 5 Marlin P. Fcrnhall B. Krahen bu hl G . The etTect of worko ut inten sit y o n runn ing econom y and mechan ics. -\bstracl. Ho ng Kong Sports Med iCine Confe rence. Hong Kong. 1987 Maughan R. Lei pe r J . AerobiC capacity and frac tIOnal utili sa tIOn of ae rohic capac it y in elite and non-elite male and femal e marathon runners. European Journal of Applied Ph ys iolog\ 52: 80-87. I QS3 Ma\ers N. G utin B. Physio logical characteristics of elite prepubertal cross-country' runners. Medicine and Scien ce in Sports II : 172-176.1979 McMiken D. Daniels J. Aerobic requirem ents and ma ximum aao bic power in treadm ill a nd track runn ing. Medicine a nd Scie nce in Spons 8: 14-17. 1976 Morga n D. EtTects of a prolonged maximal run on running econo my and running mechanics. Unpublish ed d octoral di ssertati o n. Amona Sta te l'ni\·ersit y. 1988 Mo rgan D. Raldinl F. Ma rtin P. Da y- to-day stabilit y in running econom y and step len gth among well-trained male runners. .-\bstracl. Interna tIO na l Jou rnal of Spo ns Medicine 8: 242. 1987 Morgan D. Bald ini F. Ma rtin P. Kohrt W. Ten km performance and predi cted velocit y at Y02 max among well-trained male runners. Medicine and Scie nce in Spo rts a nd Exercise 2 1: 7883. 1989 Morga n D. Martin P. EtTects of stride length alteration on racewa lking econo m y. Ca nadian Journal of Applied Sports Sciences II 211-21 7.1986 Morgan D. Martin P. Krahen buhl G . Baldini F. Da ily sta bilit y in runlng economy and running mechanics. Unpu blished manuscript. 198 8 Myers M. Steudel K . EtTect of limb mass and its d istribution on the energetic cost of running. Journ al of Experimental Biology 11 6: 36.1-.1 73.1985 Orr G W. G ree n HJ . Hugh son RL. BennCll G. A computer linea r regresSion m odel to determine the \ entdatory anaerobic threshold. Journa l of Applied PhYSiol ogy 52: 1.149-1352. 1982 Pate R. Barnes C. Miller W -\ physiological comparison of perfo rmance- matched femal e and ma le dIStance runners. Research Quarterl\ fDr Ex erCise a nd Sport 56: 245 -2 50. 198 5 Pallon J. Voge l J . Cross-sec tional and lo ngitudinal eva luations of an endurance tralOlOg program. ~kdlc ine and SCience in Spons 9: 100-1 03. 19P Petray C. Kra henbuhl G. Running trai ning. instructi o n on run ning technique. and runmng econo my in IO-yea r-o ld males. Research Quarterl\ for ExerCise and Sport 56: 251-255. 1985 Pollock M. Submax imal and ma ximal working ca pacity of elite distance runners. Pan I: CardIOrespiratory aspects. Annals of the New York .-\cademy of Sciences 301: 3 10-322. 1977 Pollock M. Jackson .-\. Anes J. Ward .-\. Linnerud -\. et al. Bod\ compoSit IOn of elite cl~ss distance runners. Annal s of the Ne';" York Academy of Sciences 30 I: 361-310. 1977 Pollock M. Jackson A. Pate R. Discnminant analysis of physiological d itTerences between good and elite dIStance runners. Research Quarterly for Exercise and Spon 51' 52 1-532. 1980 Po .... ers SK . Hopktns P. Ragsdale MR. Oxygen uptake and ven-

329

tilatory' responses to various stride lengt hs in trained women. American Correct ive Therapy Jo urnal 36: 5-8. 1982 Pugh L. Oxygen intake and track and treadmill running wi th observations o n the etTect of air resistance. Journal of Physiology (London) 207: 823-835. 1970 Pugh L. The etTect of wind resistance in running and wa lking and the mechan ical etliclcncy of work aga inst honzonta l and vertical forces. Journal of Ph ysiology I London) 213: 255-270. 197 1 Ro bin so n S. Rohlnson D. Mo untj oy R. Bullard R. Influe nce o f fatigue on the efli ciency of men during exha ustIVe runs. J o urnal of Applied Ph ysiology 12: 197-20 I. 19 58 Rowell L. Brengelm a nn G. Murray J. Kran ing K. Kusumi F. Huma n metabolic response to hype rthermia during mild to maxim al exerci se. Journal o f Applied Physiology 26: 195-402. 1969 Ro wland T. G reen G. Ph ys iological responses to treadmill exercise in fe m a Je ~;: adult-child dltTeren ces. MediclOe a nd Science in Sports and ExerCISe 20: 474-4 78. 1988 Sa ltlO Il. Astra nd P-O. Maxi mal oxygen uptake in athletes. Journal of Applied Physiology 23: 353-3 58. 1967 Sa ltin B. Sten berg J. Circulato ry response to prolonged se vere e.xerCISe. Journal o f Applied PhY SIOlogy 19: 833-838. 1964 Sargent R. The relat ion between oxygen require ment and speed in ru nning. Proceedings of the Royal Society of London (B,ologyJ 100: 10-22. 1926 Sena y L. Kok R. EtTects of trai ning a nd heat acclimatio n on blood plasma contents of exerci sing men. J o urnal of Applied Physiology 4 3: 59 1-599. 1977 Sherman W. La mb D. Nutrition and prolo nged exerci se. In Lamb & Murray (Eds) Prolonged exerCISe. pp. 213-280. Benchmark Press. India napolis. 1988 Shonen M. Woollon S. Williams C. Mechanical energy changes and the oxygen cost of running. Engin eering Medicine 10: 213217 . 1981 Sidney K. Shephard R. Maximum testing of men and wom en in the seventh. eighth . and ninth decades of li fe. Journal of .-\pplied PhYSiology 4l 280-28 7. 1 977 Sjodin B. Jaco hs I. Onset of blood lactate accumulati on and marathon runn ing performance. Interna tio nal Journal of Spons MediCine 2: 166-170. 198 1 Sjod in B. Jaco bs I. Svendenhag J . Changes in onset of blood lactate accumulation tOBlA ) and muscle enzymes after training at OBLA . European Journal of Applied Physiology 49: 45-57. 1982 Sjodin B. Svendenhag J. Applted physiology of marathon runOIng. Spons Medi cine 2: 83-99. 1985 Sk in ne r JS. Hustler R. Bergteinova V. Buskirk ER . Percept io n of etTon during ditTerent types of exercise and under ditTerent envi ro nmenta l conditions. Med icine and Scie nce In Spo ns 5: 110-1 15. 197 3 Stegmann H. Kindermann W. Compari so n of prolonged exerCise tests at the indi\'idua l a naerobic threshold and the fixed anaerobic thres hold or 4 mmol/L lacta te. Internat IOnal Journal of Spons MediCine J 105- 110. 1982 S\ endenhag J. Sjod in B. Maximal an d subma xi mal o.xygen uptakes and blood lactate le\els In eltte male middle- and lo ngdista nce runners. Internatio na l J ournal of Sports Medicine 5: 25 5-2 6 1. 1984 Svendenhag J . Sjodin B. Ph\'siologlcal characteristics of eltte male runners i n and o tT-season. Can ad ia n J ourna l o f Applied Spon Scien ces 10: 127-133. Ins Tanaka K. Matsuura Y. Kumagai S. Matsuzaka .-\. Hirako ba K. et al. Relati o nshIps of a naerobic threshold and onset of blood lactate accumulation with endurance performance. European Journal of Applied Physiology 52: 51 -56. 1983 Taylor CR. The energetics o f terrestrial locom o tion and body size in \ enebrates. In Pedle\ (Ed .) Scale etTects In animal locomo ti o n. pp. 127-14\. Ac~demic Press. New York. 1977 Taylor C R. Force de"elopment during sustained locomotIOn: a determinant of ga it. speed and metaboltc power. Journa l of Experimental Biology 115: 253-262. 198 5 Taylor C R. Energetics of locom o tion: What sets the cost ~ In

Factors Affecting Running Economy

Saltin (Ed.) Biochemistry of exercise VI. pp. 409-415. Human Kinetics. Champaign. 1986 Taylor CR, Heglund NC, Maloiy GMO. Energetics and mechanics of terrestrial locomotion. I. Metabolic energy consumption as a function of speed and i:>ody size in birds and mammals. Journal of Experimental Biology 97: 1-21. 1982 Taylor CR. Heglund NC, McMahon T A. Looney TR. Energetic cost of generating muscular force during running. Journal of Experimental Biology 86: 9-18. 1980 Taylor CR. Schmidt-Nielsen K. Rub J. Scaling of energetic cost of running to body size in mammals. American Journal of Physiology 219: 1104-1107. 1970 Taylor CR, Shkolnik A. Dmi'el R. Baharav D. Borut A. Running in cheetahs, gazelles, and goats: energy cost and limb configuration. American Journal of Physiology 227: 848-850, 1974 Walsh M, Banister E. Possible mechanisms of the anaerobic threshold. Spons Medicine 5: 269-302, 1988 Waters R, Hislop H. Perry J. Thomas L. Campbell J. Comparative cost of walking in young and old adults. Journal of Orthopaedic Research I: 73-76. 1983 Wells C, Pate R. Training for performance of prolonged exercise. In Lamb and Murray (Eds) Prolonged exercise. pp. 357-391. Benchmark Press. Indianapolis. 1988 Whipp B, Scard C, Wasserman K. Oxygen deficit-oxygen debt relationship and efficiency of anaerobic work. Journal of Applied Physiology 28: 452-456. 1970 Whipp B. Wasserman K. Oxygen uptake kinetics for various in-

330

tensities of constant load work. Journal of Applied Physiology 33: 351-356. 1972 White T. Brooks G. [u-l4Cl glucose-alanine and leucine OXIdation in rats at rest and two intensities of running. American Journal of Physiology 240: E1555. 1981 Wilcox A. Bulbulian R. Changes in running economy relative to ~02 max during a cross-country season. Journal of Spons Medicine and Physical Fitness 24: 321-326. 1984 Williams K. Cavanagh P. Relationship between distance running mechanics. running economy. and performance. Journal of Applied Physiology 63: 1236-1245. 1987 Williams KR. Cavanagh PRo Biomechanical correlates with running economy in elite distance runners. Proceedings of the Nonh American Congress on Biomechanics. Montreal. pp. 287288. 1986 Williams KR. Cavanagh PRo ZifT JL. Biomechanical studies of elite female distance runners. International Journal of Spons Medicine 8(Suppl.): 107-118. 1987 Wilmore J. Brown C, Davis J. Body physique and composition of female distance runners. Annals of the New York Academy of Sciences 301: 764-776.1977

Authors' address: Don W. Morgan. Depanment of Physical Education, Spon. and Leisure Studies. Washington State University. Pullman. WA 99164-1410 (USA).