Efficient use of available energy is a prerequisite for successful distance running performance. The concept of running economy (RE) was introduced by Hill et al. in 1924 (17). They hypothesized that, as the running human became more mechanically efficient, the metabolic cost of running at any given velocity would be reduced. The RE is defined as the steady-state O2 for a given, submaximal running velocity (24,30). The O2submax is related linearly to running velocity. Several groups of researchers have established linear regression equations to reflect this relationship. In 1980, Conley and Krahenbuhl derived the following linear equation for the RE of their highly trained subjects: y = 0.209x − 5.67 (y = O2submax [ml·kg−1·min−1]; x = running velocity [m·min−1]) (6). Their subjects ran on a treadmill at speeds of 241, 268, and 295 m·min−1. The slope of this equation, 0.209, represents the RE of their subjects. This is in close agreement with the findings of other studies of RE, using similar running velocities (Bransford and Howley, 1977 : 0.204 and 0.203; Costill and Fox 1969 : 0.204; and Daniels et al. 1977: 0.201 ).
In competitive distance runners with similar O2max values, RE has been shown to have a very strong association with distance running performance (6,9,30). Several researchers have shown RE to be a superior predictor of endurance performance compared to O2max in competitive distance runners (15,30). Data from Conley and Krahenbuhl (6) demonstrated a very strong correlation between RE and 10-km performance in competitive distance runners with similar O2max values. They found that 65.4% of the variation observed in running performance could be explained by variations in RE. Many other variables have been identified as determinants of distance running performance. Among those with the strongest correlations are: O2max (24,28,34), O2 @ lactate threshold (O2 @ LT) (30,33,34), and velocity @ lactate threshold (v @ LT) (15,34). However, other factors including muscle power, strength, and endurance and flexibility and body composition impact RE and performance.
The influence that age exerts on running performance has been investigated (11,33,34). Researchers have reported that age and O2 @ LT are reliable predictors of endurance running performance in master or older runners (33). In a follow-up study of middle-aged and elderly endurance athletes, O2max, O2 @ LT, and age showed strong correlations with distance running performance (r = 0.751, 0.781, and −0.736, respectively) (34).
Some of the variables that have a strong influence on RE are known to be affected by age. It is well documented that O2max and maximal heart rate (HRmax) decline with age (28,30,33–35). Trappe et al. in 1996 reported a 5–7% decline in O2max per decade in highly trained male distance runners (35). Rogers et al. in 1990 showed a 5.5%/decade decline in his subjects. This resulted in an absolute average decline of 2.2 ml O2 ·kg− 1·min− 1 per decade (28). In addition, muscle strength (27), power (21), and flexibility (5) tend to decrease with age and may impact RE and performance.
Understanding the interaction between RE and factors affecting it is important to the runner, coach, and exercise physiologist. To date, there is no systematic study pertaining to the effect aging has on RE, and therefore, it is not clear exactly what factors are most important and how, and to what extent, decrements in these factors affect RE as one ages. This knowledge can play a role in defining training programs for distance runners in an effort to minimize these potential physiological decrements and maintain a competitive edge as one ages. Therefore, the purposes of this study were to investigate the relationship that age has on RE and identify factors that influence RE in competitive distance runners. We hypothesized that older distance runners would possess inferior RE compared to their younger counterparts. Secondly, it was expected that older runners would show declines in factors known to affect RE compared to their younger counterparts and that these factors would play an important role in the prediction of RE.
Experimental Approach to the Problem
We employed a cross-sectional study population to investigate the impact of age on RE. Successful, age-group winning distance runners were recruited to participate and 3 age groups were formed. Subjects reported to the laboratory on 3 separate occasions and completed a lactate threshold and O2max test (to determine and compare fitness levels across the groups), an RE test with step frequency (the main dependent variable compared across the groups), and a variety of fitness tests including body composition, muscular strength, muscular endurance, flexibility, and power (dependent variables thought to affect RE). The independent variable was age, and a 1-way analysis of variance (ANOVA) was the primary statistical tool used to determine differences in the dependent variables across the age groups.
A cross-sectional examination of 51 male (n = 28) and female (n = 23) competitive distance runners (5k–10k distances), who had finished first, second, or third place in their respective age category in local large road races (>500 finishers) was performed. Subjects were categorized into 1 of 3 age groups: Young (Y) 18–39 years (n = 18); Master (M) 40–59 years (n = 22); Older (O) 60 years and over (n = 11). Subject characteristics are listed in Table 1. All subjects trained at least 5 d·wk−1 and weekly running volume ranged from 65 to 110 km. Each subject performed tests over the course of 3 visits to the environmentally controlled Robert Kertzer Exercise Physiology Lab at the University of New Hampshire in Durham, New Hampshire. Subjects were screened with a personal health history questionnaire, and they signed an informed consent document that described the testing protocols. The University of New Hampshire Institutional Review Board approved all of the testing methods involved in the experiment. On visit 1, an LT test followed by a O2max test was performed. On visit 2, an RE test at 4 speeds was performed. Visit 3 included anthropometric measures and assessments of muscle strength, muscle power, muscle endurance, and flexibility. No more than 7 days between visits were allowed, and all tests were performed at approximately the same time of the day for each subject. Further, each subject wore the same shoes and clothing for all testing sessions. Finally, all subjects were asked to perform light, steady-state exercise for 2 days before any testing and all subjects reported to the laboratory well hydrated, which was confirmed by measure of urine specific gravity (USG) to ensure that all subjects were tested in a euhydrated state (USG <1.019 g·mol−1) (2).
Lactate Threshold Test
The LT test involved a discontinuous treadmill protocol with the subjects running on a motor-driven treadmill (Quinton Inc; Bothell, WA, USA). All subjects were familiar with running on a treadmill and before any testing, subjects ran on the treadmill with the respiratory apparatus and headgear. The test protocol required subjects to run 3-minute stages at increasing velocities at a 1% grade, which was kept constant for the duration of the test. Stage 1 treadmill velocity was selected by asking the subject what their submaximal ‘sustainable’ running pace was. Then, every 3 minutes, the speed of the treadmill was increased by 13.41 m·min− 1. After each 3-minute stage, subjects were asked to stop running, and straddle the treadmill belt. Finger-stick blood samples were collected in duplicate in capillary tubes and analyzed immediately for lactate concentration, using a portable lactate analyzer (YSI 1500 Sport; Yellow Springs, OH, USA). Sterile procedures were followed for each blood draw. The LT was defined as the running velocity where 2 successive 1-mM increases in blood lactate concentration occur (8). The O2 was measured (breath × breath) using indirect calorimetry with a metabolic cart (Viasys V229; Yorba Linda, CA), and the HR was monitored continuously using a 5-lead electrocardiogram (Quinton Inc). The HR and O2 were computed and averaged over the last minute of each workload. Once the LT was identified, the subject rested in a seated position for 5–10 minutes.
A maximal oxygen consumption test was administered after the rest period, using indirect calorimetry. Subjects performed this continuous, incremental treadmill test on a motor-driven treadmill (Quinton Inc). A modified Åstrand treadmill protocol was used, where subjects ran at 215 m·min−1 at 0% grade for 1 minute, and then, the grade was increased by 2.5% every minute until volitional exhaustion. Each maximal test was preceded by a 5-minute easy running warm-up consisting of light running. Respiratory gases were analyzed using a Viasys metabolic cart (Viasys V229). The HR was monitored continuously using a 5-lead electrocardiogram (Quinton Inc; Bothell) and was printed each minute. The O2max was defined as the highest 20-second O2 value recorded in the last 2 minutes of the test. Values were reported in ml O2·kg− 1·min− 1. Verbal encouragement was provided during the test. Criteria for terminating the test was a plateau in O2 with increasing workloads, achievement of HRmax (220 age ± 5%), or an R ≥ 1.10.
Measurement of Running Economy
The RE test involved the subjects running at 4 randomized submaximal speeds (134, 161, 188, 201, 215, or 241 m·min−1) for 5 minutes each. Subjects were asked to run at 161, 188, 215, and 241 m·min−1 for 5 minutes at each velocity. Subjects who felt that they could not run at the 241 m·min−1 velocity for 5 minutes ran at speeds of 134, 161, 188, and 201 m·min−1. Five minutes of recovery followed each of the 4 speeds. Respiratory gases were measured via indirect calorimetry, using a Viasys metabolic cart (Viasys V229). The HR was monitored using a 5-lead electrocardiogram (Quinton Inc). The steady-state O2 value in mlO2·kg−1·min−1 during the last minute of running at each speed was used as the subject's O2submax for that speed. The O2 values were plotted vs. speed, and a line of best fit was created. The slope of this line was used to represent the subject's RE. Individual slopes were analyzed to detect significant differences at the p < 0.05 level, and the mean slope of each age group was calculated. Step frequency was visually counted between the third and fourth minutes of each stage. Values were reported in steps·min−1.
Body Composition Testing
Body mass (kg) was recorded using a Detecto platform scale, and each subject was measured in shorts and singlet. Skinfold thickness was measured to estimate body density as described by Jackson et al. (18). Percent body fat was determined using the Siri equation (31). Skinfold measurements were taken in duplicate using Harpenden skinfold calipers by the same trained technician. If a difference of >1.0 mm was observed, a third measure was taken, and the average of the 2 closest values was used.
Muscle Strength Testing
A 1-repetition maximum (1RM) test was performed for the bench press and leg press exercises using CYBEX selectorized horizontal leg press and upright chest press machines. All subjects had experience in similar strength training machines. The 1RM protocol as described by Kraemer et al. (20) was followed. Total weight (kg) was recorded.
Muscle Endurance Testing
After the muscle strength testing, and sufficient rest, a muscle endurance test was administered, where subjects were asked to perform the bench press and the leg press exercises on CYBEX selectorized machines at 50% of their respective 1RM, for as many repetitions as possible. The subjects were asked to lift to a specified metronome cadence of 30 lifts·min−1. The total number of lifts was recorded.
A standard sit-and-reach test using a flexibility box was performed to assess hamstring and lower back flexibility. The subjects were allowed to stretch and warm up before any measurements; they then sat on a mat with their feet flat against the flexibility box and reached forward as far as possible toward their toes with their knees fully extended. The subjects held the static contraction for approximately 2 seconds, and the distance was recorded. Three trials were performed, and the highest value (cm) was recorded.
Vertical Jump Power Testing
A modification of the Sargent jump test (29) was used to estimate relative power. Subjects warmed up and then stood with their dominant arm against a wall and extended the arm upward along a wall-mounted scale, and this height was recorded as the standing reach (cm). Subjects chalked their hand and then performed 3 countermovement jumps with the highest jump height (cm) recorded. Standing reach was subtracted, and this value was used with body mass to calculate vertical jump power using the Lewis equation (14).
The independent variable in this study was age. Dependent variables included RE slope, v @ LT, O2 @ LT, HRmax, % HRmax, O2max, % O2max, step rate, muscle strength, muscle endurance, flexibility, and vertical jump power. A 1-way ANOVA was used to detect interactions among the independent and dependent variables. A Neuman–Kuels post hoc analysis was used to detect significant differences when appropriate. In an effort to quantify the bivariate relationships, Pearson Product correlation coefficients were calculated. Finally, to predict RE, forward stepwise multiple regression analyses were conducted at 4 running speeds. Using the data from Conley and Krahenbuhl (6), a power analysis revealed that this had >80% power, and the risk of a type 1 error was <5% for detecting significant differences in the dependent variables. Effect sizes (ES) were also calculated and found to be either “moderate” (ES = 0.6–1.2) or “large” (ES > 1.2) (12) for all significant dependent variables. Statistical significance was accepted at the p ≤ 0.05 alpha level for all comparisons. All analyses were conducted using Statistica software.
Running Performance and Fitness Level
Table 2 shows that mean personal best 5-km times were significantly different among age and gender groups. Table 2 also compares New Hampshire state 5-km records for correspondent age and gender categories. These record times were significantly lower than in the subjects in this study. On average, subject 5-km personal best times were within 17.12% of New Hampshire state records. However, based on O2max data, each of our subjects would be classified as having either an “excellent” or “superior” level of fitness when compared to age-matched, gender-specific normative data and would fall in the 90–99 percentiles against these data (1).
Running Economy and Slopes
Table 3 shows oxygen uptake values for each group at each running speed. No significant differences were noted; however, only 7 out of 11 subjects in the O group were able to complete the fastest running speed. Percent of O2max values are shown in Table 4. In general, the Y group completed each speed at a significantly lower percentage of O2max than M and M was significantly lower than O group. Figure 1 shows that there were no significant differences in RE slopes among age groups observed in the present study groups. Y had a mean RE slope of 0.1827 (y = 0.1827x − 0.2974; R 2 = 0.9511), M had a mean slope of 0.1988 (y = 0.1988x − 1.0416; R 2 = 0.9697), and O had a mean slope of 0.1727 (y = 0.1727x − 3.0252; R 2 = 0.9618; where the variable y = O2submax, and the variable x is the running velocity (m·min− 1).
Maximal Cardiorespiratory Factors
Significant differences were observed in O2max between Y (64.1 ± 3.2 ml O2·kg− 1·min− 1) and O (44.4 ± 1.7 ml O2·kg− 1·min− 1), and M (56.8 ± 2.7 ml O2·kg− 1·min− 1) and O. Y and M were not significantly different from each other (Table 5). Table 5 also shows that the maximal HR was significantly different among all groups.
Submaximal Cardiorespiratory Factors
Velocity at Lactate Threshold was significantly different among all age groups (Table 5). The HR @ LT showed that the O group was significantly higher compared to the Y and M groups (Table 5). The O2 @ LT showed that the O group was significantly lower compared to the Y and M groups (Table 5).
Factors Affecting Stride Kinematics
Hamstring and lower back flexibility assessed by the sit-and-reach test was significantly lower in O, compared to Y and M. Y and M were not significantly different from each other (Table 6). In general, flexibility was negatively correlated to but significantly contributed to the prediction of RE at velocities of 188, 215, and 241 m·min−1.
No significant differences were detected in step frequency at any velocity among groups. No differences among groups were observed for lower body muscle strength or endurance in our subjects (Table 6). Upper body muscle endurance was not different among groups. However, upper body muscle strength was significantly lower in O compared to in Y and M (Table 6). Relative peak power as estimated from the vertical jump test was significantly lower for the O vs. Y groups. No difference was noted between the Y and M groups (Table 6). Upper body muscular strength was significantly and positively correlated to all running velocities, whereas lower body strength showed this same relationship at the 215 and 241 m·min− 1 velocities. Peak power was positively and significantly related to RE at the 241 m·min− 1 speed.
No differences among groups were observed for body fat % or body mass in our subjects (Table 1) when gender was collapsed. Both of these variables were negatively and significantly correlated with RE at the fastest velocity (241 m·min− 1).
Predicting Running Economy
The multiple regression analysis to predict RE at 161 m·min−1 showed that age, O2max, %O2 −161 m·min−1, and lower body strength were significant variables (F(4,33) = 272.02, p < 0.0001). The adjusted R 2 was 0.97 suggesting that 97% of the variance in predicting RE at this speed was accounted for by these 4 variables. For the 188 m·min−1 speed the following 5 variables were significant predictors: O2max, %O2 − 188 m·min−1, %O2max @ LT, %HRmax @ LT, 161 m·min−1, and flexibility were the significant predictors (F(5,34) = 231.79, p < 0.0001; R 2 = 0.97). At 215 m·min−1 the significant predictors were O2max, %O2 − 215 m·min−1, %O2max @ LT, O2 @ LT, flexibility, upper body strength, lower body strength, and age (F(8,30) = 282.66, p < 0.0001; R 2 = 0.98). Finally, for the 241 m·min−1 speed, the significant predictors were O2max, %O2 − 241 m·min−1, HRmax, %HRmax @ LT, HR @ LT, flexibility, upper body strength, lower body strength, power, and body mass (F(10,21) = 346.53, p < 0.0001; R 2 = 0.99).
One purpose of this study was to investigate the impact of age on RE in competitive distance runners. It was hypothesized that O2submax would be higher in older compared to in younger subjects (i.e., poorer RE). The results from this study do not support this hypothesis because O2submax and differences in RE slopes were not observed.
Previously reported mean RE slopes in elite distance runners were between 0.201 and 0.209 (7,34). In these studies, investigators observed linear relationships between RE and running performances. In our investigation, we recruited competitive distance runners who would qualify as ‘subelite’ athletes. These subjects produced a mean RE slope of 0.1580. The difference in slopes between the classes of athletes suggests that elite athletes have different O2submax responses to increasing running velocities. We offer the following explanations for this difference:
(a) The subjects in this study ran at treadmill velocities ranging from 134 to 241 m·min−1. The subjects in the aforementioned studies ran at velocities of 241, 268, and 295 m·min−1. For some of our subjects, the latter velocities would likely have been intensities above the LT, particularly in the O group. It would be challenging to compare running economies in differently trained subjects at very different velocities. To apply RE values to different velocities than what was actually measured requires extrapolation of the linear regression, using the slope from the measured velocities. Bickham et al. in 2004 observed that the O2 at running velocities above the LT responded in a nonlinear fashion, and that using O2–velocity data from below LT can result in overestimation of O2 values above the LT (3). Although our RE data were from steady state velocities and below the LT, this point supports the notion that the slope of the linear regression equation is influenced by running velocity and that there is a strong likelihood that, within subjects, RE slopes would vary depending on the ranges of velocities selected for their measurements.
(b) The subjects in the prior studies were deemed “elite” distance runners. In this study, we sought “successful” distance runners, that is, runners who finished first, second, or third place in their respective age divisions in local large races. After comparing 5-km state records for correspondent age and gender groups, our subjects would qualify as ‘subelite’ distance runners (Table 2). This may have impacted the performance homogeneity of the subject population in this study. In elite-cluster athletes, the performance and physiological variability among athletes may be far smaller than in subelite athletes. This could have direct implications in the RE of an athlete, particularly in how oxygen demand responds to increasing running velocities. Elite-cluster distance runners may appear to be less economical to a subelite runner, at any given running velocity, but because of a higher maximal aerobic capacity, may actually be working at a lower fraction of their aerobic capacity. This fractional use of an athlete's O2max, sometimes referred to as ‘performance O2,’ has been shown to be a strong discriminating factor for distance running performance (9,24,30).
Although, to the best of the authors' knowledge, this is the first study to address RE comparisons across age, it is not the first to detect significant differences in a variety of factors that may impact RE across age. Indeed, the second purpose of this study was to identify factors thought to influence RE in competitive distance runners. We hypothesized that older runners would show declines in factors known to affect RE compared to their younger counterparts and that these factors would play an important role in the prediction of RE. The data from this study support this hypothesis.
An investigation by Evans et al. in 1995 suggested that age-associated decrements in O2max and v @ LT had strong correlations with declines in 10-km performance in highly trained female runners (13). The authors cited that the influence that these declines exerted on distance running performance was not uniform with advancing age. Two physiological variables that we measured seemed to decline uniformly with age. The HRmax and v @ LT were progressively lower in our M and O subject groups, compared to in the younger subjects. O2max, however, was significantly lower only in the O group. Our older subjects also exhibited lower HR @ LT and lower absolute O2 @ LT. These findings are in agreement with those of similar studies (13,16,23). Higher running velocities at the LT have obvious performance implications. An athlete who can recruit less of his or her maximal capacity at any given workload can either maintain a higher level of reserve to be used at a later point in a race, or apply more effort yielding higher average running velocities, for longer periods of time. Additionally, a higher maximal heart rate suggests a higher maximal cardiac output, thereby increasing an athlete's endurance capacity.
Flexibility was shown to be significantly lower in the O subject group, possibly accounting for some of the decline in performance times seen in our subjects (Table 2). Running mechanics can be affected considerably from decreased flexibility, particularly with regard to step length and step frequency,2 major factors of running velocity. In our examination, we did not see group differences in step frequency. Therefore, it appears that the decline in flexibility observed in our O subject population did not influence step frequency. It is possible, however, that declines in step length could contribute to the slower 5-km times seen in the O subjects. Daniels et al. in 1985 suggested that older individuals, as a result of an aging musculoskeletal system may be ‘economizing’ with the force of each stride by adopting a shorter step length (11).
Our O subjects exhibited lower upper body strength values compared to M and Y subjects. Other factors known to impact stride kinematics, such as lower body muscle strength and endurance, were not different among age groups. In a 1997 investigation of female collegiate crosscountry athletes, Johnston et al. observed training-induced improvements in upper and lower body strength that were associated with improvements in RE (19). The authors suspected that these strength improvements influenced the running mechanics of the athletes, resulting in a more economical runner, thereby lowering the energy cost of running. In our older subjects, we detected differences only in upper body strength. This may account for some of the reason why age-related declines in RE were not detected among age groups.
However, this study did find that there was increasing importance placed on upper and lower body strength and muscle power as running velocity increased as shown by the regression analyses. This finding is supported by Paavolainen et al. (26) and Storen et al. (32) who both showed that through either 9 or 8 weeks, respectively, of intense strength training that RE improved significantly. Certainly, from an aging perspective, the importance of maintaining strength and power is important if the goal is to run fast and economically. Noakes (25) was the first to suggest that distance performance is impacted not only by central factors related to O2max but also by the ability of muscle to generate power. The O subjects in this study consistently showed lower upper body strength and power compared to their younger counterparts. As the aging process progresses, muscle mass decreases and muscle fibers begin to lose their nerve supply. Type 2 muscle fibers lose nerve supply more readily than type 1 fibers (22). Therefore, older individuals have less muscle mass and a greater percentage of type 1 fibers compared to younger people, and this results in lower strength values and a decreased ability to generate power. The O runners in this study did not lose a significant amount of lower body strength (although it was lower than in Y and M) presumably because of the continued use of their lower body musculature while running. However, upper body strength was significantly lower in the O compared to in the Y and M groups, and it is this strength that allows a runner to initiate a change in speed rapidly or to run uphill. This suggests, then, that upper body strengthening should be incorporated into a distance runner's training regiment. In addition, some lower body plyometric or explosive strength training should be a component of a comprehensive distance runner's training program.
This cross-sectional analysis did not show differences in RE among subjects in different age groups. This does not clearly answer the question of whether or not RE declines with age. A longitudinal study of RE in the same distance runners could potentially reflect the influence that aging poses on the energetics of distance running. This experimental design presents challenges, though. Recruiting subjects for a second, and maybe a third measurement series is difficult. Additionally, changes in training status, injuries, and subject dropout would contribute to these challenges.
In conclusion, in this diverse population of competitive distance runners, we did not detect intergroup differences in RE. Differences observed in running performance may be a function of declines in maximal and submaximal cardiovascular and hemodynamic variables and changes in muscle power.
It appears that competitive, older runners maintain their ability to run economically at submaximal velocities. Therefore, in an effort to prevent decrements in performance, these older runners and their coaches should focus some of their training regimen on factors that clearly decline with age. Maintaining muscular strength and endurance (especially upper body) through well-structured strength and plyometric programs would be advised. In addition, run training using threshold runs, intervals, fartlek runs, and hill training should be incorporated periodically and systematically. Finally, keeping muscles and muscle groups as flexible as possible through a regular stretching program may be warranted.
The authors thank the research subjects who put in their effort and gave their valuable time to this project. Funding was provided by the Undergraduate Research Opportunity Program at the University of New Hampshire. The results of this study do not constitute endorsement by the National Strength and Conditioning Association.
1. ACSM's Guidelines for Exercise Testing and Prescription
(8th ed.). Baltimore, MD: Lippincott Williams & Wilkins, 2010.
2. Armstrong, LE, Maresh, CM, Castellani, JW, Bergeron, MF, Kenefick, RW, LaGasse, KE, and Riebe, D. Urinary indices of hydration status. Int J Sports Nutr
4: 265–279, 1994.
3. Bickham, DC, Gibbons, C, and Rossignol, LE. O2
is attenuated above the lactate threshold
in endurance-trained athletes. Med Sci Sport Exerc
36: 297–301, 2004.
4. Bransford, DR and Howley, ET. Oxygen cost of running in trained and untrained men and women. Med Sci Sport Exerc
9: 41–44, 1977.
5. Chapman, EA, deVries, HA, and Swezey, R. Joint stiffness: Effects of exercise on young and old men. J Gerentol
27: 218–221, 1972.
6. Conley, DL and Krahenbuhl, GS. Running economy and distance running performance
of highly trained athletes. Med Sci Sport Exerc
12: 357–360, 1980.
7. Conley, DL, Krahenbuhl, GS, Burkett, LN, and Millar, AL. Physiological correlates of female road racing performance. Res Q Exerc Sport
52: 441–448, 1981.
8. Costill, DL and Fox, EL. Energetics of marathon running. Med Sci Sport Exerc
1: 81–86, 1969.
9. Costill, DL, Thomason, H, and Roberts, E. Fractional utilization of the aerobic capacity during distance running. Med Sci Sports Exerc
5: 248–252, 1973.
10. Coyle, EF, Sidossis, LS, Horowitz, JF, and Beltz, JD. Cycling efficiency is related to the percentage of type I muscle fibers. Med Sci Sport Exerc
24: 782–788, 1992.
11. Daniels, JT. A physiologist's view of running economy. Med Sci Sports Exerc
17: 332–338, 1985.
12. Drinkwater, EJ, Hopkins, WG, McKenna, MJ, Hunt, PH, and Pyne, DB. Modeling age and secular differences in fitness between junior basketball players. J Sport Sci
25: 869–878, 2007.
13. Evans, SL, Davy, KP, Stevenson, ER, and Seals, DR. Physiological determinants of 10-km performance in highly trained female runners of different ages. J Appl Physiol
78: 1931–1941, 1995.
14. Fox, EL and Matthews DK. Interval Training: Conditioning for Sports and General Fitness
. Philadelphia, PA: WB Saunders College, 1974.
15. Grant, S, Craig, I, Wilson, J, and Aitchison, T. The relationship between 3 km running performance
and selected physiological variables. J Sports Sci
15: 403–410, 1997.
16. Hawkins, SA, Marcell, TJ, Jaque, SV, and Wiswell, RA. A longitudinal assessment of change in O2
max and maximal heart rate in master athletes. Med Sci Sport Exerc
33: 1744–1750, 2001.
17. Hill, AV, Long, CNH, and Lupton, H. Muscular exercise, lactic acid, and the supply and utilization of oxygen. Parts VII–VIII. Proc Roy Soc B
97: 155–176, 1924.
18. Jackson, AS and Pollock, ML. Generalized equations for predicting body density of men. Br J Nutr
40: 497–504, 1978.
19. Johnston, RE, Quinn, TJ, Kertzer, R, and Vroman, NB. Strength
training in female distance runners: Impact on running economy. J Strength Cond Res
11: 224–229, 1997.
20. Kraemer, WJ, Ratamess, NA, Fry, AC, and French, DN. Strength
training: Development and evaluation of methodology. In: Physiological Assessment of Human Fitness
. Maud, PJ and Foster, C, eds. Champaign, IL: Human Kinetics, 2006. pp. 119–150.
21. Larsson, L, Grimby, G, and Karlsson, J. Muscle strength
and speed of movement in relation to age and muscle morphology. J Appl Physiol
46: 451–456, 1979.
22. Lexell, J and Downham, D. What is the effect of ageing on type II muscle fibers? J Neurol Sci
107: 250–251, 1992.
23. Mattern, CO, Gutilla, MJ, Bright, DL, Kirby, TE, Hinchcliff, KW, and Devor, ST. Maximal lactate steady state declines during the aging process. J Appl Physiol
95: 2576–2582, 2003.
24. Morgan, DW, Martin, PE, and Krahenbuhl, GS. Factors affecting running economy. Sports Med
7: 310–330, 1989.
25. Noakes, TD. Implications of exercise testing for prediction of athletic performance: A contemporary perspective. Med Sci Sports Exerc
20: 319–330, 1988.
26. Paavolainen, L, Häkkinen, K, Hamalainen, I, Nummela, A, and Rusko, H.Explosive-strength
training improves 5-km running time by improving running economy and muscle power
. J Appl Physiol
86: 1527–1533, 1999.
27. Reed, RL, Pearlmutter, L, Yochum, K, Meridith, KE, and Mooradian, AD. The relationship between muscle mass and strength
in the elderly. J Am Geriatric Soc
39: 551–561, 1991.
28. Rogers, MA, Hagberg, JM, Martin, WH, Ehsani, AA, and Holloszy, JO. Decline in O2
(max) with aging in master athletes and sedentary men. J Appl Physiol
68: 2195–2199, 1990.
29. Sargent, DA. The physical test of man. Am Phys Educ Rev
25: 188–194, 1921.
30. Saunders, PU, Pyne, DB, Telford, RD, and Hawley, JA. Factors affecting running economy in trained distance runners. Sports Med
34: 465–485, 2004.
31. Siri, WE. Body composition from fluid spaces and density—Analysis of methods. Nutr
9: 481–491, 1993.
32. Storen, O, Helgerud, J, Stoa, EM, and Hoff, J. Maximal strength
training improves running economy in distance runners. Med Sci Sports Med
40: 1087–1092, 2008.
33. Takeshima, N and Tanaka, K. Prediction of endurance running performance
for middle-aged and older runners. Br J Sports Med
29: 20–23, 1995.
34. Tanaka, K, Takeshima, N, Kato, T, Niihata, S, and Ueda, K. Critical determinants of endurance performance in middle-aged and elderly endurance runners with heterogenous training habits. Eur J Appl Physiol
59: 443–449, 1990.
35. Trappe, SW, Costill, DL, Vukovich, MD, Jones, J, and Melham, T. Aging among elite distance runners: A 22-yr longitudinal study. J Appl Physiol
80: 285–290, 1996.