Journal of Strength and Conditioning Research, 2001, 15(1), 109115

q 2001 National Strength & Conditioning Association

A New Approach to Monitoring Exercise Training

CARL FOSTER, JESSICA A. FLORHAUG, JODI FRANKLIN,
LORI GOTTSCHALL, LAURI A. HROVATIN, SUZANNE PARKER,
PAMELA DOLESHAL, AND CHRISTOPHER DODGE
Department of Exercise and Sport Science, University of WisconsinLa Crosse, Wisconsin 54601.

ABSTRACT periods of hard and easy training (2, 5, 6, 9, 1115, 18

The ability to monitor training is critical to the process of 20). These practicalities are reflected in the practice of
quantitating training periodization plans. To date, no method coaches who design highly detailed periodized train-
has proven successful in monitoring training during multi- ing programs (10, 21). Unfortunately, although period-
ple types of exercise. High-intensity exercise training is par- ized training programs are in their essence quantita-
ticularly difficult to quantitate. In this study we evaluate the tive, there has been great difficulty in finding a way
ability of the session rating of perceived exertion (RPE) meth- to effectively quantitate training using a single term.
od to quantitate training during nonsteady state and pro- Endurance athletes have often used the training vol-
longed exercise compared with an objective standard based ume (kilometers per week) as an index of training
on heart rate (HR). In a 2-part design, subjects performed
with reasonable effectiveness (13, 15). However, mea-
steady state and interval cycle exercise or practiced basket-
ball. Exercise bouts were quantitated using both the session surement of training volume ignores the critical im-
RPE method and an objective HR method. During cycle ex- portance of high-intensity training bouts (17). For ath-
ercise, the relationship between the exercise score derived letes training for strength and/or power, the use of
using the session RPE method and the HR method was high- the volume of training is an inadequate tool because
ly consistent, although the absolute score was significantly of the overriding importance of intensity.
greater with the session RPE method. During basketball, There have been several previous attempts at de-
there was a consistent relationship between the 2 methods veloping a single term for quantitating training. In the
of monitoring exercise, although the absolute score was also late 1960s, Cooper (7) proposed the concept of aero-
significantly greater with the session RPE method. Despite
bics points, which integrated exercise duration and
using different subjects in the 2 parts of the study, the re-
gression relationships between the session RPE method and
the absolute intensity of aerobic training activities. Al-
the HR method were nearly overlapping, suggesting the though this approach was highly successful in terms
broad applicability of this method. We conclude that the ses- of guiding the nonathletic public into fitness exercise,
sion RPE method is a valid method of quantitating exercise the lack of an index of the relative training intensity
training during a wide variety of types of exercise. As such, (which is much more critical as an index of how likely
this technique may hold promise as a mode and intensity- a given exercise bout is to induce a training effect)
independent method of quantitating exercise training and dictated that this method would lack the ability to ad-
may provide a tool to allow the quantitative evaluation of equately describe the training load. Banister et al. (2,
training periodization plans.
9, 18) have developed the concept of the training im-
pulse (TRIMP) as a strategy for integrating the com-
Key Words: periodization, heart rate, perceived exer- ponents of training into a single term that allows a
tion, interval training, prolonged exercise systems analysis approach to training. This method
has shown great promise relative to understanding the
Reference Data: Foster, C., J.A. Florhaug, J. Franklin, L. training response and has been extended by Busso et
Gottschall, L.A. Hrovatin, S. Parker, P. Doleshal, and
C. Dodge. A new approach to monitoring exercise al. (6), Foster et al. (11, 12, 14, 15), and Mujika et al.
training. J. Strength Cond. Res. 15(1):109115. 2001. (19, 20). There are at least 2 important limitations to
the TRIMP concept developed by Bannister et al. (2, 9,
18). First, although monitors with the capacity of in-
tegrating the heart rate (HR) responses over long pe-
Introduction riods of time are widely available, if an athlete forgets

M any studies have stressed the importance of the

training load in enhancing athletic performance
and the changes in performance attributable to varying
to use his or her HR monitor or if the HR monitor has
a technical failure during the exercise bout, informa-
tion regarding that training session is lost. Second, HR

109
110 Foster, Florhaug, Franklin, Gottschall, Hrovatin, Parker, Doleshal, and Dodge

is a comparatively poor method of evaluating very Table 1. Mean (6SD) characteristics of the subjects in
high-intensity exercise such as weight training, high- part 1.*
intensity interval training, and plyometric training.
Thus even with the most optimal HR monitoring strat- Men Women
egy, integration of the TRIMP does not translate well
Age (yr) 23.0 6 3.6 21.3 6 1.5
to very high-intensity exercise training. We have de-
Height (cm) 177 6 4 165 6 8
veloped a modification of the rating of the perceived Weight (kg) 70.8 6 7.2 63.8 6 4.3
exertion method (the session RPE), which uses RPE as Percent fat 11.1 6 4.8 20.9 6 2.7
a marker of training intensity within the TRIMP con- Peak power output (W) 315 6 34 237 6 33
cept (11, 12, 14, 15). This method has been shown to Peak power output (Wkg21) 4.52 6 0.52 3.68 6 0.38
be related to both HR and blood lactate markers of Peak VO2 (Lmin21) 3.84 6 0.30 2.94 6 0.34
exercise intensity (14). However, our previous evalua- Peak VO2 (mlmin21kg21) 54.6 6 2.4 46.2 6 3.5
tion of the session RPE method has been based pri- Peak HR (bmin21) 198 6 10 186 6 7
marily on responses during 30 minutes of steady state IAT power output (W) 228 6 25 188 6 48
exercise within a comparatively modest range of ex- IAT power output (Wkg21) 3.28 6 0.65 2.94 6 0.61
ercise intensities. Given the importance of both high- IAT HR (bmin21) 174 6 19 159 6 16
intensity training and extensive training bouts within
* HR 5 heart rate; IAT 5 individual aerobic threshold.
the training plan of contemporary athletes, informa-
tion regarding the stability of the session RPE method
vs. HR methods of monitoring training during exer-
cise other than brief steady state exercise is important. Table 2. Mean (6SD) characteristics of the subjects in
Accordingly, the intent of this study was to evaluate Part 2.
the relationship of the session RPE- and HR-based
methods of monitoring training during different forms Age (yr) 20.2 6 1.5
Height (cm) 191.4 6 4.9
of exercise training.
Weight (kg) 89.3 6 7.8
Percent fat 12.8 6 2.8
Methods Peak VO2 (Lmin21) 4.60 6 0.50
This study was conducted in 2 separate but related Peak VO2 (mlmin21kg21) 51.5 6 2.2
parts. In the first part we chose a common condition- Ventilatory threshold (Lmin21) 3.32 6 0.54
Respiratory compensation threshold
ing activity that allowed good quantitative control of
(Lmin21) 3.71 6 0.44
the exercise performed (cycle ergometry). This allowed HRpeak (bmin21) 182 6 9
an idealized approach to both steady state and interval HR at VT (bmin21) 136 6 6
exercise that we felt would be generally representative HR at RCT (bmin21) 150 6 4
of a variety of conditioning activities. During this
phase of the study, the subjects were 12 well-trained, * HR 5 heart rate.
recreational-level cyclists (m 5 6, F 5 6). In the second
part, the subjects were members of a collegiate mens
spirometry (Quinton Q-Plex, Seattle, WA). Peak VO2
basketball team (n 5 14). Each subject provided in-
was defined as the highest continuous full minute VO2
formed consent prior to participation, and the study
observed during the test. HR was measured by radio-
protocol was approved by the university institutional
telemetry (Polar Electro Oy, Port Washington, NY).
review board. Some descriptive characteristics of the
Blood lactate was measured in capillary blood ob-
subjects are provided in Tables 1 and 2.
tained from a fingertip at rest, at the end of each stage
Part 1 of exercise, and at 1, 3, 5, and 10 minutes postexercise
Prior to the experimental protocol, each subject was using an enzyme electrode system (YSI Sport, Yellow
evaluated during maximal incremental exercise on an Springs, OH). The individual anaerobic threshold
electrically braked cycle ergometer (Lode, Gronningen, (IAT) was calculated on the basis of the exercise and
Netherlands). The subjects pedaled at freely chosen recovery blood lactate concentrations according to
revolutions per minute (rpms) within the range of 60 Stegmann et al. (22).
80. The test began at a power output of 50 W for men, Subsequently, each subject performed 8 randomly
40 W for women over 60 kg body weight, and 30 W ordered exercise training bouts, which included a ref-
for women under 60 kg body weight. The power out- erence 30-minute steady state bout at a power output
put was increased by the same increment every 3 min- equivalent to 90% of the IAT, 2 additional steady state
utes until the subject could no longer continue. The exercise bouts at the same power output but of 60- and
peak power output was interpolated based on the pro- 90-minutes duration, and 5 interval bouts at the same
portional time achieved during the terminal stage. Ox- mean power output. The interval bouts were 30 min-
ygen uptake (VO2) was measured using open-circuit utes in duration and included variations in interval
 Monitoring Exercise Training 111

Figure 2. Schematic representation of the summated HR

zone method that serves as the objective basis for compari-
son for the session RPE method. Five HR zones are calcu-
lated based on percentages of the HR peak: 5060% (zone
1); 6070% (zone 2); 7080% (zone 3); 8090% (zone 4);
and 90100% (zone 5). After the exercise session, the HR
monitor is downloaded and the cumulated time is each
zone is calculated. The time in each zone is then multiplied
by the value for that zone and the results summated.

Figure 1. Modification of the category ratio rating of per-

layed securing the session RPE rating for 30 minutes
ceived exertion (RPE) scale for this study. The verbal an-
chors have been changed slightly to reflect American idio-
so that particularly difficult or particularly easy seg-
matic English (e.g., light becomes easy; strong or severe be- ments toward the end of the exercise bout would not
comes hard). Briefly, the athlete is shown the scale approxi- dominate the subjects rating. In this context, it is im-
mately 30 minutes following the conclusion of the training portant to note that the momentary RPE during the
bout and asked How was your workout? In our experi- interval bouts often varied quite substantially based on
ence, approximately 8090% of athletes will give a single the momentary activity pattern. However, the session
number representing the training session. The remaining RPE represents a single global rating of the intensity
athletes usually insist on fractionating and summating the for the entire training session. We have not encoun-
component parts of the session. tered difficulties with the subjects understanding our
intent, particularly with the slightly modified verbal
anchors presented in Figure 1, either in this study or
magnitude (610, 625, and 650% of mean power out- in our previous work with this technique (11, 12, 14,
put with a constant 60-seconds/60-seconds exercise/ 15).
recovery schedule) and in interval duration (0.5 min- An exercise score (e.g., TRIMP) for each bout was
utes/0.5 minutes, 1.0 minute/1.0 minute, and 2.0 min- computed by multiplying the duration of the exercise
utes/2.0 minutes with a constant 625% difference in bout by the session RPE for that bout. As an objective
mean power output for exercise/recovery). Through- reference method for quantitating each exercise bout,
out each exercise bout, HR was measured by radiote- the HR monitor was downloaded using software that
lemetry. At rest and at 10-minute intervals, blood lac- allowed evaluation of the accumulated time in each of
tate was measured in capillary blood obtained from a 5 HR zones based on 5060%, 6070%, 7080%, 80
fingertip and perceived exertion was obtained using 90%, and 90100% of HRpeak, as suggested by Ed-
the category ratio (e.g., 010) RPE scale of Borg (4). wards (8) and used in our previous work (11, 12, 14,
Thirty minutes following the completion of each ex- 15). An exercise score (e.g., TRIMP) for that bout was
ercise bout, the subject was shown the RPE scale with then computed by multiplying the accumulated du-
verbal anchors (Figure 1) and asked to provide a rating ration in each HR zone by a multiplier for each zone
of the overall difficulty of the exercise bout, the session (5060% 5 1; 6070% 5 2; 7080% 5 3; 8090% 5 4;
RPE. We have previously used this method in studies and 90100% 5 5) and summating the result (Figure
of monitoring exercise training (11, 12, 14, 15). This 2).
use of the RPE method is somewhat different than the
conventional approach that asks the subject to rate Part 2
with highly standardized verbal instructions how dif- Prior to the experimental protocol, each subject was
ficult they perceived the exercise to be at a particular evaluated during incremental treadmill exercise using
moment. Rather, we explained to the subject that we an Astrand protocol. Exercise was continued to voli-
wanted a global rating of the entire training bout us- tional fatigue. VO2 was measured using open-circuit
ing whatever cues they felt to be appropriate. We de- spirometry and VO2peak was defined as the highest
112 Foster, Florhaug, Franklin, Gottschall, Hrovatin, Parker, Doleshal, and Dodge

Table 3. Serial (mean 6 SD) responses of the outcome measures.

Steady state 0.5 m/0.5 m 1.0 m/1.0 m 2.0 m/ 2.0 m 610% 625% 650%

Heart rate (bmin21)

Rest 79 6 13 77 6 11 84 6 9 82 6 12 79 6 6 84 6 9 82 6 10
10 m 155 6 10 137 6 16 159 6 9 166 6 12* 153 6 10 159 6 9 155 6 13
20 m 159 6 11 140 6 17 155 6 14 169 6 12* 158 6 11 166 6 14* 164 6 11*
30 m 160 6 13 163 6 16 169 6 11 172 6 14* 163 6 9 169 6 11* 170 6 13*
60 m 164 6 12
90 m 166 6 9
Blood lactate (mmoll21)
Rest 1.1 6 0.5 0.9 6 0.3 1.5 6 0.6 1.5 6 0.8 1.0 6 0.2 1.5 6 0.6 1.1 6 0.5
10 m 2.8 6 0.8 2.9 6 0.6 4.4 6 1.8* 3.7 6 1.2* 2.7 6 0.9 4.4 6 1.8* 4.8 6 1.0*
20 m 2.8 6 0.8 3.0 6 0.8 4.2 6 1.4* 3.4 6 1.5* 2.7 6 0.8 4.2 6 1.4* 4.9 6 1.3*
30 m 2.5 6 0.8 2.8 6 0.7 3.8 6 1.6* 3.9 6 1.4* 2.2 6 0.9 3.8 6 1.6* 5.6 6 1.5*
60 m 2.3 6 0.7
90 m 2.2 6 0.6
Rating of perceived exertion
Rest 0.0 6 0.0 0.0 6 0.0 0.0 6 0.0 0.0 6 0.0 0.0 6 0.0 0.0 6 0.0 0.0 6 0.0
10 m 2.8 6 0.8 3.2 6 1.0 3.5 6 0.7 4.1 6 0.9* 3.4 6 1.1 3.5 6 0.7 3.9 6 1.0*
20 m 3.8 6 1.2 4.0 6 1.2 4.2 6 0.9 4.3 6 0.8 3.6 6 1.1 4.2 6 0.9 4.4 6 1.1*
30 m 3.8 6 1.1 4.2 6 1.5 4.2 6 0.9 4.4 6 1.1* 3.8 6 1.1 4.2 6 0.9 5.4 6 1.2*
60 m 4.3 6 1.0
90 m 4.9 6 1.0

* p , 0.05 compared to steady state.

continuous 60-second VO2 during the exercise bout. Results

Ventilatory and respiratory compensation thresholds
were determined according to changes in the slopes of Part 1
the VCO2 vs. VO2 and VE vs. VCO2 relationships, re- The serial responses of HR, blood lactate, and RPE
spectively (3). HR was measured using radiotelemetry. through the course of the 8 exercise bouts are pre-
Subsequently, each subject was monitored during sented in Table 3. Given that the mean power output
basketball practice sessions and/or during competitive was the same at 90% IAT in all exercise bouts, the dif-
matches. The pattern of exercise during these bouts ferences in HR, blood lactate, and RPE were consistent
was determined either by the coachs plan or by the with previously established responses during interval
dictates of the competitive situation. No attempt was training and during prolonged exercise (1, 16, 23).
made to experimentally influence the pattern of exer- There was a consistent pattern for longer intervals,
cise. HR responses were recorded during each exercise more variable intervals, and longer duration steady
bout using radiotelemetry, downloaded, and analyzed state exercise bouts to be associated with greater evi-
using the summated HR zone approach as in Part 1 dence of psychophysiological strain, evidenced by HR
of the study. Thirty minutes following the conclusion and blood lactate concentrations.
of each exercise bout, the subject rated the overall dif- Comparisons of the overall exercise score between
ficulty of the bout using the session RPE method, as the summated HR zone method and the session RPE
in Part 1 of the study. method are presented in Table 4. There were signifi-
cant differences between the methods for each exercise
Statistical Analyses bout, with the session RPE method consistently giving
Statistical analysis focused on comparing the exercise a larger exercise score than the summated HR zone
scores obtained during each exercise bout using the method. However, regression analyses revealed that
session RPE and summated HR zone methods of the pattern of differences was highly consistent among
quantitating exercise training. This was accomplished the various exercise bouts (Figure 3).
using repeated-measures analysis of variance (ANO-
VA). Post hoc analyses were performed using the Tu- Part 2
key procedure. Additionally, regression analyses were The comparative exercise score using the summated
performed to relate the 2 methods of quantitating ex- HR zone method and session RPE method during bas-
ercise training. ketball practices and/or games is presented in Table
 Monitoring Exercise Training 113

Table 4. Comparison of calculated exercise TRIMP scores mated HR zone method is based on only 5 zones, so
using the summated heart rate (HR) zone method and the an athlete working at maximal HR for the entire du-
session rate of perceived exertion (RPE) method. ration of an exercise bout would only have their exer-
cise duration multiplied by 5, whereas with 10 effec-
Summated HR tive zones represented by the session RPE method the
zone Session RPE
multiplier for exercise duration can be somewhat larg-
30 min steady state 110 6 24 130 6 57*
er, particularly at high intensities. In this regard, it is
60 min steady state 216 6 39 270 6 63* worth noting that Banister et al. (2, 9, 18) used a non-
90 min steady state 350 6 44 432 6 67* linear multiplier for the mean HR recorded during ex-
30 s/30 s interval 107 6 14 131 6 45* ercise, which is conceptually quite similar to the cate-
60 s/60 s interval 117 6 18 148 6 54* gory ratio RPE score. Thus although the quality of in-
120 s/120 s interval 114 6 17 146 6 47* formation available from the session RPE method is
110% interval 114 6 16 136 6 60* fairly crude relative to the highly detailed data avail-
125% interval 117 6 18 148 6 54* able from HR records, the present data suggests that
150% interval 114 6 111 161 6 46* the same critical information is contained with both
Basketball 652 6 59 744 6 84* methods. The simplicity of the session RPE method
suggests the practical value of the technique.
* p , 0.05 summated HR zone vs. session RPE.
Our experience with the session RPE method sug-
gests that most athletes can use the technique fairly
well with only minimal instruction, primarily by fo-
cusing on the verbal anchors associated with the RPE
scale while responding to the simple question How
was your workout? Approximately 20% of athletes
will attempt to separately report RPE-duration scores
for various phases of each training session, which may
then be summated. The remaining 80% of athletes will
comfortably give a single number representing the ge-
stalt of the training session. Despite these differences,
we have found that individual athletes seem to be very
consistent in their own pattern of using the session
RPE method, and that regardless of whether they are
Figure 3. Regression lines comparing the relationship be-
tween TRIMP scores generated using the summated HR
detail-oriented or more globally focused, there is a
zone method and session RPE method of monitoring exer- good relationship between their reports of training
cise during the various cycle exercise bouts and during load and subsequent performance (12). Previous work,
basketball practice and competition. Note the overall simi- which we have done with this technique, has suggest-
larity among the different exercise bouts, and that although ed that other data (monotony, strain) may be derived
different subjects are used there is a similarity of the cy- from the session RPEderived TRIMP scores reported
cling and basketball data. by the athletes (11, 15). Although one suspects that
similar data may be derived from HR-based TRIMP
4. There were significant differences between the scores, to date this has not been accomplished.
methods, with the session RPE method giving a larger The overall consistency between objective (sum-
exercise score than the summated HR method. How- mated HR zone) and subjective (session RPE) methods
ever, regression analyses revealed that the pattern of of monitoring training during highly disparate types
differences was consistent and similar to responses of exercise suggests that the session RPE method may
during steady state and interval cycle exercise ob- be useful over a very wide variety of exercise sessions.
served in Part 1 (Figure 3). Previous experience with RPE as a method of moni-
toring exercise suggests that muscularly strong indi-
viduals are comparatively poor at rating the intensity
Discussion of aerobic exercise sessions, attending more to mus-
The results of this study are consistent with our pre- cular tension than to sensations of dyspnea. However,
vious observations of a highly correlated relationship when providing an overall gestalt with familiar modes
between the session RPE and the summated HR zone of training, it may be that even athletes performing
methods of evaluating training sessions (11, 12, 14). highly intense muscular activities can provide ade-
This suggests that either method may be used as a quate ratings. Although ultrahigh-intensity exercise
method of creating a TRIMP score for the evaluation (resistance training, plyometrics) cannot objectively be
of exercise training. The methods are not, however, in- evaluated using HR criteria, the pattern of responses
terchangeable because of differences in scale. The sum- between objective and subjective measures in the pre-
114 Foster, Florhaug, Franklin, Gottschall, Hrovatin, Parker, Doleshal, and Dodge

Table 5. Schematic training diary demonstrating the calculation of training load, monotony, and strain.

Session rate of
perceived
exertion Duration
Day Training activity (RPE) (min) Load

Sunday Cycle 100 km 5* 180 940

Monday Jog 5 km 1 extensive stretch 2* 25 50
Tuesday Skate 6 3 10 min at AT pace/5 min rec 7* 120 840
Explosive weights 1 abs 7* 40 280
Wednesday Cycle 30 km 3* 60 180
Thursday Skate 10 3 3 min at 5 km pace/5-min rec 8* 75 390
Friday Jog 5 km 1 extensive stretch 2* 25 50
Saturday Skate 20 3 1 min at tempo/2-min rec 8* 75 390
Explosive weights 1 abs 7* 40 280
Weekly load 3400
Monotony (3 SD) 1.26
Strain (load 3 monotony) 4284

ing; 11, 12). Finally, the session RPE method has the
advantage of not requiring knowledge of maximal ex-
ercise responses (e.g., HRpeak) to anchor the monitor-
ing method. Although in athletic individuals deter-
mination of maximal HR is relatively risk-free, it still
represents an additional step in designing a training
monitoring scheme that is not required by the session
RPE method.

Practical Applications
The present data provide support for the use of the
session RPE method as a subjective estimate of train-
ing load during nonsteady state exercise, including
Figure 4. Schematic training periodization plan over the very high-intensity interval training and team sport
17 weeks leading up to a major weekend of competition. practice and competition. As such, it suggests that this
Note the day to day variation in the training load and the very simple method may be a useful technique for
weekly variation in training. With this scheme of monitor- quantitating training load in a wide variety of athletic
ing training, the coach can appreciate how well the athlete applications. In this regard, the present data suggest
executes the designed periodization plan. In particular, the that the session RPE method may provide a mecha-
failure of the athlete to progressively increment the nism for quantitating the exercise intensity component
heavy weeks of the training plan is very obvious and and allow calculation of a single number representa-
explains the less than satisfactory results at the time of the
tive of the combined intensity and duration of training
competitions.
sessions.
The present data and our previous experience with
sent data (where the power output of the 650% er- the session RPE technique suggest that it is easy to use,
gometer trials and during basketball practice occasion- quite reliable, and consistent with objective physiolog-
ally exceeded the peak aerobic power output) suggests ical indices of the intensity of exercise training. By sim-
that the session RPE method might be a valid ap- ply asking the athlete to rate the global intensity of the
proach to evaluating even very high-intensity exercise. exercise bout and then by multiplying by the duration
If this were so, then a single method could be used to of the training bout, a daily exercise score can be cre-
provide a quantitative basis for describing the peri- ated. This can be put into the form of an exercise diary,
odization of training plans. Some support for this sug- which can reveal the overall weekly pattern of exercise
gestion is provided by the relationship between train- (Table 5). From this, accessory indices of training, such
ing load and performance, which we have previously as monotony and strain, can be calculated, potentially
demonstrated with speed skaters (who do a wide va- providing in index of the likelihood of untoward train-
riety of aerobic, interval, and ultrahigh-intensity train- ing outcomes. Finally, the daily and weekly training
 Monitoring Exercise Training 115

loads calculated using this technique can be presented 12. FOSTER, C., E. DANIES, L. HECTOR, A. SNYDER, AND R. WELSH.
Athletic performance in relation to training load. Wis. Med. J.
graphically, allowing the coach to have a visual im-
95:370374. 1996.
pression of the periodization plan as experienced by 13. FOSTER, C., J. DANIELS, AND R. YARBROUGH. Physiological and
the athlete (Figure 4). training correlates of marathon running performance. Aust. J.
Sports Med. 9:5861. 1977.
14. FOSTER, C., L. HECTOR, R. WELSH, M. SCHRAGER, M.A. GREEN,
References AND A.C. SNYDER. Effects of specific versus cross training on
running performance. Eur. J. Appl. Physiol. 70:367372. 1995.
1. BANGSBO, J. Physiology of intermittent exercise. In: Exercise and
15. FOSTER, C., AND M. LEHMANN. Overtraining syndrome. In:
Sport Science W.E. Garrett and D.T. Kirkendall, eds. Philadel- Running Injuries. G. Guten, ed. Orlando: W.B. Saunders, Co.,
phia: Lippincott, Williams and Wilkins, 2000. pp. 3152. 1997. pp. 173188.
2. BANISTER, E.W., T.W. CLAVERT, M.V. SAVAGE, AND T. BACH. A 16. GOROSTIAGA, E.M., C.B. WALTER, C. FOSTER, AND R.C. HICK-
systems model of training for athletic performance. Aust. J. SON. Uniqueness of interval and continuous training at the
Sports Med. 7:5761. 1975. same maintained exercise intensity. Eur. J. Appl. Physiol. 63:101
3. BEAVER, W.L., K. WASSERMAN, AND B.J. WHIPP. A new method 107. 1991.
for detecting the anaerobic threshold by gas exchange. J. Appl. 17. LINDSAY, F.H., J.A. HAWLEY, K.H. MYBURGH, H.H. SCHOMEN,
Physiol. 60:20202027. 1986. T.D. NOAKES, AND S.C. DENNIS. Improved athletic performance
4. BORG, G., P. HASSMEN, AND M. LANGERSTROM. Perceived ex- in highly trained cyclists after interval training. Med. Sci. Sports
ertion in relation to heart rate and blood lactate during arm Exerc. 28:14271434. 1996.
and leg exercise. Eur. J. Appl. Physiol. 65:679685. 1985. 18. MORTON, R.H., J.R. FITZ-CLARKE, AND E.W. BANISTER. Model-
5. BROWN, D.D. Pulmonary responses to exercise and training. In: ing human performance in running. J. Appl. Physiol. 69:1171
Exercise and Sport Science. W.E. Garrett and D.T. Kirkendall, eds. 1177. 1990.
Philadelphia: Lippincott, Williams and Wilkins, 2000. pp. 117 19. MUJIKA, I. The influence of training characteristics and tapering
134. on the adaptations in highly trained individuals. A review. Int.
6. BUSSO, T., C. DENIS, R. BONNEFOY, A. GEYSSANT, AND J.R. LA- J. Sports Med. 19:439436. 1996.
COUR. Modeling of adaptations to physical training by using a 20. MUJIKA, I., T. BUSSO, L. LACOSTE, A.G. BARALE, AND J.C. CHA-
recursive least squares algorithm. J. Appl. Physiol. 82:16851693. TARD. Modeled responses to training and taper in competitive
1997. swimmers. Med. Sci. Sports Exerc. 28:252258. 1996.
7. COOPER, K.H. Aerobics. New York: Bantam, 1968. 21. ROWBOTTOM, D. Periodization of training. In: Exercise and Sport
8. EDWARDS, S. The Heart Rate Monitor Book. Sacramento: Fleet Feet Science. W.E. Garret and D.T. Kirkendall, eds. Philadelphia: Lip-
Press, 1993. pp. 113129. pincott, Williams and Wilkins, 2000. pp. 499512.
9. FITZ-CLARAKE, J.R., R.H. MORTON, AND E.W. BANISTER. Opti- 22. STEGMANN, H., W. KINDERMANN, AND A. SCHNABEL. Lactate
mizing athletic performance by influence curves. J. Appl. Phy- kinetics and individual anaerobic threshold. Int. J. Sports Med.
siol. 71:11511158. 1991. 2:160165. 1981.
10. FLECK, S.J. Periodized strength training: A critical review. J. 23. VUORIMAA, T., T. VASANKARI, AND H. RUSKO. Comparison of
Strength Cond. Res. 13:8289. 1999. physiological strain and muscular performance of athletes dur-
11. FOSTER, C. Monitoring training in athletes with reference to over- ing 2 intermittent running exercises at the velocity associated
training syndrome. Med. Sci. Sports Exerc. 30:11641168. 1997. with VO2max. Int. J. Sports Med. 21:96101. 2000.