International Journal of Sports Physiology and Performance, 2017, 12, 796  

http://dx.doi.org/10.1123/ijspp.2016-0159
© 2017 Human Kinetics, Inc. ORIGINAL INVESTIGATION

Quantifying Training Loads in Contemporary Dance

Annie C. Jeffries, Lee Wallace, and Aaron J. Coutts

Purpose: To describe the training demands of contemporary dance and determine the validity of using the session rating of
perceived exertion (sRPE) to monitor exercise intensity and training load in this activity. In addition, the authors examined the
contribution of training (ie, accelerometry and heart rate) and non-training-related factors (ie, sleep and wellness) to perceived
exertion during dance training. Methods: Training load and ActiGraphy for 16 elite amateur contemporary dancers were collected
during a 49-d period, using heart-rate monitors, accelerometry, and sRPE. Within-individual correlation analysis was used to
determine relationships between sRPE and several other measures of training intensity and load. Stepwise multiple regressions
were used to determine a predictive equation to estimate sRPE during dance training. Results: Average weekly training load
was 4283 ± 2442 arbitrary units (AU), monotony 2.13 ± 0.92 AU, strain 10677 ± 9438 AU, and average weekly vector magni-
tude load 1809,707 ± 1015,402 AU. There were large to very large within-individual correlations between training-load sRPE
and various other internal and external measures of intensity and load. The stepwise multiple-regression analysis also revealed
that 49.7% of the adjusted variance in training-load sRPE was explained by peak heart rate, metabolic equivalents, soreness,
motivation, and sleep quality (y = –4.637 + 13.817%HRpeak + 0.316 METS + 0.100 soreness + 0.116 motivation – 0.204 sleep
quality). Conclusion: The current findings demonstrate the validity of the sRPE method for quantifying training load in dance,
that dancers undertake very high training loads, and a combination of training and nontraining factors contribute to perceived
exertion in dance training.

Keywords: session RPE, microtechnology, monitoring training

Physical training provides the stimulus for performance
improvements and can be described in terms of process and out-
come.1 The training process includes the systematic repetition of
physical exercises and is usually described as a measure of external
load (ie, the work completed by the athlete).2 Training outcomes are
the consequence of the internal load, which is dependent on both
individual characteristics and the nature and quantum of the external
load.1 It is therefore recommended that coaches monitor the specific
training dose applied to an athlete to optimize training adaptations.2
Dance can be described as containing high-intensity, intermit-
tent bouts of activity, using both aerobic and anaerobic energy con-
tributions.3 Professional dancers routinely undertake high training
volumes, leading to increased levels of fatigue, impaired perfor-
mance, and high injury rates.4 Previous research has investigated
methods for measuring intensity in dance, including heart rate (HR),
oxygen consumption, and blood lactate.5 However, due to practical
limitations associated with these methods, such as expense and the
requirement for technical expertise,2 monitoring training load (TL)
in dancers remains a relatively uncommon practice.
Session rating of perceived exertion (sRPE) was proposed as a
simple, noninvasive, and inexpensive method for assessing internal
TL. sRPE can be determined as the product of training duration
(min) and a rating-of-perceived-exertion (RPE) score according to
the Borg Category Ratio 10-point scale (CR-10).2 This method is
now accepted as a valid gestalt measure of TL across a wide variety
of sports and exercise modalities.2,6,7 However, despite widespread
validation in many sports, to the best of our knowledge, no studies
to date have examined the validity of the sRPE method in dance.
The authors are with the Sport and Exercise Discipline Group, University of
Technology Sydney (UTS), Moore Park, NSW, Australia. Jeffries (Annie.
Jeffries@uts.edu.au) is corresponding author.
Moreover, no studies have assessed the relationship between
external TL and perceived exertion, together with training and
nontraining factors that contribute to perceived exertion in dance.
Accordingly, the aims of this study were to describe the training
demands of contemporary dance and to determine the validity of
using sRPE to monitor internal TL for this activity. In addition, we
examined the contribution of training- and non-training-related
factors to perceived exertion during dance training.
Methods
Participants
Sixteen elite amateur contemporary dancers, 5 male (age 18.8 ± 2.4
y, height 178 ± 0.08 cm, body mass 70.2 ± 5.9 kg, dance experience
9.4 ± 4.4 y) and 11 female (age 19.0 ± 2.0 y, height 165 ± 0.05 cm,
body mass 57.5 ± 4.9 kg, dance experience 14.1 ± 3.2 y), participated
in this study. Before the study, dancers completed a 2-week rest
period to allow recovery from previous workloads. Dancers were
familiarized with all procedures before the commencement of the
study. Written consent forms were obtained from each subject, in
addition to ethical approval being granted by the institutional human
research ethics committee before the commencement of the study.
Study Design
This study used a prospective longitudinal research design where
TL data were collected during a 49-day period. TL measures were
assessed via HR, waist accelerometers, and perception of effort
(sRPE) during each dance class and rehearsal. Both HRrest and
HRpeak were determined for each dancer before the commencement
of the study to allow determination of individual HR training zones
for the calculation of internal TL. Resting HR was calculated from

796

the lowest HR observed for each dancer while at rest using Polar
HR straps (Team 2, Polar, Oy, Finland). Maximum HR was calcu-
lated from the highest HR achieved during the Yo-Yo Intermittent
Recovery Test Level 1 (YIRT) conducted according to previously
described methods.8
The dance training was prescribed and implemented by both
the director of the school and visiting choreographers during the
study period. A typical week consisted of 3 or 4 daily dance classes
or rehearsals. All sessions could be described as low to moderate
intensity with brief periods of high-intensity exercise and were
broadly classified into the following training modes: ballet and
contemporary class and contemporary rehearsals.
To monitor the psychophysiological recovery process between
days of dance training, each participant was required to complete
a psychometric questionnaire before the commencement of class
each day. The questionnaire, developed from previous recommenda-
Downloaded by Ebsco Publishing aellsworth@ebsco.com on 08/24/17, Volume 12, Article Number 6
tions,9,10 was used to assess general indicators of dancer wellness.
The custom-made questionnaire comprised 5 questions related to
perceived fatigue, general muscle soreness, pain, motivation, and
overall recovery, with each question scored on a 10-point scale
(scores 1–10, with 1 representing poor and 10 representing very
good).
To monitor sleep, dancers wore the ActiGraph (GTX3, Acti-
Graph, Pensacola, FL, USA) on their nondominant wrist during
sleep. In accordance with manufacturer guidelines, epochs of 1
minute were taken,11 while the frequency was set at 100 Hz. The
ActiGraph data were downloaded and analyzed using the Actilife
software (ActiGraph 2015, Actilife 6.12, Pensacola, FL, USA).
Sleep variables that were recorded included total sleep time, sleep
efficiency %, total time in bed (min), wake after sleep onset, and
number of awakenings (n).
The internal TL for each session was calculated using the
sRPE method for each dancer during the study period.6 sRPE was
calculated as the product of exercise intensity and exercise dura-
tion. Exercise intensity was determined by sRPE according to the
Borg CR-10 scale12 and was collected within 10 minutes after each
class.13 Training monotony and strain were also calculated based
on the method of Foster.14 Monotony was calculated by dividing
the mean daily sRPE TL over each week by the standard deviation.
Strain was determined as the product of accumulated weekly TL
and monotony. Monotony is a measure of the variability of training,
while strain reflects the overall training stress.14 In addition, sRPE
sessions were classified as low (sRPE <3 AU), moderate (sRPE 3–5
AU), and high (RPE >5 AU) intensity. These zones were based on
the methods previously reported by Lovell et al15 and then adjusted
to reflect the nature of dance training and the verbal anchors of the
CR-10 scale.
HR was collected during each training session every 5 seconds,
using polar HR transmitters located on the anterior chest wall, as a
further measure of exercise intensity (Polar Team2, Oy, Finland).
HR has been shown to be a valid and reliable measure of exercise
intensity.16 The HR data in this study were stored in the transmitter
and downloaded and analyzed on a weekly basis using Polar Team2
software version 1.4.5, Oy, Finland.
Two HR-based TRIMP measures were used to determine
internal TL. First, the TRIMP based on the method of Banister17
was calculated according to training duration, HRmax, HRrest, and
HRex and was expressed in the following formula: TL = D (ΔHR
ratio) eb(ΔHR ratio), where D = duration of training session, ΔHR ratio
is average change in HRreserve, and b = 1.67 for females and 1.92 for
males.18 The second TRIMP was based on the method of Edwards19
but was modified to include 3 HR zones.20 The internal TL was
IJSPP Vol. 12, No. 6, 2017
Monitoring Dance Training   797
calculated from the time spent in each HR zone multiplied by the
corresponding zone coefficient (ie, <70% of HRmax = 1, 70–85% of
HRmax = 2, and >85% of HRmax = 3) followed by a summation of
the results. This method has been validated as an accurate measure
of intensity and load during intense, prolonged exercise.20
External load of body load/vector magnitude, metabolic equiva-
lent of task (METs), and estimated energy expenditure (EEE) were
measured during each class and rehearsal using a GTX3 ActiGraph
accelerometer (ActiGraph, Pensacola, FL, USA). The ActiGraph
was worn at the waist on the nondominant side using an elastic belt.
Sampling rate was selected at 1 second and frequency at 100 Hz.
ActiGraph uses a triaxial accelerometer to collect motion data in 3
axes—vertical (y), horizontal right-left (x), and horizontal front-back
(z)—and includes the vector summed value (x2 +y2 + z2)–2, known
as vector magnitude. This measure provides an indication of total
stress resulting from accelerations, decelerations, and changes in
direction and is calculated automatically using a custom algorithm
included in the proprietary software provided by the manufacturers.
The ActiGraph data were downloaded and analyzed using Actilife
software, which has previously been validated for the measurement
of physical activity.21
Statistical Analyses
The assumptions of normality were verified before parametric statis-
tical analysis. A multivariate analysis of variance (MANOVA) with
Scheffé post hoc was used to investigate the differences in weekly
TL, monotony, and strain. Within-individual correlations between
sRPE-TL and HR and accelerometry measures of exercise intensity
and load were analyzed using the Pearson correlation coefficient.
The following criteria were employed to interpret the magnitude of
the correlation between test measures: <.10 trivial, .10 to .29 small,
.30 to .49 moderate, .50 to .69 large, .7 to .89 very large, and .90
to 1.0 almost perfect.22 Differences between the within-individual
correlations between each of the methods were assessed using a
1-way ANOVA for repeated measures with Scheffé post hoc to
locate differences. A stepwise multiple regression was used to
determine a predictive equation to estimate sRPE of dance train-
ing from %HRmax, METs, soreness, motivation, and sleep quality.
Effect size was determined using Cohen effect-size statistic.23 Effect
size was assessed using the following criteria: <0.20 trivial, 0.20 to
0.59 small, 0.60 to 1.19 moderate, 1.20 to 1.99 large and ≥2.0 very
large.24 Standard statistical methods were used for the calculation
of means, SDs, and Pearson product–moment correlation coef-
ficients. The multiple regression, ANOVA analysis, and effect-size
calculation were conducted using Statistica 12 (Statsoft Inc, Tulsa,
OK, USA). Statistical significance was set at P < .05. The data are
presented as mean, SD, or 90% confidence interval (CI) as stated.
Results
A total of 954 individual training sessions were completed over
49 days, with the dancers completing 59 ± 16 sessions. The mean
accumulated weekly TL over the 49 days of the study was 4283
± 2442 AU, monotony 2.13 ± 0.92 AU, strain 10,677 ± 9438 AU,
and weekly vector magnitude load 1809,707 ± 1015,402 AU. The
mean session duration was 118 ± 45 minutes, and the mean session
TLs expressed as sRPE-TL, vector magnitude, and EEE were 476 ±
312 AU, 82 ± 56 AU, and 76 ± 94 kcal, respectively. The Bannister
TRIMP and modified TRIMP TLs were 58 ± 37 AU and 130 ± 50
AU. The mean session intensity expressed as RPE and HR were
4.1 ± 1.9 AU and 53.2% ± 8.0%HRpeak, respectively. The TL and
 798  Jeffries, Wallace, and Coutts

intensity measures for the different types of dance training sessions
are shown in Table 1. The dancers slept for 437 ± 69 minutes each
night, with sleep efficiency, total time in bed, wake after sleep onset,
and number of awakenings being 89.1% ± 9.5%, 485 ± 66 minutes,
47.9 ± 27.1, and 16.5 ± 7.9, respectively.
When classified according to HR response, 97.6% (n = 931),
2.3% (n = 22), and 0.1% (n = 1) of all sessions were classified as
low (<70% HRpeak), moderate (70–85% HRpeak), and high (>85%
HRpeak), respectively. In contrast, when categorized according to
sRPE, 40.8% (n = 384), 33.2% (n = 327), and 26% (n = 248) of all
sessions were classified as low (sRPE <3 AU), moderate (sRPE 3
to 5 AU), and high (RPE >5 AU) intensity.
Figure 1 shows changes in the weekly training-dose measures.
MANOVA showed significant main effects (F24,336.1 =1.958, P
= .005) for changes in weekly TL. Univariate analysis revealed
significant differences for vector magnitude, training strain, and
Downloaded by Ebsco Publishing aellsworth@ebsco.com on 08/24/17, Volume 12, Article Number 6
Discussion
The main purpose of this study was to examine the validity of sRPE
for monitoring TL in contemporary dance and to describe the TLs
experienced by high-level contemporary dancers. In addition, the
internal, external, sleep, and perceptual factors that contribute to
perceived exertion in dance were examined. The key finding of this
study was the significant within-individual correlations between
sRPE-TL and various other internal and external measures of load.
The results also showed that a combination of internal, external,
perceptual, and sleep factors explains more variance of sRPE-TL
in dance more accurately than any individual measure alone. These
results demonstrate the validity of sRPE-TL as a measure of TL
in dance training.
The current study showed that contemporary dancers undertake
very high TLs, especially in comparison with many other high-level

sRPE-TL. Subsequent post hoc analyses showed greater vector
magnitude load in week 3 than in weeks 1, 2, and 5. Similarly, strain
was greater in week 3 than in weeks 1 and 5, while sRPE-TL was
greater in week 3 than in week 5.
Within-individual correlations between the RPE-based mea-
sures of intensity and load and various other training-intensity and
load measures are presented in Table 2. A 1-way repeated-measures
ANOVA showed significant differences for the strength of the
within-individual correlations between sRPE and the different
constructs of load (F6,105 = 3.775, P = .002; ES 0.33), Scheffé post
hoc revealed higher correlations with Bannister TRIMP compared
with accelerometer counts and EEE. Group correlations with vari-
ous measures of load and intensity across training modes are shown
in Table 3.
The stepwise multiple-regression analysis revealed that
49.7% of the adjusted variance in sRPE-TL could be explained
by %HRpeak, METs, soreness, motivation, and sleep quality (y =
–4.637 + 13.817%HRpeak + 0.316 METS + 0.100 soreness + 0.116
motivation – 0.204 sleep quality; adjusted R2 = .497; F 79.883; P
= .012). The collinearity statistics were acceptable with tolerance
levels at .954, .979, .916, .919, and .958.
athletes.10,25 The mean accumulated weekly TLs (4283 ± 2442 AU)
were much higher than those reported in adolescent ballet dancers
(596 ± 153 AU)26 and professional Australian football players (rang-
ing from 2027 AU preseason to 1651 AU in-season).25 The observa-
tion period was representative of a typical training period that would
be completed in preparation for a major performance. It was also
notable that the monotony and strain values were much greater than
previous reports from adolescent ballet dancers (monotony 0.9 ± 0.2
AU and strain 4365 ± 1810 AU).26 High levels of TL, monotony, and
strain have been identified as factors related to negative adaptations
to training.14 For example, training-monotony values of (>2.0 AU)
are considered to be of great concern for athletes27 as monotonous
training combined with high TLs may lead to the onset of overtrain-
ing.14 Furthermore, high training strain has been associated with
the incidence of illness and poor performance.28
High TLs have been related with increased risk of injury and
illness.25 The present results show that contemporary dancers
undertake high TLs, with relatively little variation in daily stimulus.
This approach to training might explain high injury rates previously
reported in dancers29 and provide support for the implementation
of monitoring systems in high-level dancers.

Table 1  Load and Intensity Measures for All Training Sessions (Mean ± SD)
Measure Contemporary (n = 142) Rehearsal (n = 559) Ballet (n = 240)
Duration (min) 91.5 ± 19.7b 137.5 ± 45.7a,c 90.3 ± 23.6b
s-RPE TL (AU) 323 ± 210b,c 545 ± 361a,c 417 ± 173a,b
Heart rate (%HRpeak) 0.51 ± 0.08c 0.52 ± 0.07c 0.56 ± 0.07b
s-RPE (AU) 3.4 ± 1.7c 3.9 ± 2.0c 4.6 ± 1.6a,b
<70% HRpeak 84.5 ± 19.7b 127.1 ± 44a,c 78.0 ± 25b
70–85% HRpeak 5.9 ± 9.5c 8.5 ± 10.1c 11.1 ± 10.6a,b
85–100% HRpeak 0.8 ± 2.5b 1.9 ± 4a 1.4 ± 2.9
Bannister TRIMP (AU) 41 ± 29b,c 64 ± 42a,c 54 ± 26a,b
Edwards TRIMP (AU) 99 ± 26b 150 ± 53a,c 104 ± 29b
METs 3.0 ± 1.1b 3.9 ± 2a,c 2.8 ± 1.1b
Estimated energy expenditure (kcal) 68 ± 58 80 ± 108 68 ± 71
Vector-magnitude counts (mean) 195,788 ± 131,959 217,578 ± 153,143c 170,198 ± 100,779a
Vector-magnitude average counts 82 ± 51c 89 ± 63c 64 ± 36a,b
Abbreviations: sRPE, session rating of perceived exertion; TL, training load; AU, arbitrary units; HR, heart rate; TRIMP, training impulse; METs, metabolic equivalents.
a Significantly different from contemporary, b significantly different from rehearsal, c significantly different from ballet.

IJSPP Vol. 12, No. 6, 2017


Downloaded by Ebsco Publishing aellsworth@ebsco.com on 08/24/17, Volume 12, Article Number 6
Figure 1 — Mean weekly measures of vector magnitude, session rating
of perceived exertion, monotony, and training strain.
In agreement with previous research from a variety of sports
involving endurance, steady-state, and intermittent exercise,2,6 the
current study reported large to very large correspondence between
sRPE and other measures of internal TL (r = .69–.77). These results
were similar to those described in a study of gymnastics, involv-
ing complex motor skills and technical difficulty, with very large
correlations reported (r = .92–.96).30 A new finding from this study
was the large to very large within-individual correlations between
sRPE and various accelerometer-derived measures of external
load. Specifically, the estimated MET equivalents derived from
accelerometry showed a very large relationship with sRPE (r = .71),
while large correlations were found for vector magnitude counts (r
= .63), vector magnitude average (r = .66), and EEE (r = .63). The
vector magnitude, which is a composite of counts from 3 planes
of motion, was used to estimate body load.31 It is reported by the
manufacturer to provide an indication of total stress resulting from
accelerations, decelerations, and changes of direction. While no
known previous studies have examined the relationships between
IJSPP Vol. 12, No. 6, 2017
Monitoring Dance Training   799
sRPE and constructs of external load in dance, similar findings have
been reported from Australian football training activities, where
large correlations (r = .86) were found for accelerometer-derived
measures of body load (ie, vector magnitude).32 The slightly lower
correlations between these variables observed in the current study
may be explained by the more intermittent nature of the activity
profile in dance. Indeed, dance involves highly complex physical
activities that are interspersed with long periods of low-intensity
activity.3 For example, an adage section of a dance routine, which is
common in most ballet and contemporary training sessions, requires
controlled movements of precision and skill, followed by explosive
bursts of actions such as a series of jumps.
Although the collective results from all training sessions
demonstrated sRPE to be a valid global measure of internal TL,
an assessment of the relationships between the sRPE and the other
load constructs for each of the training modes provides a more
detailed understanding of the factors that influence sRPE in dance
training. The group correlations obtained for each of the training
modes (Table 3) demonstrated moderate to very large correlations
between sRPE-TL and various measures of TL. While rehearsal and
contemporary sessions demonstrated moderate to very large cor-
relations between sRPE and the various load constructs (rehearsal
r = .44–.73; contemporary r = .52–.72), ballet revealed poorer
relationships (r = .32–.58). The lower correlations reported in ballet
may be related to a combination of factors associated with the
unique nature of this dance genre. For example, ballet consists of
complex technical movements involving both artistic and aesthetic
elements.26 This is thought to place increased stress on the anaerobic
energy system, with more jumps and changes of direction, longer
rest periods, and a strong focus on eccentric muscle actions than
used in contemporary dance.33 In addition, the dancers involved in
this study were predominantly trained in contemporary dance, with
varying degrees of exposure to ballet training. Indeed, comparable
results from resistance- and sprint-training studies have shown that
similar factors affect sRPE.34
The present results also demonstrated that a combination
of load and other measures explains the variance in sRPE-TL
better than any individual measures alone. For example, the step-
wise multiple regression for sRPE-TL showed that measures of
%HRpeak, METs, soreness, motivation, and sleep quality accounted
for 49.7% of the variance in dance TL. While no previous studies
have examined how these specific non-training-related factors
affect sRPE (ie, sleep, soreness, motivation), others have examined
the effect of other non-training- and training-related factors.15,32
Collectively, these results suggest that a myriad of training and
nontraining factors contribute to sRPE-TL. Moreover, the present
results show that contributing factors to the perception of effort
in dancers may include fatigue, sleep quality, and muscle sore-
ness. It should be noted that in this study there were 16 subjects
per determinant variable for the prediction equation, which is a
low number for multivariate regression analyses. Accordingly, we
suggest that the present results be interpreted in light of this limita-
tion. Notwithstanding this, the results of this study further extend
our understanding that not only internal load but also nontraining
and external factors affect RPE, further demonstrating that it is a
global indicator of internal load.
To our knowledge, no previous research has compared the dis-
tribution of training according to intensity using various measures
in dance. The current results show different distribution of training
intensities when sessions are categorized into low-, moderate-, and
high-intensity zones by both %HRpeak and RPE. For example, in a
dance class, %HRpeak may be low while sRPE may be high. While
 800  Jeffries, Wallace, and Coutts

Table 2  Within-Individual Correlations Between sRPE-TL and Various Measures of Load and Intensity (Mean ±
SD)
Participant %HRpeak Bannisters TRIMP Edwards TRIMP METs Accelerometer counts Vector magnitude EEE
1 .69 .73 .69 .44 .52 .47 .50
2 .52 .60 .86 .74 .69 .75 .69
3 .59 .72 .79 .80 .74 .72 .67
4 .65 .70 .57 .63 .63 .67 .66
5 .72 .76 .72 .70 .68 .68 .61
6 .75 .90 .79 .71 .79 .73 .60
7 .76 .75 .62 .79 .79 .76 .78
8 .55 .81 .79 .75 .58 .58 .59
9 .49 .80 .90 .80 .40 .47 .45
Downloaded by Ebsco Publishing aellsworth@ebsco.com on 08/24/17, Volume 12, Article Number 6

10 .64 .69 .74 .61 .58 .60 .50

11 .81 .77 .82 .64 .47 .61 .58
12 .80 .85 .73 .83 .82 .83 .83
13 .69 .76 .68 .64 .55 .69 .72
14 .79 .84 .85 .76 .61 .59 .63
15 .72 .74 .68 .75 .69 .69 .60
16 .87 .88 .36 .73 .58 .68 .62
Mean ± SD .69 ± .11 .77 ± .08bc .72 ± .13 .71 ± .10 .63 ± .12a .66 ± .10 .63 ± .10a
Maximum .87 .90 .90 .83 .82 .83 .83
Minimum .49 .60 .36 .44 .40 .47 .45
Abbreviations: sRPE, session rating of perceived exertion; TL, training load; HR, heart rate; TRIMP, training impulse; METs, metabolic equivalents; EEE, estimated
energy expenditure.
a Significantly different from Bannister TRIMP, b significantly different from accelerometer count, c significantly different from EEE, all P < .05.

Table 3  Group Correlations for sRPE-TL With Various Measures of Load and
Intensity (Mean ± SD) Across Training Modes
Ballet Contemporary Rehearsal
Dancers’ sessions 16 (240 ± 18.2) 16 (142 ± 15.5) 16 (559 ± 21.6)
sRPE-TL
  Bannister TRIMP .51 .59 .73
  Edwards TRIMP .58 .72 .67
 METs .39 .53 .63
  Vector-magnitude counts .40 .56 .64
  Vector-magnitude average .41 .56 .64
 EEE .32 .52 .44
sRPE
 %HRpeak .49 .59 .68
Abbreviations: sRPE, session rating of perceived exertion; TL, training load; TRIMP, training impulse; METs, metabolic
equivalents; EEE, estimated energy expenditure; HR, heart rate.

previous studies have employed individual delineation points for
low, moderate, and high intensities,15 it is possible that these arbi-
trary cutoff points for sRPE and HR may not have true equivalence
(eg, exercise at 85% HRmax may not always correspond to an RPE
of 5). The variation in the distribution of training intensities may be
attributed in part to the differentiation in delineation points between
these methods. Such variations in training intensities have also been
found in rugby league,15 suggesting that different monitoring meth-
ods produce anomalous information regarding the stimulus applied
to individuals during training. Therefore it appears that increased
effort during dance training may not be accurately assessed using
HR or accelerometry measures independently, further supporting

IJSPP Vol. 12, No. 6, 2017


the use of sRPE in evaluating global TL in dance. Moreover, these
findings demonstrate that when comparing TL between studies or
within a dance setting, consideration must be given to the effect of
employing different monitoring methods.
In summary, the present findings provide additional evidence
of the efficacy of sRPE as a measure of global TL. Furthermore,
the current study is the first to validate sRPE in dance. The study
also observed high weekly TLs, monotony, and strain. Furthermore,
much of the variance of sRPE-TL was found to relate to a combina-
tion of factors including %HRpeak, METs, soreness, motivation, and
sleep quality. Since TL measures involving HR and accelerometry
incur costs, require monitors, involve expertise, and are time-
consuming, we suggest that the sRPE method provides a simple,
versatile, and valid alternative method for dance teachers to monitor
and optimize dance training.
Downloaded by Ebsco Publishing aellsworth@ebsco.com on 08/24/17, Volume 12, Article Number 6
Practical Applications
Currently, there are few valid and practical methods for monitoring
TL in intermittent exercise such as dance. The sRPE method pro-
vides a simple, low-cost, and noninvasive method for monitoring TL
and intensity in dance. This method provides teachers and coaches
a means to evaluate and compare the training stress imposed on
an individual dancer, particularly the internal training dose. Fur-
thermore, the use of the sRPE method may allow dance teachers
to monitor TLs in individual dancers and introduce periodization
strategies, particularly as high levels of TL, monotony, and strain
were observed in this study.
Conclusion
sRPE is a valid method for quantifying internal TL in dance. It was
also observed that dancers experience very high TLs, monotony,
and strain. The study found that a combination of internal, external,
perceptual, and sleep factors predicts sRPE-TL in dance training
better than any individual measure alone. Collectively, these find-
ings provide new evidence to support the use of sRPE as a global
measure of TL in dance.
References
1. Impellizzeri FM, Rampinini E, Marcora SM. Physiological assess-
ment of aerobic training in soccer. J Sports Sci. 2005;23(6):583–592.
PubMed doi:10.1080/02640410400021278
2. Wallace LK, Slattery KM, Impellizzeri FM, Coutts AJ. Establishing the
criterion validity and reliability of common methods for quantifying
training load. J Strength Cond Res. 2014;28(8):2330–2337. PubMed
doi:10.1519/JSC.0000000000000416
3. Wyon M. Cardiorespiratory training for dancers. J Dance Med Sci.
2005;9(1):7–12.
4. Wyon M. Preparing to perform: periodization and dance. J Dance
Med Sci. 2010;14(2):67–72. PubMed
5. Schantz PG, Astrand PO. Physiological characteristics of classi-
cal ballet. Med Sci Sports Exerc. 1984;16(5):472–476. PubMed
doi:10.1249/00005768-198410000-00009
6. Foster C, Florhaug JA, Franklin J, et al. A new approach to monitoring
exercise training. J Strength Cond Res. 2001;15(1):109–115. PubMed
7. Minganti C, Capranica L, Meeusen R, Piacentini M. The use of ses-
sion-RPE method for quantifying training load in diving. Int J Sports
Physiol Perform. 2011;6:408–418. PubMed doi:10.1123/ijspp.6.3.408
IJSPP Vol. 12, No. 6, 2017
Monitoring Dance Training   801
8. Bangsbo J, Iaia FM, Krustrup P. The Yo-Yo Intermittent Recov-
ery Test: a useful tool for evaluation of physical performance
in intermittent sports. Sports Med. 2008;38(1):37–51. PubMed
doi:10.2165/00007256-200838010-00004
9. Hooper SL, Mackinnon LT. Monitoring overtraining in athletes:
recommendations. Sports Med. 1995;20(5):321–327. PubMed
doi:10.2165/00007256-199520050-00003
10. McLean BD, Coutts AJ, Kelly V, McGuigan MR, Cormack SJ.
Neuromuscular, endocrine, and perceptual fatigue responses during
different length between-match microcycles in professional rugby
league players. Int J Sports Physiol Perform. 2010;5(3):367–383.
PubMed doi:10.1123/ijspp.5.3.367
11. Sadeh A, Sharkey KM, Carskadon MA. Activity-based sleep-wake
identification: an empirical test of methodological issues. Sleep.
1994;17(3):201–207. PubMed doi:10.1093/sleep/17.3.201
12. Borg G, Ljunggren G, Ceci R. The increase of perceived exer-
tion, aches and pain in the legs, heart rate and blood lactate during
exercise on a bicycle ergometer. Eur J Appl Physiol Occup Physiol.
1985;54(4):343–349. PubMed doi:10.1007/BF02337176
13. Uchida MC, Teixeira LF, Godoi VJ, et al. Does the timing of measure-
ment alter session-RPE in boxers? J Sports Sci Med. 2014;13(1):59–
65. PubMed
14. Foster C. Monitoring training in athletes with reference to overtraining
syndrome. Med Sci Sports Exerc. 1998;30(7):1164–1168. PubMed
doi:10.1097/00005768-199807000-00023
15. Lovell TW, Sirotic AC, Impellizzeri FM, Coutts AJ. Factors affect-
ing perception of effort (session rating of perceived exertion) during
rugby league training. Int J Sports Physiol Perform. 2013;8(1):62–69.
PubMed doi:10.1123/ijspp.8.1.62
16. Esposito F, Impellizzeri FM, Margonato V, Vanni R, Pizzini G, Veic-
steinas A. Validity of heart rate as an indicator of aerobic demand
during soccer activities in amateur soccer players. Eur J Appl Physiol.
2004;93(1-2):167–172. PubMed doi:10.1007/s00421-004-1192-4
17. Banister EW. Modeling Elite Athletic Performance. Champaign, IL:
Human Kinetics; 1991.
18. Morton RH. Modelling human power and endurance. J Math Biol.
1990;28(1):49–64. PubMed doi:10.1007/BF00171518
19. Edwards S. High Performance Training and Racing. The Heart Rate
Monitor Book. Sacramento, CA: Fleet Feet Press; 1993.
20. Padilla S, Mujika I, Orbananos J, Santisteban J, Angulo F, Jose Goiri-
ena J. Exercise intensity and load during mass-start stage races in
professional road cycling. Med Sci Sports Exerc. 2001;33(5):796–802.
PubMed doi:10.1097/00005768-200105000-00019
21. Kelly LA, McMillan DG, Anderson A, Fippinger M, Fillerup G,
Rider J. Validity of ActiGraphs uniaxial and triaxial accelerometers
for assessment of physical activity in adults in laboratory conditions.
BMC Med Phys. 2013;13(1):5. PubMed doi:10.1186/1756-6649-13-5
22. Hopkins WG, Marshall SW, Batterham AM, Hanin J. Progres-
sive statistics for studies in sports medicine and exercise science.
Med Sci Sports Exerc. 2009;41(1):3–13. PubMed doi:10.1249/
MSS.0b013e31818cb278
23. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd
ed. New York, NY: Academic Press; 1969.
24. Batterham AM, Hopkins WG. Making meaningful inferences about
magnitudes. Int J Sports Physiol Perform. 2006;1(1):50–57. PubMed
doi:10.1123/ijspp.1.1.50
25. Rogalski B, Dawson B, Heasman J, Gabbett TJ. Training and game
loads and injury risk in elite Australian footballers. J Sci Med Sport.
2013;16(6):499–503. PubMed doi:10.1016/j.jsams.2012.12.004
26. da Silva CC, Goldberg TB, Soares-Caldeira LF, Dos Santos Oliveira
R, de Paula Ramos S, Nakamura FY. The effects of 17 weeks of ballet
training on the autonomic modulation, hormonal and general bio-
 802  Jeffries, Wallace, and Coutts
chemical profile of female adolescents. J Hum Kinet. 2015;47:61–71.
PubMed
27. Foster C, Lehmann M. Overtraining Syndrome. Philadelphia, PA:
W.B. Saunders; 1997.
28. Putlur P, Foster C, Miskowski JA, et al. Alteration of immune function
in women collegiate soccer players and college students. J Sports Sci
Med. 2004;3(4):234–243. PubMed
29. Jacobs CL, Hincapie CA, Cassidy JD. Musculoskeletal injuries
and pain in dancers: a systematic review update. J Dance Med Sci.
2012;16(2):74–84. PubMed
30. Minganti C, Capranica L, Meeusen R, Amici S, Piacentini MF. The
validity of session rating of perceived exertion method for quantifying
training load in teamgym. J Strength Cond Res. 2010;24(11):3063–
3068. PubMed doi:10.1519/JSC.0b013e3181cc26b9
31. Howe CA, Staudenmayer JW, Freedson PS. Accelerometer prediction
of energy expenditure: vector magnitude versus vertical axis. Med
Downloaded by Ebsco Publishing aellsworth@ebsco.com on 08/24/17, Volume 12, Article Number 6
IJSPP Vol. 12, No. 6, 2017
Sci Sports Exerc. 2009;41(12):2199–2206. PubMed doi:10.1249/
MSS.0b013e3181aa3a0e
32. Gallo T, Cormack S, Gabbett T, Williams M, Lorenzen C. Character-
istics impacting on session rating of perceived exertion training load
in Australian footballers. J Sports Sci. 2015;33(5):467–475. PubMed
doi:10.1080/02640414.2014.947311
33. Wyon MA, Twitchett E, Angioi M, Clarke F, Metsios G, Koutedakis Y.
Time motion and video analysis of classical ballet and contemporary
dance performance. Int J Sports Med. 2011;32(11):851–855. PubMed
doi:10.1055/s-0031-1279718
34. Lockie RG, Murphy AJ, Scott BR, Janse de Jonge XA. Quantifying ses-
sion ratings of perceived exertion for field-based speed training meth-
ods in team sport athletes. J Strength Cond Res. 2012;26(10):2721–
2728. PubMed doi:10.1519/JSC.0b013e3182429b0b
Copyright of International Journal of Sports Physiology & Performance is the property of
Human Kinetics Publishers, Inc. and its content may not be copied or emailed to multiple
sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.