SALIVARY CORTISOL RESPONSES AND PERCEIVED EXERTION
DURING HIGH INTENSITY AND LOW INTENSITY BOUTS OF RESISTANCE EXERCISE
Department of Exercise Sport Science, University of Wisconsin La Crosse,
10 September 2003
Journal of Sports Science and Medicine (2004) 3, 8-15
Google Scholar for Citing Articles
purpose of this study was to measure the salivary cortisol response
to different intensities of resistance exercise. In addition, we
wanted to determine the reliability of the session rating of perceived
exertion (RPE) scale to monitor resistance exercise intensity. Subjects
(8 men, 9 women) completed 2 trials of acute resistance training
bouts in a counterbalanced design. The high intensity resistance
exercise protocol consisted of six, ten-repetition sets using 75%
of one repetition maximum (RM) on a Smith machine squat and bench
press exercise (12 sets total). The low intensity resistance exercise
protocol consisted of three, ten-repetition sets at 30% of 1RM of
the same exercises as the high intensity protocol. Both exercise
bouts were performed with 2 minutes of rest between each exercise
and sessions were repeated to test reliability of the measures.
The order of the exercise bouts was randomized with least 72 hours
between each session. Saliva samples were obtained immediately before,
immediately after and 30 mins following each resistance exercise
bout. RPE measures were obtained using Borg's CR-10 scale following
each set. Also, the session RPE for the entire exercise session
was obtained 30 minutes following completion of the session. There
was a significant 97% increase in the level of salivary cortisol
immediately following the high intensity exercise session (P<0.05).
There was also a significant difference in salivary cortisol of
145% between the low intensity and high intensity exercise session
immediately post-exercise (P<0.05). The low intensity exercise did
not result in any significant changes in cortisol levels. There
was also a significant difference between the session RPE values
for the different intensity levels (high intensity 7.1 vs. low intensity
1.9) (P<0.05). The intraclass correlation coefficient for the session
RPE measure was 0.95. It was concluded that the session RPE method
is a valid and reliable method of quantifying resistance exercise
and that salivary cortisol responds promptly to the exercise load.
WORDS: Weight lifting, stress, endocrine effects.
many studies the endocrine response of subjects to varying types
of stress has been tested. Cortisol, the principal glucocorticoid
in humans, plays a major role in metabolism and immune function.
It has been shown that acute exercise induces a change in plasma
cortisol concentrations, which is dependent on the type of exercise
(Lac and Berton,
2000; Jacks et al., 2002).
Several studies have investigated the effect of both acute and chronic
resistance exercise on adrenocortical function (Mulligan et al.,
et al., 1999;
Fry et al., 2000;
Nindl et al., 2001;
Smilios et al., 2003).
However, there appear to be no studies that have measured salivary
cortisol responses to different intensities of resistance exercise.
measures of cortisol have been shown to be a valid and reliable
reflection of serum cortisol (Obminski and Stupnicki, 1997).
Salivary cortisol may actually provide a better measure than serum
cortisol of the stress response as it more accurately measures the
amount of unbound cortisol compared to serum measures (Vining et
There is also evidence that suggests fitter individuals show increased
cortisol responses compared to less trained individuals (Marthur
et al., 1986;
Luger et al., 1987).
The Borg 15-category scale for the rating of perceived exertion
(RPE) during physical activity has been widely researched for its
use in both clinical and exercise settings (Noble and Robertson,
based the RPE scale on the idea that a measure of perceived exertion
is the level of strain and/or heaviness experienced during physical
effort, as estimated by a specific rating method (Borg, 1998).
Since the unveiling of the original scale over forty years ago,
the CR-10 RPE scale has become a standard method to evaluate perceived
exertion in exercise testing, training, and rehabilitation and has
been validated against objective markers of exercise intensity (Borg
et al., 1985,
Noble et al., 1983).
However, to date this scale has not been evaluated to the same extent
for other high intensity exercises such as resistance training.
A recent study by Gearhart et al. (2001)
showed that the Borg CR-10 RPE scale can be used effectively during
single set resistance training sessions and that it is a valid measure
of exercise intensity. A second study by Gearhart et al. (2002)
yielded similar results thus expressing promise to its application
of the rating of single set perceived exertion.
A series of studies by Foster (1998),
Foster et al. (1996;
2001) and Day
et al.(in press)
have suggested that a single session RPE rating may accurately reflect
the intensity of an exercise session. A recent study conducted by
Day et al. (in press)
demonstrated that the session RPE could be used to quantify the
intensity of a resistance training session. However, the exercise
protocol used by Day et al (in press)
used a single set format. In addition, the research done thus far
using RPE during resistance exercise has not adequately addressed
the efficacy of its use during typical multi-set, higher intensity
weight training sessions undertaken by many trainees, particularly
athletes (Gearhart et al., 2001). It has been suggested that combined
psychological and physiological changes during high intensity training
provide important indicators for monitoring training stress (Filaire
et al., 2001).
Several studies have evaluated training load and prescribing exercise
periodization using session RPE (Foster et al., 1996; 2001).
Foster et al. (1996)
reported that self-directed increases in training load, using the
session RPE scale as a marker of intensity multiplied by exercise
duration in minutes to yield an index of the total training load,
improved athletic performance during cycling time trials. However,
another study by Foster et al. (1998)
revealed that a sudden increase in training load above normal training
limits caused a decrease in endurance performance and led to injury
or illness. Periodization or variation of training intensity should
be utilized within a weekly training plan and can be monitored using
session RPE values obtained by the individual after each exercise
session. Session RPE could also lead to optimal athletic performance
with a reduced injury/illness cost due to overtraining with endurance
The purpose of this study was to measure the salivary cortisol responses
to different intensities of resistance exercise. The secondary purpose
was to evaluate the effectiveness of using the session RPE scale
to measure physical effort during bouts of resistance training exercise,
as well as to examine the validity of this scale in rating entire
resistance training sessions of different intensities.
design and approach to the problem
This study used a randomized, crossover design, in which subjects
completed two experimental trials twice. For this study, subjects
performed a low intensity protocol, and a high intensity protocol
for two exercises (the bench press and the squat, respectively). For
the purpose of safety and the elimination of possible external variables,
which potentially could have affected results, both of these exercises
were performed on the Smith Machine. Each subject completed a total
of five sessions, on nonconsecutive days. Day one consisted of a familiarization
session that included informed consent procedures, instruction on
the use of CR-10 RPE scale and session RPE to rate perceived exertion.
The subsequent four sessions consisted of two high intensity workouts
at 75% of 1-RM and two low intensity workouts at 30% of 1-RM. The
order of these sessions was randomized with least 72 hours between
each session. The purpose of conducting the resistance exercise sessions
on two occasions was to determine the reliability of the session RPE
method for rating the intensity of the workout.
Seventeen volunteers between the ages of 18 and 25 were recruited
for this study including 8 men (Mean ± SD; 21.6 ± 1.2 years; 1.8 ±
0.1m; 86.6 ± 11.0 kg; 11.0 ± 5.2% body fat) and 9 women (20.0 ± 0.9
years; 1.6 ± 0.1m; 60.6 ± 8 kg; 21.2 ± 2.8% body fat). These volunteers
were required to meet the following requirements prior to participation
in this study: absence of any skeletal, muscle, cardiovascular, or
endocrine limitations; a history of a resistance-training program
of at least two sessions per week for at least three weeks prior to
participation in this study; and free of controlled and performance-enhancing
drugs for at least one month prior to and for the duration of the
study. Subjects provided informed consent as per the university's
Institutional Review Board. All subjects were measured for height,
body mass, and percent body composition during the first testing session.
During the duration of the study subjects were required to refrain
from intense exercise 24 hours prior to each testing session, to follow
the same diet on each day of each trial, and not to eat for at least
3-4 hours prior to any testing session. In addition, subjects were
instructed to abstain from alcohol and caffeine for a minimum of 24
hours prior to any testing session. Percent body fat was estimated
using four site measurements of skinfolds as per Durnin and Wormesley
At least one week prior to the acute resistance exercise protocol,
each subject had their one repetition maximum (1RM) determined on
the Smith machine squat and Smith machine bench press as previously
described (Kraemer et al., 1998).
This involved a number of warm-up trials being performed using 30%
(8-10 reps), 50% (4-6), 70% (2-4), and 90% (1 repetition) of an estimated
RM or 1-1.5 times subjects body weight (McBride et al.,
2002). Following the warm-up the subjects' resistance was increased
where the individual completed a number of maximal efforts to determine
Acute resistance exercise protocol
The high intensity resistance exercise protocol consisted of 6 sets
of 10 RM squats (75% of 1RM) and 6 sets of 10RM bench presses (75%
of 1RM) with 2 minutes of rest between each set. A similar protocol
has been previously used and has shown to result in changes in endocrine
function (Hymer et al., 2001).
If the subject failed to perform the 10 repetitions on any given set
due to fatigue the load was immediately adjusted to permit completion
of the remaining repetitions. The low intensity resistance exercise
protocol consisted of 3 sets of 10 repetitions at 30% of 1RM of the
same exercises as the high intensity protocol with 2 minutes of rest
between each exercise.
Rating of perceived exertion measures
During the familiarization session, each subject was given instructions
on the use of the modified CR-10 category RPE scale (Noble et al.,
1983; Borg et
al., 1985). The
session RPE measure, which was developed by Foster et al. (1996;
2001), was used
to rate intensity of the entire workout (Figure
1). A series of anchoring tests was used as previously described
by Gearhart et al. (2001)
to establish high and low perceptual anchors. In addition, the subject
was shown the scale 30 minutes following conclusion of the training
bout and asked, "How was your workout?" (Foster et al., 2001).
RPE was taken 30 minutes post-exercise to prevent particularly difficult
or easy elements near the end of the exercise session from skewing
the overall rating of the session. As a comparison to studies where
duration was multiplied by the session RPE to calculate the training
load, we multiplied the session RPE by the number of sets performed,
the number of repetitions performed and weight lifted to create a
term representative of the training load. The goal of the session
RPE is to encourage the subject to view the training session globally
and to simplify the myriad of exercise intensity cues during the exercise
Saliva samples were collected at the beginning of each testing session
(without stimulation, by spitting directly into a plastic tube), immediately
following completion of the resistance exercise session and 30 minutes
following completion of the last exercise (at the same time as the
obtaining the session RPE). Samples were obtained for all the testing
sessions from ten of the subjects who completed the study. Salivary
cortisol has been shown to have a circadian rhythm (Thuma et al.,
1995; Raff et
al.1998). To avoid
any confounding effects due to variations in circadian rhythm all
testing sessions were performed at the same time of day. Samples were
stored at -80ºC until analyzed. There is a strong relationship between
salivary and serum unbound cortisol both at rest (r = .93) and during
exercise (r = .90) (O'Connor and Corrigan, 1987).
Saliva measures of cortisol concentrations are independent of saliva
flow rate (Riad-Fahmy et al., 1983).
Salivary cortisol concentrations were determined in duplicate by Enzyme
Immunoassay using a Diagnostic Systems Laboratories Salivary Cortisol
Enzyme Immunoassay Kit (DSL, Webster, Texas). Assay plates were read
using an Opsys MRTM Microplate Reader (Dynex Technologies,
Chantilly, USA). Intra-assay variance was 7.2% and the sensitivity
of the assay was 0.011 µg·dL-1.
Statistical significance was set at the p < 0.05 level. Changes within
groups for salivary cortisol measures and RPE values were analyzed
using two way repeated measures analysis of variance. Comparisons
among the groups were made using analysis of variance. The Tukey post-hoc
test was used to identify significantly different group means. Each
subject's RPE values was averaged and compared to his/her session
RPE rating. This test was completed to identify if significant differences
exist between the session RPE rating and the accumulated RPE ratings
obtained during each resistance training session. Interclass correlation
coefficients (ICC) were calculated to establish the reliability of
the session RPE method. Bivariate relationships were calculated using
Pearson's product moment correlations to examine the relationship
between changes in cortisol levels and session RPE. Lastly bivariate
correlations were computed relating changes in cortisol and training
was a significant increase in the level of salivary cortisol immediately
following the high intensity exercise session (Figure
2). There was a significant difference between the low intensity
and high intensity exercise session immediately post-exercise. The
low intensity exercise did not result in any significant changes in
There was a significant difference between the mean RPE values for
each intensity (p < 0.05) (Figure
3 and Figure 4). There was
also a significant difference between the session RPE values for each
intensity of lifting (high intensity 7.1 vs. low intensity 1.9) (p
< 0.05) (Figure 5). There was
no significant difference between the average RPE values and the session
RPE values for the squat exercise. However, there was a significant
difference between the average RPE value for the bench press exercise
and the session RPE value during each intensity for the bench press
exercise (p < 0.05).
A test for reliability of the session RPE to predict the same value
across two different trials of the same intensity was performed. The
ICC was 0.95 with the 95% confidence interval of 0.90-0.97.
There were no significant correlations between salivary cortisol levels
immediately post or 30 minutes post exercise and measures of RPE (both
average RPE and session RPE). There were also no significant correlations
between salivary cortisol levels and load.
studies have shown that resistance exercise has a significant effect
on cortisol levels following exercise (Mulligan et al., 1996;
Nindl et al., 2001).
The present study showed that salivary cortisol responses were significantly
different immediately post exercise between the low intensity and
high intensity exercise sessions (145%). Immediately following the
high intensity acute resistance exercise bout there was a significant
elevation of 97% in salivary cortisol from baseline. This increase
in salivary cortisol was significantly larger than the cortisol response
for the low intensity resistance exercise session.
Salivary cortisol levels have been shown to increase following acute
exercise with the response dependent on the intensity and duration
of activity (Lac and Berthon, 2000;
Jacks et al., 2002).
To our knowledge there are no studies that have measured salivary
cortisol following different intensities of resistance exercise. However,
there is limited research investigating salivary cortisol responses
to resistance exercise. Salivary cortisol provides a stress free,
non-invasive procedure (Vining and McGinley, 1987)
that avoids additional stress caused by venipuncture (Lac et al.,
cortisol may also be a better measure of adrenocortical function as
it represents more accurately the level of unbound cortisol (Vining
et al., 1983).
In the present investigation we showed that a high intensity bout
of resistance exercise significantly increases the level of salivary
cortisol. This in agreement with the majority of studies that have
shown that high intensity resistance exercise elevates serum cortisol
(Mulligan et al., 1996;
Fry et al., 2000;
Nindl et al., 2001;
Smilios et al., 2003).
A recent study by Smilios et al. (2003)
showed that different resistance exercise protocols produce different
hormonal response patterns depending on the number of sets that are
Previous research has shown that the volume of resistance exercise
as measured by the number of sets performed, affects the hormonal
concentrations. Studies have demonstrated that performing three sets
of each exercise results in higher levels of cortisol, testosterone
and growth hormone compared to one set of each exercise (Mulligan
et al., 1996;
Gotshalk et al., 1997).
Our data also confirmed that salivary cortisol levels were significantly
elevated when a higher volume of work was performed (6 sets versus
3 sets). In addition to the number of sets performed, another factor
that may be affect the salivary cortisol response is the training
status of the subjects. The majority of subjects in the present study
were only recreationally trained, as reflected by the average squat
1RM to bodyweight ratio (1.62 for men and 1.16 for women). It has
been suggested that multiple set workouts are more effective for well
trained individuals (ACSM, 2002)
and hormonal responses are influenced by the training status of the
individual (Fry et al., 2000).
For the present study there was a strong correlation (r = .54, p =
0.08) between the squat 1RM to bodyweight ratio and the percentage
change in salivary cortisol concentrations from pre- to post the exercise
bout. Although this was not a significant relationship it does suggest
that the training status of the subjects is related to the hormonal
response. There was however a wide range of individual responses of
salivary cortisol to the different intensities of resistance exercises,
as demonstrated by the large standard deviations (Figure
1). In addition, the inclusion of both men and women in the present
study could provide a confounding factor. There was no significant
differences between the men and women in the study for both salivary
cortisol responses and session RPE values. However, this could have
contributed to the large variation of salivary cortisol responses
as there may have been variation in the menstrual status of the female
subjects. We attempted to control for the time of day by having the
subjects complete their testing for all sessions at the same time.
All testing was conducted in the morning as salivary cortisol has
been shown to have a circadian rhythm (Thuma et al., 1995;
Raff et al.1998).
The second purpose of this study was to evaluate the reliability and
effectiveness of the session RPE method to rate an overall resistance
training session using multi-set exercises. Based on the data, the
session RPE method appears to be a reliable method of quantifying
resistance training intensities. We have previously shown that the
session RPE method is a reliable measure for evaluating single-set
exercise bouts (Day et al., in
press). However, multiple set workouts have been recommended for
optimal development of strength and power (ACSM, 2002).
Therefore, we felt it was important to determine the reliability of
this measure with multiple set training. This was the rationale for
using two sessions of each exercise intensity. There was a significant
difference between the session RPE values for each intensity for the
two exercises (bench press and squat). The average RPE value and session
RPE value for the bench press exercise were also significantly different.
For the squat exercise however, no significant difference was found
between average RPE value and session RPE value.
These findings are similar to a previous study conducted in our laboratory
(Day et al., in press).
Another study conducted recently in our laboratory compared session
RPE during easy, moderate, and high intensity resistance training
to session RPE during comparable intensities of steady state aerobic
exercise on a cycle ergometer (Sweet et al., in press).
Session RPE and the mean RPE all increased as the % 1-RM increased
despite a decrease in repetitions and total workload. The results
of this series of studies supports the idea that the session RPE is
a valid method for quantifying the intensity of resistance training,
and is generally comparable to aerobic training. The difference between
the bench press and squat exercises for mean RPE versus session RPE
is similar to our previous findings, in that it appears that RPE measurements
taken after each set varied widely depending on the type of resistance
exercise being performed, ie. large muscle mass exercises versus small
muscle mass exercises (Day et al., in
press; Sweet et al., in press).
Many factors could have influenced variations in RPE measurements
such as motor unit recruitment and energy expenditure.
In the present study we used an exercise protocol designed to clearly
delineate between higher volume, high intensity bouts of exercise
and a low intensity workout to investigate the hormonal response and
perceived exertion. However, there was no attempt to equate the exercise
protocols for total work performed, although the rest periods were
the same for both protocols. It is also important to note that the
exercise bouts used in the present study do not necessarily reflect
the type of workout that would be used traditionally by exercisers.
Future studies could investigate the effect of exercise protocols
designed to improve maximum strength, hypertrophy and power on the
hormonal responses, in addition to session RPE.
Other researchers have investigated perceived exertion and resistance
exercise (Gearhart et al., 2001;
training consists of a complex milieu of variables including sets,
repetitions, rest periods and type of exercise performed. Therefore,
resistance exercise represents a unique mode to study perceived exertion
and hormonal responses. The RPE values for each set were taken in
addition to the session rating. The purpose of taking the set RPE
values was to further familiarize the subjects with rating their perceived
effort on the modified CR-10 scale. We believed this would increase
the accuracy of the session RPE value. This study provides further
evidence validating the findings of Gearhart et al. (2002)
and Day et al. (in press),
where fewer repetitions of a heavier resistance was perceived to be
more difficult than performing more repetitions of a lighter resistance.
Training volume in resistance exercise is a composite of the number
of sets, number of repetitions and the amount of resistance lifted.
This is an important difference from non resistance type of training
where the total duration of exercise in minutes is the appropriate
duration measure. Because of the long periods of recovery required
in resistance training, particularly in high intensity resistance
training, time per se is probably an inappropriate measure of training
volume. Previous studies by Day et al. (in
et al. (2001)
and Sweet et al. (in press)
have shown that RPE is most influenced by exercise intensity and not
by the volume of exercise being performed. Further research is required
to investigate the role of training volume during resistance exercise,
on both RPE and salivary hormonal responses.
conclusion, there was a significant difference between the salivary
cortisol responses immediately following the high and low intensity
exercise protocols. This study has demonstrated that salivary measures
of cortisol can be used to delineate between high and low intensity
workouts. The results of this study have also shown that session RPE
is a reliable and useful tool of measuring the intensity of a resistance
training session. This scale would be a beneficial tool for researchers,
strength coaches, recreational weightlifters, and athletes as they
strive to rate the work intensity of a resistance training session.
Overall, the session RPE scale was shown to be a reliable tool to
quantify work of low intensity and high intensity workouts.
The present study showed that salivary cortisol responses were
significantly different immediately post exercise between the
low intensity and high intensity exercise sessions
measures of cortisol can be used to delineate between high and
low intensity resistance exercise bouts.
session RPE method appears to be a reliable method of quantifying
Michael R. MCGUIGAN
Employment: Ass. Prof. in the
Department of Exercise and Sport Science, Univ. of Wisconsin-La
Research interests: Hormonal and muscular responses to
Employment: Undergraduate student in the Depart. of Exercise
and Sport Science, University of Wisconsin-La Crosse.
Employment: Prof. in the Depart. of Exercise and Sport Science,
University of Wisconsin-La Crosse.
Research interests: Clinical and exercise physiology,
high performance physiology, exercise physiology.