Journal of Sports Sciences, June 2007; 25(8): 845 – 850

Quantifying delayed-onset muscle soreness: A comparison of

unidimensional and multidimensional instrumentation

DANIEL J. CLEATHER1 & SHARON R. GUTHRIE2

1
English Institute of Sport, St. Mary’s College, Twickenham, UK and 2Department of Kinesiology, California State
University, Long Beach, CA, USA

(Accepted 11 July 2006)

Abstract
Unidimensional pain instrumentation, whereby participants simply rate the intensity of their pain on one evaluative level, has
been the most common method of assessing delayed-onset muscle soreness (DOMS). However, pain has been shown to be a
multidimensional phenomenon including sensory, affective, and evaluative aspects. The aims of this study were two-fold: (1)
to compare the DOMS pain responses derived from a multidimensional instrument (i.e. the McGill Pain Questionnaire –
MPQ) with those using a unidimensional measure (i.e. a visual analogue scale), and (2) to identify the MPQ descriptors
most commonly used to characterize DOMS among a sample of 14 male (mean age ¼ 24.7 years, s ¼ 4.4) and 9 female
participants (mean age ¼ 24.6 years, s ¼ 3.5). Although the results demonstrated no significant differences between the pain
ratings of the two instruments (mean values of the pain rating indices had a Spearman rank correlation coefficient of
r ¼ 1.00), suggesting no significant advantage to be gained in using the MPQ, a clearer description of DOMS emerged. The
most frequently selected DOMS descriptors were ‘‘tight’’ (95% of participants chose this descriptor at least once), ‘‘sore’’
(86%), ‘‘tender’’ (86%), ‘‘annoying’’ (86%), and ‘‘pulling’’ (68%). These findings may be of use to researchers and sports
medicine professionals in their deliberations about which instrumentation to use in quantifying DOMS and in distinguishing
such pain from other, potentially more serious, musculoskeletal damage.

Keywords: Delayed-onset muscle soreness, pain scales

in the battery of tests used to measure exercise-

Introduction
Exercise-induced muscle damage, caused by the
strenuous use of muscles unaccustomed to exercise,
particularly exercise dominated by eccentric muscle
actions, has been well documented. This damage,
which occurs at the cellular level, is manifested by
several markers of muscle function, including pain,
tenderness, inflammation, loss of strength, change in
the length – tension relationship of the muscle, an
increase in muscle stiffness, and biochemical markers
(e.g. creatine kinase, lactate dehydrogenase, and
glutamic oxaloacetic transaminase) (Clarkson &
Sayers, 1999).
Pain of this type, which is referred to as delayed-
onset muscle soreness (DOMS) because it generally
peaks 1 – 3 days after exercising, has been one of the
key variables reported in the quantification of
exercise-induced muscle damage (Connolly, Sayers,
& McHugh, 2003). Thus, it is common in experi-
mental research to include a measurement of DOMS
induced muscle damage (Barlas et al., 2000; Hilbert,
Sforzo, & Swensen, 2003; Nosaka, Sakamoto,
Newton, & Sacco, 2001). In attempting to quantify
DOMS, however, most researchers have used uni-
dimensional instruments that obtain merely an
evaluation of the intensity of one’s pain, for example
simple numerical scales (e.g. Chen & Hsieh, 2001),
graphic rating scales (e.g. Allen, Mattacola, & Perrin,
1999), and most often visual analogue scales (e.g.
Nosaka et al., 2001).
Unfortunately, unidimensional assessment tools
are grounded in the notion of pain as a sensation
varying in intensity only. This thinking runs counter
to scientific evidence indicating pain to be a multi-
dimensional phenomenon consisting of at least three
different dimensions: (a) a sensory dimension that
includes temporal, spatial, pressure, and thermal
aspects; (b) an affective dimension that encompasses
a person’s emotional response to the pain; and
(c) an evaluative dimension incorporating a person’s

Correspondence: D. J. Cleather, English Institute of Sport, St. Mary’s College, Waldegrave Road, Twickenham TW1 4SX, UK.
E-mail: daniel.cleather@eis2win.co.uk
ISSN 0264-0414 print/ISSN 1466-447X online Ó 2007 Taylor & Francis
DOI: 10.1080/02640410600908050
846 D. J. Cleather & S. R. Guthrie
cognitive interpretation of the pain (Melzack, 1975;
Melzack & Torgerson, 1971). Although multidimen-
sional pain instruments do exist, including the
McGill Pain Questionnaire (MPQ; Melzack, 1975)
and the Differential Descriptor Scale (DDS; Gracely
& Kwilosz, 1988), such measures have rarely been
used in studies of DOMS.
This paucity of research is surprising given that
researchers have advocated the use of multidimen-
sional pain measures (Hilbert et al., 2003; Mattacola,
Perrin, Gansneder, Allen, & Mickey, 1997). More-
over, there is evidence that instruments measuring
both sensory and affective components of pain may
be more effective than unidimensional instrumenta-
tion in detecting small differences in pain (Doctor,
Slater, & Atkinson, 1995). Although researchers
using multidimensional instruments (e.g. the MPQ,
Barlas et al., 2000; the DDS, Hilbert et al., 2003)
have demonstrated a temporal pattern of DOMS
similar to that found with unidimensional tools, they
have not conducted a comparison of such instru-
mentation. It therefore remains to be determined
whether a multidimensional approach to measuring
DOMS is superior to the more commonly used
unidimensional intensity-based approaches. The
present study was designed to address this issue.
Delayed-onset muscle soreness responses were
evaluated using both unidimensional and multi-
dimensional pain instrumentation following an
exercise protocol designed to induce DOMS, and
these responses were compared. A visual analogue
scale (VAS) and the McGill Pain Questionnaire
were selected because they are frequently used in
DOMS research and clinical pain research, respec-
tively. Comparisons were made between the pain-
rating indices produced by each instrument and
their ability to detect significant differences in the
pain ratings of DOMS (i.e. their sensitivity) on
seven successive days. A secondary purpose was to
identify the MPQ adjectives most commonly used
by the participants in this study to describe their
DOMS. It was hypothesized that these adjectives
would differ from those used to describe other
forms of musculoskeletal injury and that such
information, if reproduced in additional studies,
would be of value to sports medicine professionals
in identifying DOMS.
Methods
Participants
Ethical approval for this study was obtained from the
Institutional Review Board of California State Uni-
versity, Long Beach. Each participant was screened
to ensure his or her physical readiness to take part in
the study and to confirm that no pain was currently
being experienced in the non-dominant arm. In
accordance with the Protection of Human Subjects
protocol, participants provided written informed
consent before the study began.
Fourteen males (mean age ¼ 24.7 years, s ¼ 4.4)
and nine females (mean age ¼ 24.6 years, s ¼ 3.5),
recruited from the student body of a large urban
university in the southern California area, served as
participants in this study. The participants included
collegiate athletes (n ¼ 1), recreational athletes
(n ¼ 20), and sedentary individuals (n ¼ 2). Of the
recreational athletes, nine were former collegiate
athletes. In terms of participation in weight training
at the time of the study, five had never trained, ten
were training one to two times per week, and eight
were training three to four times a week. One
participant was excluded from the study due to an
extreme physical reaction to the exercise protocol;
thus, 22 participants were included in the final data
sample.
Instrumentation
Visual analogue scale. The VAS was selected because
it is commonly used in DOMS research. The VAS,
which has been shown to be valid and reliable
in previous research (Price, McGrath, Rafii, &
Buckingham, 1983), consists of a line 100 mm long,
anchored at the left by the words ‘‘no pain’’ and
at the right by the words ‘‘unbearable pain’’ (see
Figure 1). Participants were asked to draw a vertical
line on the scale representing the current intensity of
their pain. A pain-rating index was calculated from
each VAS, with the intensity of pain considered to be
the distance (in millimetres) of the participant’s mark
from the left-hand side of the scale.
McGill Pain Questionnaire. The MPQ (Melzack,
1975) was selected because it is widely used in
clinical pain research and has well-established
psychometric properties (Graham, Bond, Gerkovich,
& Cook, 1980). The instrument consists of 78
words, divided into twenty subcategories, ten of
which represent the sensory dimension of pain, five
the affective dimension, one the evaluative dimen-
sion, and four miscellaneous subcategories. An
individual is asked to select those adjectives that best
describe his or her pain. A completed MPQ yields a
group of descriptors that can be used to characterize
an individual’s pain. The words chosen, and the
intensities of each dimension, vary across different
Figure 1. Visual analogue scale (VAS).

pain syndromes (e.g. dental, back, menstrual pain)
(Melzack, 1975, 1987).
The participants were directed to select only the
words that described their current pain and were
limited to choosing only one word from each
subcategory. They were not, however, required to
select a word from each subcategory. Each word on
the MPQ is assigned a scale numerical value,
according to its strength relative to predetermined
anchor words. Subcategories are ordered by the scale
values of the words, with the lowest value word
appearing first and the highest value word last. These
scale values were later used to calculate pain-rating
indices based on the words chosen for each
completed MPQ. Pain-rating indices were calculated
for the entire MPQ, as well as for the sensory and
affective subclasses of MPQ descriptors.
Procedures
After the screening process, the participants engaged
in an exercise protocol designed to induce DOMS in
the elbow flexors of their non-dominant arm. They
then reported the pain experienced, first using a VAS
and then the MPQ. During this process, each person
was observed and instructed by the investigator to
ensure that the instruments were completed correctly
onsite and would be completed correctly in the
following unsupervised conditions. Subsequently,
participants were each given seven copies of both
the VAS and MPQ with the following directives: (a)
complete each instrument at the same time of day as
the initial questionnaires, on each of the next seven
successive days; (b) do not examine or compare the
questionnaire responses completed on previous days;
(c) do not take any painkillers, receive massages, ice
the affected limb, or use any other measures to
control DOMS; and (d) return all of the completed
questionnaires to the investigator at the end of the
seven-day period following the DOMS-inducing
exercise.
DOMS-inducing exercise. After a warm-up consisting
of arm swings and upper body stretching, the one-
repetition maximum (1-RM) of the non-dominant
arm of each participant was determined using a
preacher curl machine. This was accomplished by
having the individual lift progressively heavier
weights, resting for 30 s between lifts, until the
maximum weight that could be lifted once was
ascertained. Following this protocol, the participants
were allowed a 30-s rest.
Next, the participants completed three sets of
eccentric preacher biceps curls with their non-
dominant arm, using their 1-RM and lowering the
weight to a 3-s count. The researcher then lifted the
weight back to the starting position, and the process
Quantifying delayed-onset muscle soreness 847
was repeated. The researcher also ‘‘spotted’’ the
eccentric phase of the bicep curl, preventing the
weight from being lowered too quickly, to allow
the participants to work to complete muscular
failure. This process was repeated until the partici-
pants could no longer control the rate of descent
of the weight. The participants then took a 30-s
rest before attempting a second set, and an addi-
tional 30-s rest before a third and final set.
Data analysis
The PC version of SPSS (Statistical Package for
Social Sciences, Chicago, IL) was used for all
inferential statistical tests, while correlation and
descriptive statistics were calculated using Microsoft
Excel. Spearman’s rank correlation coefficients were
calculated to determine the relationships between the
various pain-rating indices produced by the two
instruments. Friedman’s test was computed on the
VAS and MPQ data to compare the instruments in
terms of their ability to detect significant differences
in the pain rating of DOMS on different days.
Friedman’s test, which is the non-parametric analo-
gue of a repeated-measures analysis of variance, was
used as the data were ordinal. Alpha was set at 0.05 a
priori. Finally, descriptive statistics were used to
identify the most common words and subcategories
of words used to describe DOMS among this sample
of 22 participants.
Results
All 23 participants experienced pain following the
experimental protocol. The protocol was therefore
considered to be novel and of sufficient volume to
induce DOMS in all participants.
Correlation between the MPQ and the VAS
Figure 2 shows the mean values of the pain-rating
indices for the MPQ and the VAS. The mean values
of the pain-rating indices were highly correlated and
the Spearman rank correlation coefficient was
r ¼ 1.00. In contrast, individual responses to the
MPQ and VAS were not as highly correlated when
considered on a daily basis. Although responses on
day 2 and on days 5 – 7 yielded moderately high
correlations (r ¼ 0.71, r ¼ 0.54, r ¼ 0.73, and
r ¼ 0.72, respectively), responses on days 2 – 4 were
only moderately correlated (r ¼ 0.47, r ¼ 0.35, and
r ¼ 0.45, respectively).
Sensitivity of the MPQ and the VAS
Friedman’s test was used to calculate w2 values for
both the VAS and the MPQ. The w2 values
848 D. J. Cleather & S. R. Guthrie

Figure 2. Mean scores of the pain-rating indices for the MPQ and VAS over time.

calculated for the VAS and the MPQ were w2 ¼ 109.1 Table I. MPQ descriptors chosen at least once by 50% or more of
and w2 ¼ 91.3, respectively. This finding indicates participants.
that both instruments were individually capable of
% participants choosing
detecting significant differences in pain ratings Descriptor descriptor at least once
between successive days where alpha was set at
0.05 a priori. Tight 95
Sore 86
Tender 86
Analysis of MPQ pain descriptor responses Annoying 86
Pulling 68
Table I presents the words most commonly used by Cramping 59
the participants to describe their DOMS. The term Dull 59
‘‘tight’’ was chosen at least once by 95% of Nagging 59
Pressing 50
participants, followed by ‘‘sore,’’ ‘‘tender’’ and Tugging 50
‘‘annoying’’ (86%), and ‘‘pulling’’ (68%). The words Troublesome 50
most often chosen to describe DOMS belonged to a
number of subcategories. As Table II shows,
descriptors from subcategories 18 (miscellaneous),
9 (sensory: dullness), and 16 (evaluative) were variation in the temporal stages of DOMS because it
chosen most often. assesses several dimensions of pain.
There are several possible explanations for this
finding. One way in which a multidimensional
Discussion
instrument could be found to be more sensitive than
The main aim of this study was to compare DOMS a unidimensional tool is if the pain-rating indices for
pain responses derived from both unidimensional each pain dimension (in this case, the sensory and
and multidimensional instrumentation. We found no affective subscales of the MPQ) showed different
significant differences between the pain rating of temporal patterns. In this study, however, the
the temporal changes in DOMS between the VAS sensory and affective subscales were highly correlated
(a unidimensional pain instrument) and the MPQ with one another and the complete MPQ; thus, the
(a multidimensional pain instrument). Previous power of using two dimensions for discriminatory
research using multidimensional instrumentation purposes was significantly reduced.
(Doctor et al., 1995) has suggested that the MPQ is The administrative protocol of the MPQ may
more sensitive than the VAS in distinguishing the account, at least partially, for the correlation between
 Quantifying delayed-onset muscle soreness 849

Table II. MPQ subcategories chosen at least once by at least 50% damage. As a result, the condition, in comparison
of participants to describe DOMS. with other lesser known pain syndromes, may not
% participants have evoked a strong affective response, which in
choosing descriptor turn would have reduced the effect of the affective
Subcategory at least once dimension on the complete MPQ. Forty-two words
belong to the sensory subcategory on the MPQ, and
18: Miscellaneous (tight, numb, 100
drawing, squeezing, tearing)
these words were chosen a total of 459 times
9: Sensory: dullness (dull, 95 (i.e. sensory words were chosen 10.9 times per
sore, hurting, aching, heavy) word). In contrast, 14 words belong to the affective
16: Evaluative (annoying, 95 subcategory on the MPQ, and words within this
troublesome, miserable, group were chosen only 50 times (i.e. 3.6 times per
intense, unbearable)
6: Sensory: traction pressure 86
word). Thus, significantly fewer affective descriptors
(tugging, pulling, wrenching) were selected than sensory descriptors (P 5 0.05),
10: Sensory: miscellaneous 86 even when adjusted for the greater number of
(tender, taut, rasping, splitting) sensory descriptors.
5: Sensory: constrictive pressure 82 The structure of the MPQ also may affect its
(pinching, pressing, gnawing,
cramping, crushing)
ability to detect significant differences in DOMS on
20: Miscellaneous (nagging, 73 successive days. In this study, some of the more
nauseating, agonizing, commonly chosen descriptors of DOMS were in the
dreadful, torturing) miscellaneous subcategories. This may have reduced
1: Sensory: temporal 55 the sensitivity of the MPQ, as the terms within the
(flickering, quivering, pulsing,
throbbing, beating, pounding)
miscellaneous groups are more heterogeneous. For
3: Sensory: punctuate pressure 55 example, in subcategory 18, there are words descrip-
(pricking, boring, drilling, tive of constrictive pressure (‘‘tight’’ and ‘‘squeez-
stabbing, lancinating) ing’’), dullness (‘‘numb’’ and ‘‘drawing’’), and
11: Affective: tension 50 incisive pressure (‘‘tearing’’). This contrasts with
(tiring, exhausting)
17: Miscellaneous (spreading, 50
the subcategories of the sensory, affective, and
radiating, penetrating, piercing) evaluative dimensions, which consist of more homo-
geneous adjectives (Melzack, 1975; Melzack &
Torgerson, 1971). Similarly, some of the words
most commonly chosen to describe DOMS may
the sensory and affective subscales. When complet- have been viewed by some participants as synon-
ing the MPQ, the protocol demands that participants ymous, giving unjustified weight to that level of pain;
respond to all of the sensory subcategories before for example, some people might find the words
responding to the affective subcategories. It is ‘‘taut’’ and ‘‘tight’’ to be identical, yet because they
possible that participants’ responses to the sensory appear in two different subcategories this facet of
subcategories, and hence their evaluation of the pain may have been reported twice.
sensory component of their pain, may have influ- The limitations of the MPQ highlighted may
enced their responses to the affective subcategories. certainly explain why it was not more sensitive to
This could have occurred through two mechanisms. changes in DOMS than the VAS. It could be argued
First, participants may have selected affective words that these limitations make the MPQ a less robust
that had similar rank positions to those they had measure of DOMS than the VAS. However, a
already selected from the sensory categories. Second, multidimensional measure of pain that does not
the process of focusing on and evaluating their exhibit these limitations may still be more sensitive
sensory experience of DOMS may have affected the than the VAS to changes in the sensation of DOMS.
participants’ emotional response to DOMS. Such It is possible that a modified MPQ, reflecting the
response bias could be avoided by randomizing the aforementioned considerations, may enhance the
order in which subcategories are presented for tool’s ability to detect differences in DOMS on
response and altering this order each time the successive days. Developing a DOMS-specific MPQ
questionnaire is administered. Such an approach is that includes only the most common DOMS
employed when using the Differential Descriptor descriptors is another possibility. Alternatively, re-
Scale (Gracely & Kwilosz, 1988). search employing different multidimensional pain
Another possible explanation is associated with the instruments (e.g. the Differential Descriptor Scale)
participants’ prior knowledge and experience with may reveal an instrument that is more sensitive than
DOMS. Most of the individuals in this study were the VAS.
familiar with the sensation of DOMS and realized The fact that the correlation between the MPQ
that their pain was not indicative of serious muscular and the VAS was diminished between days 2 and 4
850 D. J. Cleather & S. R. Guthrie
warrants some discussion, especially as this is the
period when DOMS symptoms are known to peak.
As the VAS is primarily an evaluative measure, these
results could highlight a difference between the
sensory and affective experience of DOMS and
the evaluative dimension. This may be explained
by the fact that many participants in the study were
experienced weight trainers who were accustomed to
the experience of DOMS.
Identifying the most frequently used MPQ words
to describe DOMS was a secondary aim of
this study. The pain descriptors ‘‘tight’’, ‘‘sore’’,
‘‘tender’’, ‘‘annoying’’, and ‘‘pulling’’ were most
commonly reported, followed by ‘‘cramping’’,
‘‘dull’’, ‘‘nagging’’, ‘‘pressing’’, ‘‘tugging’’, and
‘‘troublesome’’. Moreover, descriptors associated
with subcategories 18 (miscellaneous), 9 (dullness),
and 16 (evaluative) were chosen most often by
participants experiencing DOMS. Further research
is needed to determine the MPQ descriptors
associated with other musculoskeletal injuries. If
future research produces consistent findings, athletic
trainers and other sports medicine professionals may
find these pain descriptors useful in distinguishing
DOMS from other forms of pain indicating other,
potentially more serious, musculoskeletal damage.
References
Allen, J. D., Mattacola, C. G., & Perrin, D. H. (1999). Effect of
microcurrent stimulation on delayed-onset muscle soreness:
A double-blind comparison. Journal of Athletic Training, 34,
334 – 337.
Barlas, P., Craig, J. A., Robinson, J., Walsh, D. M., Baxter, G. D.,
& Allen, J. M. (2000). Managing delayed-onset muscle
soreness: Lack of effect of selected oral systemic analgesics.
Archives of Physical Medicine and Rehabilitation, 81, 966 – 972.
Chen, T. C., & Hsieh, S. S. (2001). Effects of a 7-day eccentric
training period on muscle damage and inflammation. Medicine
and Science in Sports and Exercise, 33, 1732 – 1738.
Clarkson, P. M., & Sayers, S. P. (1999). Etiology of exercise-
induced muscle damage. Canadian Journal of Applied Physiology,
24, 234 – 248.
Connolly, D. A. J., Sayers, S. P., & McHugh, M. P. (2003).
Treatment and prevention of delayed onset muscle soreness.
Journal of Strength and Conditioning Research, 17, 197 – 208.
Doctor, J. N., Slater, M. S., & Atkinson, J. H. (1995). The
Descriptor Differential Scale of pain intensity: An evaluation of
item and scale properties. Pain, 61, 251 – 260.
Gracely, R. H., & Kwilosz, D. M. (1988). The Descriptor
Differential Scale: Applying psychophysical principles to
clinical pain assessment. Pain, 35, 279 – 288.
Graham, C., Bond, S. S., Gerkovich, M. M., & Cook, M. R.
(1980). Use of the McGill Pain Questionnaire in the assess-
ment of cancer pain: Replicability and consistency. Pain, 8,
377 – 387.
Hilbert, J. E., Sforzo, G. A., & Swensen, T. (2003). The effects of
massage on delayed onset muscle soreness. British Journal of
Sports Medicine, 37, 72 – 75.
Mattacola, C. G., Perrin, D. H., Gansneder, B. M., Allen, J. D., &
Mickey, C. A. (1997). A comparison of visual analogue and
graphic rating scales for assessing pain following delayed onset
muscle soreness. Journal of Sport Rehabilitation, 6, 38 – 46.
Melzack, R. (1975). The McGill Pain Questionnaire: Major
properties and scoring methods. Pain, 1, 277 – 299.
Melzack, R. (1987). The short-form McGill Pain Questionnaire.
Pain, 30, 191 – 197.
Melzack, R., & Torgerson, W. S. (1971). On the language of pain.
Anesthesiology, 34, 50 – 59.
Nosaka, K., Sakamoto, K., Newton, M., & Sacco, P. (2001). How
long does the protective effect on eccentric exercise-induced
muscle damage last? Medicine and Science in Sports and Exercise,
33, 1490 – 1495.
Price, D. D., McGrath, P. A., Rafii, A., & Buckingham, B. (1983).
The validation of visual analogue scales as ratio scale measures
for chronic and experimental pain. Pain, 17, 45 – 56.