Content uploaded by Jonas Tesarz
Author content
All content in this area was uploaded by Jonas Tesarz on Oct 17, 2017
Content may be subject to copyright.
Pain perception in athletes compared to normally active controls: A systematic
review with meta-analysis
Jonas Tesarz
a,
⇑
, Alexander K. Schuster
b
, Mechthild Hartmann
a
, Andreas Gerhardt
a
, Wolfgang Eich
a
a
Department of General Internal Medicine and Psychosomatics, Medical Hospital, University of Heidelberg, Germany
b
Mannheimer Institute of Public Health, Medical Faculty Mannheim, University of Heidelberg, Germany
Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.
article info
Article history:
Received 6 September 2011
Received in revised form 5 March 2012
Accepted 5 March 2012
Keywords:
Athletes
Pain perception
Pain threshold
Pain tolerance
Systematic review
Meta-analysis
abstract
This study systematically reviewed differences in pain perception between athletes and normally active
controls. We screened MEDLINE, Sport-Discus, EMBASE, Web of Science, PsycINFO, PSYNDEX, and the
citations of original studies and systematic reviews. All studies on experimentally induced pain that com-
pared pain perception between athletes and normally active controls were eligible. The main outcome
measures were pain tolerance and pain threshold. Effects are described as standardized mean differences
and were pooled using random-effects models. Fifteen studies including 899 subjects met the inclusion
criteria. Twelve of these studies assessed pain tolerance, and 9 studies examined pain threshold. A meta-
analysis of these studies revealed that athletes possessed higher pain tolerance compared to normally
active controls (effect size calculated as Hedges’ g = 0.87, 95% confidence interval [CI
95
] 0.53–1.21;
P< 0.00001), whereas available data on pain threshold were less uniform (Hedges’ g = 0.69, CI
95
0.16–
1.21; P= 0.01). After exclusion of studies with high risk of bias, differences between groups in pain
threshold were not significant any longer. Our data suggest that regular physical activity is associated
with specific alterations in pain perception. Psychological and biological factors that may be responsible
for these alterations are discussed.
Ó2012 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
1. Introduction
Pain perception in athletes is commonly believed to differ from
pain perception in normally active persons. This belief is primarily
based on anecdotes of athletes who continue to exercise in the face
of severe injury. Researchers have also postulated that long-stand-
ing physical activity may alter pain perception and have often con-
cluded that athletes possess higher pain thresholds and higher pain
tolerance [58,60]. However, the available scientific data on pain
perception in athletes are inconsistent and partially contradictory
[24,25,50,57,69].
Therapeutic exercise is included in most multidisciplinary
treatment programs, and it is recommended in numerous treat-
ment guidelines for pain patients [7,29]. However, exercise therapy
is not a uniform method but often includes variations in content,
dose, and mode of delivery. Most studies of different exercise
modes and the long-term effects of exercise in patients have been
conducted essentially for the further development of therapeutic
exercise programs. However, these studies are expensive and time
consuming. Studies in athletes offer the opportunity for an evalu-
ation of somatic and psychological effects of regular physical activ-
ity on pain perception, which might foster the development of
effective types of exercise for the relief of symptoms in pain
patients.
Furthermore, pain is a natural mechanism of protection against
injuries and overuse, representing an important diagnostic feature
[28]. Therefore, a more profound knowledge of the impact of phys-
ical activity on pain perception and processing will impact the
medical care of pain patients in general, and rehabilitation pro-
cesses in athletes in particular.
Physical activity results in both acute and long-standing effects
on pain perception [47]. In athletes, analgesia during and directly
after physical activity (so-called ‘‘acute exercise-induced analge-
sia’’) must be differentiated from a general alteration of pain per-
ception at rest.
Concerning acute effects of physical activity, there is consistent
evidence that following a bout of intense exercise, pain perception
is reduced for a limited time period [37]. In contrast, studies of pain
perception in athletes at rest (ie, long-standing effects) have re-
vealed contradictory results. Both elevated pain tolerance and pain
0304-3959/$36.00 Ó2012 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.pain.2012.03.005
⇑
Corresponding author. Address: Department of General Internal Medicine and
Psychosomatics, Medical Hospital, University of Heidelberg, Im Neuenheimer Feld
410, D-69120 Heidelberg, Germany. Tel.: +49 6221 56 37862; fax: +49 6221 56
8450.
E-mail address: jonas.tesarz@med.uni-heidelberg.de (J. Tesarz).
PAIN
Ò
153 (2012) 1253–1262
www.elsevier.com/locate/pain
threshold have been reported in some studies [24,25,69]. Other
studies demonstrate normal [57] or lower pain thresholds but ele-
vated pain tolerance [50]. In general, pain tolerance is only weakly
related to pain threshold [14].
These observations raise the question of whether athletes gen-
erally differ in pain perception from normally active persons.
Although evidence suggests that an athlete’s regular exposure to
painful training may contribute to an altered pain perception
[3,20,24,26,34,50,58,69,71], there is no consensus on the long-last-
ing effects of regular physical activity on pain perception.
This review examined whether athletes and normally active
controls differ in pain perception at rest. We summarize the cur-
rent evidence of differences in pain threshold and pain tolerance
of experimentally induced pain between athletes and normally ac-
tive controls.
2. Methods
2.1. Procedures
The review was performed according to the recommendations
of the Cochrane Collaboration [30] when appropriate, and is re-
ported after the PRISMA statement [41]. All steps and methods of
the review were specified in advance in a predetermined review
protocol developed within the RevMan software (Version 5.0; de-
tailed protocol is available from the corresponding author).
We searched Medline, EMBASE, Sport-Discus, Web of Science,
PsycINFO, and PSYNDEX. The search strategy was adapted for each
database if necessary (see the web Appendix for complete search
strategy). Additionally, congress abstracts and reference lists from
identified articles and reviews of all types of pain assessments in
athletes [32,33,37,47,63] were screened for published and unpub-
lished data, and all promising references were scrutinized. For
promising abstracts, complete publications were retrieved. In case
of identified unpublished data, authors were contacted for further
information. In addition, a citation search on the included articles
was performed. Searches were performed independently by 2
reviewers (J. Tesarz [J.T.] and A. Schuster [A.S.]). There were no
restrictions regarding language, publication status, and type of
publication. Finally, leading scientists in the field were contacted
to locate unpublished studies.
The 2 reviewers independently scanned the titles and abstracts
of eligible studies. Both reviewers independently scanned the full
text articles to determine whether the articles met the selection
criteria. Disagreements between the 2 reviewers were resolved
by discussion, and if agreements between the 2 reviewers could
not be achieved, a third reviewer was consulted.
2.2. Eligibility criteria
We selected all studies that compared experimentally induced
pain threshold or pain tolerance in athletes with normally active
controls. There is no standardized definition for athletes, and scien-
tists use a variety of different criteria for the definition of athletes
or nonathletes. Based on clinical considerations, this review ac-
cepted studies in which athletes were described as participating
in competitions or training for at least 3 hours per week. There is
evidence that ‘‘at least 30 minutes of regular, moderate-intensity
physical activity on most days of the week’’ reduces the risk of
multiple diseases [53,78]. By setting our cutoff point of at least
3 hours per week, we were orientating on these recommendations,
assuming that there is evidence for this amount of physical exer-
cise to have beneficial impact on health and well-being.
To exclude physically inactive sports (eg, mental exercises,
handicraft work), athletes must have been classified in at least
one of the following sport groups: endurance, game (including
dance), or strength. The decision to include dancers in our analysis
was primarily based on the consideration that dancing – character-
ized by physical and motor fitness skills – is commonly accepted as
an independent kind of sport in sport scientific literature [4,21,38].
Further, recent data emphasized the ‘‘physical load’’ of dance sport
(eg, [8]) and demonstrated similar energy expenditure values as in
‘‘classical’’ sport disciplines [1].
Controls were adequate when physical activity was <3 hours
per week with no active participation in any form of organized
sports. In addition, studies must have quantified pain threshold
or pain tolerance using continuous variables with means and SDs
or have provided sufficient information to reconstruct these values.
Repetitive pain measures are common in pain research. How-
ever, bias (eg, by the stimulation of descending inhibitory neuronal
circuits) induced by the effects of preexposure to pain testing itself
is a controversial and often discussed issue, especially in the field
of sport sciences [51]. Therefore, we used only the first measured
results for our meta-analysis in studies with repetitive measure-
ments to minimize potential pain test reactivity artifacts.
2.3. Outcomes
Both pain threshold and pain tolerance were analyzed to obtain
a better understanding of pain perception. Pain threshold is the
minimum intensity of a stimulus that is perceived as painful. Pain
tolerance is measured by the length of time an individual is willing
to endure a noxious stimulus or by the maximum stimulus inten-
sity that one will endure in a given situation [39].
2.4. Data collection
Two reviewers (J.T., A.S.) independently extracted data (study
characteristics, study results, and assessments of risk of bias) using
a prespecified data extraction form. All discrepancies were double
checked, and if disagreements between the 2 reviewers arose, a
third reviewer was consulted.
2.5. Assessment of risk of bias
Because no standardized criteria have been established to as-
sess the quality of (psychophysiological) studies on experimentally
induced pain, we developed a priori a standardized checklist of risk
of bias that was a modified version of the checklist developed by
Williamson [77].
Quality assessment was divided into 3 sections: patient sam-
pling, pain assessments used, and analysis. Similar to the QUA-
DAS-2 quality assessment for diagnostic accuracy studies [75],
signaling questions for each section were developed to assist judg-
ments. They were answered as ‘‘criteria met,’’ ‘‘criteria partially
met,’’ or ‘‘criteria not met.’’ Risk of bias was judged based on the
signaling questions as ‘‘low,’’ ‘‘moderate,’’ or ‘‘high’’ for each sec-
tion separately. For deciding on the overall study quality, the min-
imal judgment of all sections was used (according to [77]). This
means that only studies judged as being at low risk of bias on all
3 sections were classified as ‘‘low risk of bias studies,’’ and studies
being at moderate risk of bias or better for all 3 sections were clas-
sified as ‘‘moderate risk of bias studies.’’ All other studies were
deemed at ‘‘high risk of bias studies.’’ This procedure of quality
assessment is rather conservative compared to relying on sum
scores, and guarantees better validity and transparency [30].
2.6. Statistical analysis
We recoded pain measures in which higher scores indicated
higher pain thresholds or tolerance. Standardized mean differences
were calculated as Hedges’ g for pain threshold and pain tolerance
1254 J. Tesarz et al. / PAIN
Ò
153 (2012) 1253–1262
by considering the groups of athletes and normally active controls
for each study. To compare and combine the different studies, we
used a DerSimonian-Laird random-effects model [15] to calculate
pooled estimates with 95% confidence intervals (CI
95
). A random-
effects model was chosen because studies differed in the nature
of physical activity (eg, endurance sport, game sport, strength
sport), the type of noxious stimulus applied (eg, heat, cold, ische-
mic, pressure, or electrical) and other factors (eg, sex, age) [30].
Heterogeneity among the studies was described using the I
2
sta-
tistic, and I
2
values over 50% indicated strong heterogeneity [31].
We hypothesized that the characteristics of pain assessment (eg,
ischemia, cold, heat, pressure, or electrical) and the nature of phys-
ical activity (eg, endurance, game, or strength sports) may explain
the heterogeneity in the effect size. Therefore, in the case of at least
2 studies available, we conducted subgroup analyses of pain induc-
tion methods, types of athletic participation, and sex to identify
possible moderators.
Sensitivity analyses were performed to determine the effect of
outliers (by excluding every single study) and low quality studies
(by excluding studies with high risk of bias).
Cohen’s categories were used to evaluate the magnitude of the
effect size, with a g < 0.5 indicating a small effect size, g = 0.5–0.8
as a moderate effect size, and g > 0.8 as a large effect size [13].
Potential small-study bias (ie, the association of publication
probability with the statistical significance of study results) was
investigated using visual examination of the funnel plot (plots of
effect estimates against its standard error).
All calculations were performed in RevMan analyses software
(Version 5.0).
3. Results
3.1. Characteristics of included studies
An initial database search identified 1333 studies. After adjust-
ing for duplicates, 1152 studies remained. Of these studies, 1090
studies were discarded because after reviewing the abstracts, it ap-
peared to both reviewers that these papers clearly did not meet the
inclusion criteria. The full text of the remaining 62 citations was
examined in more detail (see the web Appendix); 46 studies were
excluded due to the following reasons. Forty-one studies did not
fulfill the predefined inclusion criteria. In addition, there was one
double publication [72,73], and 6 studies initially provided insuffi-
cient information. Of these 6 potentially relevant studies, we con-
tacted the corresponding authors for missing information. Three
authors provided additional data [25,59,60]. Two of these studies
met the inclusion criteria [25,59], but one study failed [60]. In the
remaining 3 studies, the authors were not available. These studies
were excluded because of insufficient information [49,56,79]. Two
measurements in one publication [34] that assessed pain percep-
tion in 2 different studies in 2 different populations were regarded
as independent units ([34] a,b). After this, 17 studies from 6 coun-
tries met the inclusion criteria. However, we excluded 2 studies
that assessed pain perception in athletes due to conceptual flaws.
In these studies, the most tolerant subjects were excluded prior
to analysis, which constituted a strong ‘‘floor effect’’ [9,35]. More-
over, we excluded the values for heat pain thresholds in an older
study by Ryan and Kovacic [57] because of an irreproducible trans-
formation of the original data by the authors. In addition to initial
database searching, 3 studies were identified based on abstracts
and reference lists and 2 by web search in the field of sport sciences.
Responsible authors were contacted. One provided additional data,
thus the study could be included (see also the web Appendix).
Finally, 15 studies that allocated 568 athletes (202 women, 366
men) and 331 normally active controls (147 women, 184 men)
provided sufficient data for our meta-analysis (Fig. 1). Eight of these
15 studies were conducted in the United States [19,20,34,57,
59,64,73], 2 of these studies were conducted in Canada [18,52],1
study was performed in Australia [24], and 4 studies were per-
formed in Europe [25,50,58,69]. Endurance athletes were included
in 9 studies [18,19,25,34,50,58,59,64]. Athletes that were active in
game sports were assessed in 8 studies [18–20,52,57,64,69,73],
and strength sports were studied in 2 studies [24,25] (Table 1).
3.2. Meta-analysis
3.2.1. Pain tolerance
Twelve studies reported pain tolerance outcomes (Fig. 2). The
average pain tolerance was significantly elevated in athletes with
a pooled standardized mean difference of 0.87 (CI
95
0.53–1.21;
I
2
= 76%). Five studies revealed statistically significant elevations
in pain tolerance in athletes [50,57,58,69,73], and 7 studies demon-
strated no significant differences [18–20,34,52,64]. A sensitivity
analysis to exclude studies with a high risk of bias did support
the assumption of an overall significant difference in pain tolerance
(Hedges’ g = 0.93, CI
95
0.52–1.34; I
2
= 73%; n = 7). The funnel plot
was symmetrical, with no evidence of relevant small study bias
(see the web Appendix). The exclusion of any single study did not
alter the magnitude or statistical results of the summary estimate.
Subgroup analyses (Table 2): Exploratory subgroup analyses
demonstrated that the effects on pain tolerance varied with differ-
ent types of physical activities, but not with different types of nox-
ious stimuli.
In endurance athletes (Hedges’ g = 0.65, CI
95
0.42–0.88; I
2
= 6%;
n = 8), pain tolerance was consistently characterized by a moderate
effect size and low heterogeneity, but in game sport athletes, the
effect was high and was characterized by high heterogeneity
(Hedges’ g = 0.98, CI
95
0.40–1.57; I
2
= 86%; n = 8). In strength
sports, only one study [20] was available and reported no differ-
ences between athletes and normally active controls (Hedges’
g = 0.07, CI
95
0.73-0.87).
Significant differences between athletes and normally active
controls were observed for both ischemic stimulation (Hedges’
g = 0.72, CI
95
0.33–1.10; I
2
= 42%; n = 4) and cold pain stimulation
(Hedges’ g = 0.73, CI
95
0.16–1.30; I
2
= 81%; n = 5). However, the
heterogeneity was high. With one study per type of stimulus, data
for electrical, heat, and pressure pain tolerance were not further
analyzed.
Separate analyses of male (Hedges’ g = 0.87, CI
95
0.40–1.35;
I
2
= 77%; n = 8) and female athletes (Hedges’ g = 1.18, CI
95
0.34–
1.68; I
2
= 84%; n = 3) showed significant sex differences.
3.2.2. Pain threshold
Nine studies reported data on differences in pain threshold be-
tween athletes and normally active controls (Fig. 3). The pooled
standardized mean difference was 0.69 with a CI
95
0.16–1.21. Five
studies showed a statistically significant elevation of pain thresh-
old in athletes [24,25,34,69], and 4 studies showed no significant
differences [50,58,59,73]. One of the studies [50] showed an effect
size in the opposite direction (Hedges’ g = 0.57) and another
study was characterised by an extraordinary high effect size of
2.22 [24]. The exclusion of any single study did not alter the mag-
nitude or the statistical result of the summary estimate. A sensitiv-
ity analysis that excluded studies with a high risk of bias did not
support the assumption of an overall significant difference in pain
threshold (Hedges’ g = 0.75, CI
95
0.02-1.53; I
2
= 90%; n = 6). Due
to the limited number of studies (<10), we did not perform a funnel
plot for pain threshold [65].
3.2.2.1. Subgroup analyses (Table 3). Assessment methods and
the nature of physical activity varied across studies. Therefore,
J. Tesarz et al. / PAIN
Ò
153 (2012) 1253–1262 1255
exploratory subgroup analyses were performed. Cold pain thresh-
old was assessed in 3 studies [34,69], and ischemic pain threshold
was measured in 4 studies [34,50,58]. Electrical pain threshold was
measured in 2 studies [25,73]. Subgroup analyses for different
stimuli resulted in a reduction in heterogeneity (0.0% to 56%). No
significant differences between athletes and normally active con-
trols were observed for ischemic (Hedges’ g = 0.24, CI
95
0.79-
0.32; I
2
= 52%; n = 2) or electrical stimulation (Hedges’ g = 0.80,
CI
95
0.19-1.78; I
2
= 56%; n = 2). Significant differences were ob-
served in cold-pressor pain thresholds (Hedges’ g = 1.00, CI
95
0.70–1.29; I
2
= 0%; n = 3).
Endurance sports were assessed in 6 studies [24,34,50,58,59].
Game sports were measured in 2 studies [69,73], and strength
sports were measured in one study [24]. Subgroup analyses of
endurance athletes showed no significant differences (Hedges’
g = 0.35, CI
95
0.14-0.84; I
2
= 73%), but the 2 studies that assessed
game sports (Hedges’ g = 0.82, CI
95
0.10–1.55; I
2
= 76%) and the
single study that assessed strength sports (Hedges’ g = 2.22, CI
95
1.56–2.87) demonstrated significant differences between athletes
and normally active controls.
A separate analysis of males and females yielded no signifi-
cant effects in the male subgroup (Hedges’ g = 0.51, CI
95
0.16-1.19; I
2
= 82%; n = 5), but female athletes were character-
ized by significantly heightened pain thresholds compared to
normally active controls (Hedges’ g = 0.56, CI
95
0.04–1.08;
I
2
= 50%; n = 3).
4. Discussion
The present study analyzed whether differences in pain percep-
tion exist between athletes and normally active persons. We con-
ducted a meta-analysis of pain tolerance and pain threshold as
essential characteristics of pain perception.
The most important finding was that pain perception differed in
athletes compared to normally active controls. Athletes possessed
consistently higher pain tolerance than normally active controls.
However, the available data on pain thresholds were rather sparse
and less uniform.
The generally heightened pain tolerance in athletes was consis-
tently characterized by moderate to large effect sizes (Fig. 2). These
effects remained significant when only high-quality studies were
included in the analysis. These results support the hypothesis that
athletes generally possess elevated pain tolerance. Compared to ef-
fect size estimates in other experimental pain studies showing only
low-to-moderate effects by gender or genetic variants [16], the ef-
fects reported in our analysis were strong and exceeded effect sizes
of most classical pain interventions [17,36,61]. The data emphasize
a possible impact as well as clinical relevance of athletic status on
pain tolerance.
It is postulated that pain threshold is relatively constant in an
individual, but pain tolerance is strongly modulated by psycholog-
ical and psychosocial factors [11,27,42,54,55,62]. Coping skills can
increase pain control [6,22,23,68,76]. For example, self-efficacy and
Fig. 1. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram. Study selection process.
1256 J. Tesarz et al. / PAIN
Ò
153 (2012) 1253–1262
Table 1
Characteristics of the studies.
Study Assessed
outcomes
Assessment tools No of
athletes
(f/m)
No of
controls
(f/m)
Characteristics of study populationand authors’ conclusion Risk of bias
Egan 1987 [18] Tolerance CPT (time in min to
withdrawal)
50 (male) 10 (male) Athletes were randomly chosen from inter-university athletes (M = 22 y; football, boxing, fencing, karate and cross-
country skiing). A control group of nonathletes (M = 22 y) was also included in the experiment. All the subjects in the
experimental group trained for at least 3 90 min/wk and competed on a regular basis during their competitive
season. Excluded from the study were subjects who had frostbite, cardiac disorders, poor circulation, hypertensive
disease, or who were frequent recipients of cryotherapy. The cultural background and pain status at rest or during
activity of the subjects were not controlled
Athletes tolerated significant more pain than normally active persons. In particular, football players and the cross-
country skiers tolerated more pain than karate and fencers. These findings do not support the hypothesis that contact
sports tolerate more pain than noncontact sports
£75.0%
(92.0/62.5/
62.5)
Moderate
Eitter 1980 [19] Tolerance Pressure in mm Hg
(gross pressure device
according to Ryan)
Ischemia
(number of fist
contractions)
29 (male) 16 (male) Subjects were male university students. They were classified into 3 groups: 1) contact sports athletes (varsity football
players); 2) endurance sports athletes (cross-country runners), and 3) nonathletes. These 2 groups of athletes
participated in the study while they were not in their competitive season. The nonathletes were drawn from a physical
activity class in personal fitness. Only students who had not participated in any form of regular physical exercise or
varsity athletics were used in this sample. No information is given about age, current or past pain experiences
There were significant differences in gross pressure pain tolerance among these 3 groups: contact athletes tolerated
more pain, whereas endurance athletes tolerated less pain than controls. No differences in ischemic pain tolerance
were found among the 3 groups
£50.0%
(66.7/50.0/
25.0)
High
Ellison and
Freischlag
1975 [20]
Tolerance Finger flexion (Time of
performance in s)
72 (male) 12 (male) Undergraduate male students were selected randomly from intercollegiate athletic teams (baseball, basketball,
football, track distance/sprint) and from the nonathlete male population at the university. Subjects were tested
individually, told only that pain tolerance was the factor being investigated, and had no knowledge that comparisons
were to be made among athletes and nonathletes. No information is given about age, or current or past pain
experiences
No differences in pain tolerance were found among groups
£61.0%
(75.0/50.0/
63.0)
Moderate
Granges and
Littlejohn
1993 [24]
Threshold Pressure (in kg/cm at
first painful sensation)
30 (26/4) 30 (27/3) Healthy controls (M = 32 y) exercising less than 2 h/wk were selected from 2 different hospital physiotherapy staffs.
Exercising subjects (M = 36 y) were selected from different fitness centers. To enter the study, the mean exercise time
(regular aerobic/strength exercises) was set at more than 6 h/wk. While these patients were not formally assessed for
aerobic fitness, it was considered that they represented the upper echelon of the exercising public. No subject had
reported any significant musculoskeletal pain in the 6 months before assessment
The fit subjects significantly differed from the unfit for pain threshold
£82.1%
(100/75.0/
62.5)
Moderate
Guieu et al. 1992
[25]
Threshold Electrical stimulation
in mAmp for leg flexion
nociceptive reflex
threshold
6 (2/4) 8 (1/7) Controls consisted of subjects (M = 24 y), who had undergone no intensive sports training in the past; athletes
(M = 22 y; runner, bodybuilder, weightlifter) regularly participated in national or international sporting events and
underwent 1 or 2 intensive training sessions every day. No information is given about current or past pain experiences
The only study using objective measurement tools (leg flexion reflex threshold).The nociceptive flexion reflex threshold
at rest was found to be significantly higher in athletes than in the controls
£71.4%
(75.0/62.5/
75.0)
Moderate
Janal et al. 1994
(part I) [34]
Threshold
tolerance
Ischemia(time in s to
first painful sensation/
withdrawal) CPT (time
in s to first painful
sensation/withdrawal)
12 (male) 18 (male) Post hoc analysis of old data (the runners’ data were from the prerun condition of an exercise-induced analgesia study;
control subjects’ data were from the placebo condition of a drug study): controls consisted of subjects (M = 27 y) who
did not participate in a regular aerobic training program. Athletes consisted of male runners (M = 39 y) training an
average of 69.2 km/wk. For the CPT, 4 additional runners (M = 44 y: mean training: 33.0 km/wk) were recruited. All
subjects were paid for their participation. No information is given about current or past pain experiences
These data indicate that runners are less sensitive than controls only to noxious cold, but more sensitive to heat
stimulation. Signal detection theory measurements demonstrated that runners discriminated noxious thermal stimuli
better than controls. Thus, runners do not appear to be generally ‘‘stoical’’
£54.0%
(75.0/50.0/
25.0)
High
Janal et al. 1994
(part II) [34]
Threshold
tolerance
Ischemia (time in s to
first painful sensation/
withdrawal) CPT (time
in s to first painful
sensation/withdrawal)
36 (male) 24 (male) Athletes consisted of male runners (M = 30 y) who had been training at least 30 km/wk for at least 3 mo. Controls
consisted of subjects (M = 28 y) who were not engaged in any regular aerobic training program. All subjects were paid
for their participation. No information is given about current or past pain experiences
Regular athletic training seems to produce effects similar to cold acclimatization, blunting the sensory response to
noxious cold. In general, however, runners do not appear to be any less sensitive to noxious stimulation than normally
active controls
£57.0%
(75.0/25.0/
62.5)
High
Ord and Gijsbers
2003 [50]
Threshold
tolerance
Ischemia (time in min
to first painful
sensation/withdrawal)
20 (male) 20 (male) The group of athletes (M = 20 y) engaged in training for competitive rowing; controls (M = 22 y) were not training for any
specific sport at the time of testing. No information is given about current or past pain experiences Pain tolerance, but not
threshold, was significantly higher for athletes. Pain tolerance was correlated with the number and quality of coping
strategies used during testing. Pain threshold of the control group was characterized by high variability
£71.4%
(66.7/75.0/
75.0)
Moderate
(continued on next page)
J. Tesarz et al. / PAIN
Ò
153 (2012) 1253–1262 1257
Table 1 (continued)
Study Assessed
outcomes
Assessment tools No of
athletes
(f/m)
No of
controls
(f/m)
Characteristics of study populationand authors’ conclusion Risk of bias
Paparizos et al.
2005 [52]
Tolerance CPT (time in s to
withdrawal)
47
(female)
26
(female)
The group of athletes consisted of conveniently sampled women dancers from the Queen’s Dance Club (mean training’s
history of 10 y; M = 20 y), controls were conveniently sampled from the Queen’s University Undergraduate Psychology
Subject Pool (M = 19 y). No restriction in age or dance ability was enforced within the dance sample, while the control
group consisted of women who had never experienced professional training. Anyone with history of cardiac disease or
extreme hypertension, or anyone who had experienced frostbite to their nondominant hand was excluded from the
study. No information is given about current or past pain experiences
Dancers had significantly higher pain tolerance than controls. High-skill dancers had significantly higher pain tolerance
than low-skill dancers. No difference in pain sensitivity was found. Despite an ability to withstand higher amounts of
pain, dancers still found it as painful as controls but were able to manage longer exposure
£71.4%
(83.3/75.0/
50.0)
Moderate
Ryan and Kovacic
1966 [57]
Tolerance Pressure (mm Hg at
withdrawal) Ischemia
(number of
contractions = time in
s)
37 (male) 18 (male) Subjects were male university students and based on a questionnaire classified into 3 groups: 1) contact sports athletes
(boxing, football, or wrestling); 2) endurance sports athletes (golf, tennis, track), and 3) nonathletes (physical
inactive).The subjects assumed they had been randomly selected, and were unaware that athletic participation was a
factor or that the experiment was in any way related to the questionnaire administered earlier. No information is given
about age, current or past pain experiences
There were no significant differences between groups (contact/noncontact/control) in pain threshold, but a highly
significant difference between groups in pain tolerance, wherein the contact athletes tolerated more pain than the
noncontact athletes, who in turn tolerated more pain than the normally active persons
£50.0%
(58.3/37.5/
50.0)
High
Scott and
Gijsbers 1981
[58]
Threshold
tolerance
Ischemia (numbers of
fist contraction to first
pain sensation/
withdrawal)
60 (31/29) 26 (16/10) The group of athletes consisted of swimmers (highly conditioned swimmers of the Scottish national squad and
competitive club swimmers). Controls were made up of undergraduate students of whom none had experience of sport
at a competitive level. No information is given about age, current or past pain experiences
While pain threshold showed little difference between the groups, pain tolerance was significantly different. Pain
tolerance of the competitive swimmers varied according to the stage of the training season (so the author suggested
that the enhanced pain tolerance of the competitive swimmers would seem to lie in their systematic exposure to brief
periods of intense pain)
£78.6%
(67.0/75.0/
100)
Moderate
Smith 2004 [59] Threshold Heat (temperature of
first painful sensation)
26 (13/13) 26 (13/13) The subject population was composed of Haverford College students, including an equal number of male and females in
each category. Athletes were recruited from the Haverford track team and nonathletes were recruited from the general
campus population via signs posted around campus and in athletic buildings. Subjects who answered yes to any
questions of diseases or pain were excluded. No information is given about age. Eligible subjects were monetarily
compensated at the conclusion of their participation
There were no differences between athletes and normally active persons (or sexes) in heat pain threshold, but an
interaction between sex and athletic status on cold-pressor ratings (‘‘sensitivity’’) could be observed: female athletes’
pain ratings were significantly lower than ratings in normally active females, but male athletes did not differ from
normally active male subjects
£39.3%
(50.0/37.5/
25.0)
High
Sternberg et al.
1998 [64]
Tolerance Heat (time in s to
withdrawal)
67 (33/34) 20 (14/6) The group of athletes consisted of active members of college teams (NCAA Division III; basketball, fencers, track
athletes). Controls were recruited from an introductory psychology class. All subjects (athletes and nonathletes) were
given the option of participating in the study for course credit (if applicable) or for potential monetary compensation.
No information is given about age, current or past pain experiences
There were no significant differences between athletes’ pain values. But the results of this and normally active persons’
baseline suggest that athletic competition modulates responses to noxious stimuli. The degree and direction of this
modulation (inhibition or potentiation) depended on the pain test, the body region tested, and the sport in question
£50.0%
(62.0/25.0/
87.5)
High
Tajet-Foxell and
Rose 1995
[69]
Threshold
tolerance
CPT (time in min to
first painful sensation/
withdrawal)
52 (26/26) 53 (26/27) Athletes consisted of professional ballet dancers from a national ballet company (M = 25 y), controls were selected from
students at a university institution (M = 24 y). No information is given about current or past pain experiences
Dancers were found to have significantly higher pain threshold and pain tolerance than age-matched controls in the
CPT. They also reported a more acute experience of the sensory aspects of pain
71.4%
(66.7/75.0/
75.0)
Moderate
Walker 1971 [73] Threshold
tolerance
Electrical stimulation
(mAmp)
24
(female)
24
(female)
Female college students (18–24 y) enrolled in 2 colleges served as subjects. Athletes consisted of active members of the
varsity basketball teams; controls (selected from required physical education classes), who had never participated in
any type of athletic competition on an inter-school basis, served as a nonathletic group. Each had an A health rating
from her college medical service. No information is given about current or past pain experiences
Athletes showed significantly higher pain tolerance than normally active persons. Threshold did not vary by group
classification
78.6%
(91.7/75.0/
62.5)
Moderate
CPT, cold pressor test; m, male; f, female; N, number; y, years (range of age), M, mean age; Risk of bias, The quality scoring was divided into 3 sections (patient sampling/pain assessments used/analysis). Each article was than
graded as a low risk of bias (80% or above for all 3 sections), moderate risk of bias (50% or above for all 3 sections), or high risk of bias (scored <50% for any one section) study (for further information, see text).
1258 J. Tesarz et al. / PAIN
Ò
153 (2012) 1253–1262
pain acceptance are significant correlates of pain tolerance [5,43].
Athletes are frequently exposed to unpleasant sensory experiences
during their daily physical efforts, and high physical and psycho-
logical resistances must be overcome during competitions or very
exhausting activities. However, athletes are forced to develop effi-
cient pain-coping skills because of their systematic exposure to
brief periods of intense pain. Therefore, pain coping is an integral
part of athletic training, and coping skills are important features
in the development of athletic character [40]. Moreover, the men-
tal attitude of athletes towards pain and physical discomfort signif-
icantly differs from that of normally active controls [45,50].
Overall, psychological factors play an important role in pain toler-
ance in athletes. However, we can only hypothesize that these psy-
chological factors may be responsible for the observed effects
because they were not directly evaluated.
Available data on pain thresholds do not give a uniform picture.
Although data on pain thresholds demonstrated significant
differences in pooled analysis, there are some arguments that
extenuate the explanatory power of these results. First, as the
number of studies and athletes are rather low, the power of the
analysis may be insufficient and there is a serious risk of random
error [74]. Second, no differences between the groups were found
after exclusion of high risk of bias studies. In addition, the interpre-
tation of a consistent effect requires caution because of the very
high heterogeneity of these results (Fig. 3), which includes even
inconsistent directions in effect sizes. Some may argue that in such
a situation, pooling of effects should not be undertaken [30].In
sum, although it appears that athletes possess higher pain thresh-
olds, closer consideration reveals that the present situation is
rather conflicting, and further studies are needed to highlight pain
threshold differences in athletes and nonathletes.
The findings that regular exercise was clearly associated with
higher pain tolerance, but pain detection thresholds were affected
more ambiguously, corresponds with some clinical observations.
Fig. 2. Overall effect on pain tolerance. Standard mean differences (SMDs) were calculated as Hedges’ g. A DerSimonian-Laird random-effects model was used to calculate
pooled estimates with 95% confidence intervals (95% CI). Heterogeneity among the studies was described using the I
2
statistic.
Table 2
Exploratory subgroup analyses for pain tolerance.
Outcome title Number of
studies
Number of subjects
(athletes/controls)
Effect size (Hedges’ g)
95% CI
Test for overall
effect P-value
Heterogeneity
I
2
(%);
s
2
Induction stimulus
Cold 5 201/131 0.73 (0.16–1.30) 0.01 81; 0.33
Ischemia 4 181/74 0.72 (0.33–1.10) 0.003 42; 0.06
Electrical 1 24/24 1.36 (0.73–1.99) <0.001
Heat 1 67/20 0.50 (0.00–1.01) 0.05
Pressure 1 27/18 2.45 (1.65–3.25) <0.001
Kind of physical activity
Game sport 8 240/179 0.98 (0.40–1.57) 0.001 86; 0.60
Endurance sport 8 198/146 0.65 (0.42–0.88) <0.001 6; 0.01
Strength sport 1 12/12 0.07 (0.73-0.87) 0.86
Sex
Female 3 97/76 1.18 (0.34–1.68) 0.006 84; 0.46
Male 8 276/145 0.87 (0.40–1.35) <0.001 77; 0.36
Risk of bias
Moderate/low 7 325/171 0.93 (0.52–1.34) <0.001 73; 0.22
High 5 175/96 0.80 (0.17–1.43) 0.01 81; 0.42
Exploratory subgroup analyses were performed for pain induction methods, types of athletic participation, sex and risk of bias to identify possible
moderators of heterogeneity. Effect sizes were calculated as Hedges’ g. A DerSimonian-Laird random-effects model was used to calculate pooled estimates
with 95% confidence intervals (95% CI). Heterogeneity among the studies was described using the I
2
statistic.
J. Tesarz et al. / PAIN
Ò
153 (2012) 1253–1262 1259
Numerous studies of the effect of physical exercise in pain patients
demonstrate a consistent impact on quality of life and functioning
without an effect on pain scores [48,70]. The assumption that
physical activity mainly affects pain tolerance and omits most
modalities of pain thresholds could be a coherent explanation.
Therefore, it may be advisable to expect the restoration of function,
quality of life, and the development of pain-coping skills rather
than the direct alleviation of pain threshold in exercise treatment
in pain patients.
Although the primary purpose of this review was to summarize
all commonly used pain induction methods to show general alter-
ations in pain perception, the exploratory subgroup analyses may
shed further light on the underlying mechanisms. Subgroup analy-
ses showed that the pain induction method and the nature of phys-
ical activity may contribute to the heterogeneity. This observation
requires further attention.
Exploratory subgroup analyses for pain tolerance (Table 2)
demonstrated high heterogeneity in all subgroups stratified by
pain induction methods. Surprisingly, stratification by the nature
of physical activity revealed that the subgroup of endurance ath-
letes was characterized by low heterogeneity. This result suggests
that endurance athletes form a more homogenous population,
while athletes involved in game sports do not represent a homoge-
neous self-contained group but offer a wide range of different clus-
ters of athletes with different physical and psychological profiles.
Exploratory subgroup analyses for pain threshold (Table 3) sug-
gest that heterogeneity may be explained by the difference in pain
induction methods.
Noteworthy, pain induction by cold was characterized by
strong and consistent findings [34,69], which raises the question
of underlying mechanisms. Previous findings show that cold-pres-
sor pain stimulates the autonomic nervous system [44] and
induces pain-modulating processes, such as the hypothalamic-
pituitary-adrenal axis [67] and baroreflex-mediated analgesia
[2,10,66]. As baroreflex sensitivity and autonomic nervous system
affect pain processing [12], physical activity may act as a mediator,
resulting in adaptations of the autonomic nervous system, which
for its part may selectively contribute to the alterations in cold-
pressor pain [46]. However, this hypothesis has never been
directly targeted, and is based on only 3 studies within our review.
Therefore, it remains speculative and has to be supported with fur-
ther data.
Fig. 3. Overall effect on pain threshold. Standard mean differences (SMDs) were calculated as Hedges’ g. A DerSimonian-Laird random-effects model was used to calculate
pooled estimates with 95% confidence intervals (95% CI). Heterogeneity among the studies was described using the I
2
statistic.
Table 3
Exploratory subgroup analyses for pain threshold.
Outcome title Number of
studies
Number of subjects
(athletes/controls)
Effect size (Hedges’ g)
95% CI
Test for overall
effect P-value
Heterogeneity I
2
(%);
s
2
Induction stimulus
Cold 3 104/95 1.00 (0.70–1.29) <0.001 0; 0.00
Ischemia 2 80/46 0.24 (0.79-0.32) 0.40 52; 0.09
Electrical 2 30/32 0.80 (0.19-1.78) 0.11 56; 0.31
Heat 1 26/26 0.06 (0.46-0.63) 0.76
Pressure 1 30/30 2.22 (1.56–2.87) <0.001
Kind of physical activity
Game sport 2 76/77 0.82 (0.10–1.55) 0.03 76; 0.21
Endurance sport 6 138/96 0.35 (0.14-0.84) 0.16 73; 0.26
Strength sport 1 30/30 2.22 (1.56–2.87) <0.001
Sex
Female 3 63/63 0.56 (0.04–1.08) 0.04 50; 0.11
Male 5 111/102 0.51 (0.16-1.19) 0.13 82; 0.48
Risk of bias
Moderate/low 6 192/161 0.75 (0.02-1.53) 0.06 90; 0.82
High 3 78/68 0.56 (0.05–1.07) 0.003 55; 0.40
Exploratory subgroup analyses were performed for pain induction methods, types of athletic participation, sex and risk of bias to identify possible
moderators of heterogeneity. Effect sizes were calculated as Hedges’ g. A DerSimonian-Laird random-effects model was used to calculate pooled estimates
with 95% confidence intervals (95% CI). Heterogeneity among the studies was described using the I
2
statistic.
1260 J. Tesarz et al. / PAIN
Ò
153 (2012) 1253–1262
4.1. Limitations
Some limitations of our review should be considered. A primary
limitation that is common to all meta-analyses is the dependency
of the analysis on the quality and number of the studies on which it
is based. Detailed analysis of study quality revealed some serious
deficits in all phases of study conduct. As pain threshold results de-
pend on study quality in our meta-analysis, results on pain thresh-
old should be interpreted with caution, and further research is
necessary to determine this question. Moreover, the fact that there
are actually no ‘‘low risk of bias’’ studies emphasizes the need for
well-designed studies using standardized techniques and well-
defined populations. A second limitation is that pain research in
athletes is primarily based on laboratory pain tests. However,
laboratory pain tests do not necessarily represent ‘‘natural’’ pain
stimuli. Therefore, the transfer of our results to ‘‘real life’’ must
be considered with caution because the available data were gener-
ally collected under controlled conditions.
Another important limitation of this review is that causal and
mechanism-based interpretations were not possible within this
meta-analysis. Whether athletes acquire the ability to tolerate pain
because they engage in physical activity, or whether they engage in
physical activity because they can more easily tolerate pain re-
quires further analysis. To date, there has been only one controlled
interventional study that directly examined the effect of physical
activity on pain tolerance in healthy subjects [3]. Demonstrating
an elevation of pain tolerance after 12 weeks of aerobic training,
these data support the concept of direct alterations in pain pro-
cessing by physical exercise and indicate that alterations in pain
perception are not restricted to high-performance athletes, but
are also observable in ‘‘activated’’ normally active controls. Further
evidence comes from a study by Scott and Gijsbers [58], who dem-
onstrated that pain tolerance in competitive swimmers varied sig-
nificantly according to the stage of the training season. In contrast,
studies exploring specifically the hypothesis of an ‘‘athletic selec-
tion process’’ of pain-insensitive individuals are totally missing.
In general, direct evidence is lacking and further studies are
needed.
Nevertheless, our review provides some essential information
for further research. Studies of pain perception in athletes require
a critical appraisal of methodological quality from pain induction
procedure, over patient sampling methods, to data analysis. Fur-
thermore, future research should focus on the underlying mecha-
nisms to identify the involved psychological factors and
neurobiological processes. Controlled interventional studies are
also required to directly compare the effect of different types of
physical activity on pain perception in pain patients.
4.2. Conclusions
In conclusion, our data indicate that regular physical activity is
associated with specific alterations in pain perception. The effects
are generalized over different types of physical activities and pain
assessment methods, which suggest a common trend in pain per-
ception in athletes. Nevertheless, there is also evidence for differ-
ential effects in pain tolerance and pain threshold, with pain
threshold showing more ambiguous results. This observation is
consistent with the concept that pain perception is modifiable by
physical activity, which provides promise for patients with chronic
pain conditions for the use of noninvasive methods with few side
effects. However, further research is required to clarify the exact
relationship between physical activity and modifications in pain
perception and to identify underlying mechanisms. These findings
emphasize the potential role of physical exercise in the manage-
ment of pain.
Conflicts of interest statement
The submitted manuscript does not contain information about
medical device(s)/drug(s). There are no conflicts of interest. No
benefits in any form have been or will be received from a commer-
cial party directly or indirectly related to the subject of this
manuscript.
Acknowledgments
The authors thank Dr. rer. nat. Gerta Rücker from the Institute
of Medical Biometry and Medical Informatics of University Medical
Center Freiburg for her helpful comments.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.pain.2012.03.005.
References
[1] Ainsworth BE, Haskell WL, Herrmann SD, Meckes N, Bassett Jr DR, Tudor-Locke
C, Greer JL, Vezina J, Whitt-Glover MC, Leon AS. 2011 Compendium of physical
activities: a second update of codes and MET values. Med Sci Sports Exerc
2011;43:1575–81.
[2] al’Absi M, Petersen KL. Blood pressure but not cortisol mediates stress effects
on subsequent pain perception in healthy men and women. Pain
2003;106:285–95.
[3] Anshel MH, Russell KG. Effect of aerobic and strength training on pain
tolerance, pain appraisal and mood of unfit males as a function of pain
location. J Sports Sci 1994;12:535–47.
[4] Association of the IOC Recognized International Sports Federations (ARISF),
<http://www.arisf.org> [accessed 31.12.11].
[5] Baker SL, Kirsch I. Cognitive mediators of pain perception and tolerance. J Pers
Soc Psychol 1991;61:504–10.
[6] Birrer D, Morgan G. Psychological skills training as a way to enhance an
athlete’s performance in high-intensity sports. Scand J Med Sci Sports
2010;20:78–87.
[7] Bouwmeester W, van Enst A, van Tulder M. Quality of low back pain guidelines
improved. Spine 2009;34:2562–7.
[8] Bria S, Bianco M, Galvani C, Palmieri V, Zeppilli P, Faina M. Physiological
characteristics of elite sport-dancers. J Sports Med Phys Fitness
2011;51:194–203.
[9] Broucek MW, Bartholomew JB, Landers DM, Linder DE. The effects of relaxation
with a warning cue on pain tolerance. J Sport Behav 1993;16:239–50.
[10] Bruehl S, Chung OY. Interactions between the cardiovascular and pain
regulatory systems: an updated review of mechanisms and possible
alterations in chronic pain. Neurosci Biobehav Rev 2004;28:395–414.
[11] Chen AC, Dworkin SF, Haug J, Gehrig J. Human pain responsivity in a tonic pain
model: psychological determinants. Pain 1989;37:143–60.
[12] Chung OY, Bruehl S, Diedrich L, Diedrich A, Chont M, Robertson D. Baroreflex
sensitivity associated hypoalgesia in healthy states is altered by chronic pain.
Pain 2008;138:87–97.
[13] Cohen J. Statistical power analysis for the behavioral sciences. 2nd
ed. Hillsdale, NJ: Lawrence Erlbaum Associates; 1988.
[14] Defrin R, Shramm L, Eli I. Gender role expectations of pain is associated with
pain tolerance limit but not with pain threshold. Pain 2009;145:230–6.
[15] DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials
1986;7:177–88.
[16] Doehring A, Kusener N, Fluhr K, Neddermeyer TJ, Schneider G, Lotsch J. Effect
sizes in experimental pain produced by gender, genetic variants and
sensitization procedures. PLoS One 2011;6:e17724.
[17] Eccleston C, Palermo TM, Williams AC, Lewandowski A, Morley S.
Psychological therapies for the management of chronic and recurrent pain in
children and adolescents. Cochrane Database Syst Rev 2009;2:CD003968.
[18] Egan S. Acute-pain tolerance among athletes. Can J Sport Sci 1987;12:175–8.
[19] Eitter TJ. Pain tolerance training applied to the athletic environment. Diss
Abstr Int 1980;41:2005–6.
[20] Ellison K, Freischlag J. Pain tolerance, arousal, and personality relationships of
athletes and nonathletes. Res Q 1975;46:250–5.
[21] Ewalt KL. Athletic training in dance medicine and science. J Dance Med Sci
2010;14:79–81.
[22] Fernandez E, Turk DC. The utility of cognitive coping strategies for altering
pain perception: a meta-analysis. Pain 1989;38:123–35.
[23] Geisser ME, Robinson ME, Pickren WE. Differences in cognitive coping
strategies among pain-sensitive and pain-tolerant individuals on the cold-
pressor test. Behav Ther 1992;23:31–41.
J. Tesarz et al. / PAIN
Ò
153 (2012) 1253–1262 1261
[24] Granges G, Littlejohn GO. A comparative study of clinical signs in fibromyalgia/
fibrositis syndrome, healthy and exercising subjects. J Rheumatol
1993;20:344–51.
[25] Guieu R, Blin O, Pouget J, Serratrice G. Nociceptive threshold and physical
activity. Can J Neurol Sci 1992;19:69–71.
[26] Hall EG, Davies S. Gender differences in perceived intensity and affect of pain
between athletes and nonathletes. Percept Mot Skills 1991;73:779–86.
[27] Harkins SW, Price DD. Assessment of pain in the elderly. In: Turk DC, Melzack
R, editors. Handbook of pain assessment. New York: Guilford Press; 2001. p.
317–21.
[28] Hasselström J, Liu-Palmgren J, Rasjö-Wrååk G. Prevalence of pain in general
practice. Eur J Pain 2002;6:375–85.
[29] Hauser W, Thieme K, Turk DC. Guidelines on the management of fibromyalgia
syndrome – a systematic review. Eur J Pain 2010;14:5–10.
[30] Higgins J, Green S, editors. Cochrane handbook for systematic reviews of
interventions, version 5.1.0 [updated March 2011]. The Cochrane
Collaboration, 2009. Available from: <www.cochrane-handbook.org>.
[31] Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat
Med 2002;21:1539–58.
[32] Hoffman MD, Lee J, Zhao H, Tsodikov A. Pain perception after running a 100-
mile ultramarathon. Arch Phys Med Rehabil 2007;88:1042–8.
[33] Janal MN. Pain sensitivity, exercise and stoicism. J R Soc Med 1996;89:376–81.
[34] Janal MN, Glusman M, Kuhl JP, Clark WC. Are runners stoical? An examination
of pain sensitivity in habitual runners and normally active controls. Pain
1994;58:109–16.
[35] Jaremko ME, Silbert L, Mann T. The differential ability of athletes and
nonathletes to cope with two types of pain: a radical behavioral model.
Psychol Rec 1981;31:265–75.
[36] Keller A, Hayden J, Bombardier C, van Tulder M. Effect sizes of non-surgical
treatments of non-specific low-back pain. Eur Spine J 2007;16:1776–88.
[37] Koltyn KF. Analgesia following exercise: a review. Sports Med 2000;29:85–98.
[38] Koutedakis Y, Jamurtas A. The dancer as a performing athlete: physiological
considerations. Sports Med 2004;34:651–61.
[39] Loeser JD, Treede RD. The Kyoto protocol of IASP basic pain terminology. Pain
2008;137:473–7.
[40] Meyers MC, Stewart CC, Laurent CM, Leunes AD, Bourgeois AE. Coping skills of
olympic developmental soccer athletes. Int J Sports Med 2008;29:987–93.
[41] Moher D, Liberati A, Tetzlaff J. For the PRISMAGroup. Preferred reporting items
for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern
Med 2009;151:1–7.
[42] Moloney NA, Hall TM, O’Sullivan TC, Doody CM. Reliability of thermal
quantitative sensory testing of the hand in a cohort of young, healthy adults.
Muscle Nerve 2011;44:547–52.
[43] Motl RW, Gliottoni RC, Scott JA. Self-efficacy correlates with leg muscle pain
during maximal and submaximal cycling exercise. J Pain 2007;8:583–7.
[44] Mourot L, Bouhaddi M, Regnard J. Effects of the cold pressor test on cardiac
autonomic control in normal subjects. Physiol Res 2009;58:83–91.
[45] Nicholls AR, Polman RC. Coping in sport: a systematic review. J Sports Sci
2007;25:11–31.
[46] Nielsen CS, Stubhaug A, Price DD, Vassend O, Czajkowski N, Harris JR.
Individual differences in pain sensitivity: genetic and environmental
contributions. Pain 2008;136:21–9.
[47] O’Connor PJ, Cook DB. Exercise and pain: the neurobiology, measurement, and
laboratory study of pain in relation to exercise in humans. Exerc Sport Sci Rev
1999;27:119–66.
[48] Oesch P, Kool J, Hagen KB, Bachmann S. Effectiveness of exercise on work
disability in patients with non-acute non-specific low back pain: systematic
review and meta-analysis of randomised controlled trials. J Rehabil Med
2010;42:193–205.
[49] Oktedalen O, Solberg EE, Haugen AH, Opstad PK. The influence of physical and
mental training on plasma beta-endorphin level and pain perception after
intensive physical exercise. Stress Health 2001;17:121–7.
[50] Ord P, Gijsbers K. Pain thresholds and tolerances of competitive rowers and
their use of spontaneous self-generated pain-coping strategies. Percept Mot
Skills 2003;97:1219–22.
[51] Padawer WJ, Levine FM. Exercise-induced analgesia: fact or artifact? Pain
1992;48:131–5.
[52] Paparizos AL, Tripp DA, Sullivan MJL, Rubenstein ML. Catastrophizing and pain
perception in recreational ballet dancers. J Sport Behav 2005;28:35–50.
[53] Pate RR, Pratt M, Blair SN, Haskell WL, Macera CA, Bouchard C, Buchner D,
Ettinger W, Heath GW, King AC, Kriska A, Leon AS, Marcus BH, Morris J,
Pfaffenbarger RS, Patrick K, Pollock ML, Rippe JM, Sallis J, Willmore JH. Physical
activity and public health. A recommendation from the centers for disease
control and prevention and the American college of sports medicine. JAMA
1995;273:402–7.
[54] Price DD, Harkins SW. Psychophysical approaches to pain measurement and
assessment. In: Turk DC, Melzack R, editors. Handbook of pain
assessment. New York: Guilford Press; 2001. p. 112–21.
[55] Quiton RL, Greenspan JD. Across- and within-session variability of ratings of
painful contact heat stimuli. Pain 2008;137:245–56.
[56] Robinson ME. The effect of exercise on the perception of pain in athletes and
non-athletes: a signal detection theory analysis. Dissertation 1988.
[57] Ryan ED, Kovacic CR. Pain tolerance and athletic participation. Percept Mot
Skills 1966;22:383–90.
[58] Scott V, Gijsbers K. Pain perception in competitive swimmers. Br Med J
1981;283:91–3.
[59] Smith LD. The effects of competition and exercise on pain perception.
Dissertation 2004.
[60] Spector TD, Harris PA, Hart DJ, Cicuttini FM, Nandra D, Etherington J, Wolman
RL, Doyle DV. Risk of osteoarthritis associated with long-term weight-bearing
sports. Arthritis Rheum 1996;39:988–95.
[61] Stein C, Reinecke H, Sorgatz H. Opioid use in chronic noncancer pain:
guidelines revisited. Curr Opin Anaesthesiol 2010;23:598–601.
[62] Sternbach RA. Psychophysiology of pain. Int J Psychiatry Med 1975;6:63–73.
[63] Sternberg WF. Athletes: pain and pain inhibition. APS Bull 1999;9:18–9.
[64] Sternberg WF, Bailin D, Grant M, Gracely RH. Competition alters the perception
of noxious stimuli in male and female athletes. Pain 1998;76:231–8.
[65] Sterne JA, Sutton AJ, Ioannidis JP, Terrin N, Jones DR, Lau J, Carpenter J, Rucker
G, Harbord RM, Schmid CH, Tetzlaff J, Deeks JJ, Peters J, Macaskill P, Schwarzer
G, Duval S, Altman DG, Moher D, Higgins JP. Recommendations for examining
and interpreting funnel plot asymmetry in meta-analyses of randomised
controlled trials. BMJ 2011;343:d4002.
[66] Streff A, Kuehl LK, Michaux G, Anton F. Differential physiological effects during
tonic painful hand immersion tests using hot and ice water. Eur J Pain
2010;14:266–72.
[67] Suzuki K, Maekawa K, Minakuchi H, Yatani H, Clark GT, Matsuka Y, Kuboki T.
Responses of the hypothalamic-pituitary-adrenal axis and pain threshold
changes in the orofacial region upon cold pressor stimulation in normal
volunteers. Arch Oral Biol 2007;52:797–802.
[68] Syrjala KL, Donaldson GW, Davis MW, Kippes ME, Carr JE. Relaxation and
imagery and cognitive-behavioral training reduce pain during cancer
treatment: a controlled clinical trial. Pain 1995;63:189–98.
[69] Tajet-Foxell B, Rose FD. Pain and pain tolerance in professional ballet dancers.
Br J Sports Med 1995;29:31–4.
[70] van Middelkoop M, Rubinstein SM, Kuijpers T, Verhagen AP, Ostelo R, Koes BW,
van Tulder MW. A systematic review on the effectiveness of physical and
rehabilitation interventions for chronic non-specific low back pain. Eur Spine J
2011;20:19–39.
[71] Varrassi G, Bazzano C, Edwards WT. Effects of physical activity on maternal
plasma beta-endorphin levels and perception of labor pain. Am J Obstet
Gynecol 1989;160:707–12.
[72] Walker J. Pain parameters of athletes and non-athletes. Diss Abstr Int
1970;31:1607–8.
[73] Walker J. Pain and distraction in athletes and non-athletes. Percept Mot Skills
1971;33:1187–90.
[74] Wetterslev J, Thorlund K, Brok J, Gluud C. Trial sequential analysis may
establish when firm evidence is reached in cumulative meta-analysis. J Clin
Epidemiol 2008;61:64–75.
[75] Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, Leeflang
MM, Sterne JA, Bossuyt PM. QUADAS-2: a revised tool for the quality
assessment of diagnostic accuracy studies. Ann Intern Med 2011;155:529–36.
[76] Whitmarsh BG, Alderman RB. Role of psychological skills training in increasing
athletic pain tolerance. Sport Psychol 1993;7:388–99.
[77] Williamson E, Williams M, Gates S, Lamb SE. A systematic literature review of
psychological factors and the development of late whiplash syndrome. Pain
2008;135:20–30.
[78] World Health Organisation. Global strategy on diet, physical activity and health.
World Health Organization, Geneva, <http://www.who.int/dietphysicalactivity/
strategy/eb11344/strategy_english_web.pdf>; 2004 [accessed 31.12.11].
[79] Yamaguchi AY, Bucton BP. Differences in response and anxietyy between
athletes and non-athletes. J Athl Train 1997;32:45.
1262 J. Tesarz et al. / PAIN
Ò
153 (2012) 1253–1262
