Anaesthesia 2012 doi:10.1111/j.1365-2044.2012.07267.

Original Article
Modulation of pain sensation by stress-related testosterone and
cortisol
J. C. Choi,1 M. I. Chung2 and Y. D. Lee3

1 Associate Professor, 3 Resident, Department of Anesthesiology and Pain Medicine, Brain Research Group, Yonsei
University Wonju College of Medicine, South Korea
2 Director, GunSei Biotec Inc., Gwanak-gu, Seoul, South Korea

Summary
Stress increases cortisol and decreases testosterone. It is not known whether pain is affected by stress-related
testosterone. Therefore, we investigated whether stress can affect pain perception by decreasing testosterone and
increasing cortisol. Pain thresholds, pain and anxiety ratings and salivary testosterone and cortisol levels were measured
in 46 healthy men during resting and stressful conditions. Pain was induced by electrical stimulation. Stress was induced
by having participants perform a medical test. Stress significantly increased anxiety ratings and salivary cortisol levels,
but decreased salivary testosterone levels. Stress also increased pain ratings and decreased pain thresholds. During stress,
cortisol levels were negatively correlated with pain thresholds and testosterone levels were positively correlated with pain
thresholds. Results indicated that testosterone can decrease and cortisol can increase pain induced by electrical
stimulation, suggesting that acute clinical pain may be relieved by controlling stress and managing consequent stress-
related testosterone and cortisol.
. ..............................................................................................................................................................
Correspondence to: J. C. Choi
Email: jaechan@yonsei.ac.kr
Accepted: 20 June 2012

Stress activates the hypothalmic pituitary adrenal axis,
leading to the enhancement of cortisol secretion [1, 2].
Anticipatory stress has also been known to increase
cortisol levels [3]. Although cortisol regulates essential
physiological systems to restore homoeostasis and
protect against damage, this stress response can be
detrimental, resulting in both short- and long-term
damage that negatively impacts many organ systems
including the brain [1, 2]. It has been reported that
anxiety lowers the pain threshold, leading to an increase
in pain perception [4, 5]. Although many situations lead
to a stress response, we chose to focus on stressful
situations most commonly experienced by medical
students. Examinations have a very important role in
evaluating learning outcomes and competence [6]. In
particular, medical students are assumed to experience
high levels of anxiety during medical examinations.
Previous studies report that examination stress increases
anxiety, salivary cortisol level and blood pressure [6, 7].
Students self-report higher stress levels before examin-
ations, which are also associated with higher salivary
cortisol levels, and these students also tend to have lower
examination scores [7, 8]. Collectively, these studies
suggest that stress is linked to increased cortisol levels
that may possibly impact performance.
Anticipatory stress has also been shown to decrease
testosterone levels [3]. Low serum testosterone levels in
humans have been reported during psychological and

Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland 1

Anaesthesia 2012 Choi et al. | Modulation of pain sensation by stress-related testosterone and cortisol
physical stress, as well as during the acute stress of
surgery [9]. Men with low levels of testosterone tend to
be more anxious and irritable than men with normal
testosterone levels [10]. Animal studies have also shown
that testosterone increases anxiety, has an anti-analgesic
effect [11] and modulates opioid analgesia [12]. Our
previous studies with humans have indicated that
testosterone levels are positively correlated with activa-
tion in the middle frontal cortex during electrical and
thermal pain stimulation, leading to a decrease in pain
perception [13, 14].
Existing evidence suggests that increased cortisol
and decreased testosterone contribute to increased
stress. In a previous report that studied experimental
ischaemic arm pain in healthy volunteers, verbally
induced nocebo hyperalgesia was associated with
increased plasma adrenocorticotropic hormone (ACTH)
and cortisol concentrations [15]. Although it is known
that the testosterone:cortisol ratio is significantly lower
in patients with chronic migraine than in healthy
controls [16], it is not known whether pain ratings
and pain thresholds are affected by stress-related
testosterone in experimental electrical pain. Therefore,
we investigated whether stress affects pain perception by
modulating the activities of the hypothalamic pituitary
adrenal and hypothalamic pituitary axes during stress.
Methods
Forty-six healthy male medical students were recruited
for this study. Participants were excluded if they had a
history of chronic pain, chronic opioid consumption,
chronic psychiatric disorders or drug abuse, or were
taking regular analgesia in the week before the study.
After ethical approval from the medical ethics commit-
tee of Yonsei University Wonju College of Medicine, all
participants provided written, informed consent.
The Objective Structured Clinical Examination
(OSCE) was completed in the Department of Anesthe-
siology and Pain Medicine of Yonsei University Wonju
College of Medicine. The OSCE includes three basic
medical anaesthetic technique tests performed with
dummies (Test 1: intubation practice; Test 2: venepunc-
ture practice; Test 3: local anaesthesia practice). The
OSCE was performed on the last day of the polyclinic in
the Department of Anesthesiology and Pain Medicine.
The OSCE was worth one credit and was evaluated
2 Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland
Table 1 Experimental sequence. This sequence was
identical in the resting and stress conditions.
Step 1 Step 2 Step 3
Sampling saliva Pain threshold Pain rating
Asking anxiety rating measurement measurement
Blood pressure and at 20 mA
heart rate
measurement
based on a letter grade system (A, B, C, etc.). The same
pain experiments (Table 1) were conducted twice, once
during rest 2 days before the actual OSCE (the resting
condition) and once during stress (the actual OSCE).
The pain experiment with the participants under stress
was completed between the venepuncture and the local
anaesthesia tests.
Participants sat in a chair with their left hand
comfortably resting on their thigh. They were asked to
rate their anxiety before pain stimulation, both in the
resting and OSCE conditions. Before applying pain
stimulation, blood pressure and heart rate were mea-
sured in the arm using an automatic sphygmomano-
meter both in the resting and stressful conditions. Pain
was induced experimentally using an electrical stimula-
tor (Multistim Vario; PAJUNK, Geisingen, Germany).
One silver chloride electrode (Bio Protech Inc., Wonju,
Korea) was attached 2 cm proximal to the left palmar
distal wrist crease. The other electrode was attached
2 cm proximal to the left dorsal wrist crease. The pain
threshold was defined as the lowest electrical current
required for the participants to report a sensation of
pain. Electrical current was incrementally increased in
steps of 0.1 mA (2 Hz) until the participant first
reported pain. After the pain threshold was reached,
the current was incrementally increased to 20 mA
(2 Hz) in steps of 1 mA and maintained for 15 s. If
participants felt an intolerable pain sensation due to the
increasing electrical current, the current was not
increased further. Participants were asked to rate their
pain sensation after electrical stimulation at 20 mA for
15 s. Although a current higher than 20 mA might be
required to induce pain sensation similar to that in a
clinical situation, a maximum of 20 mA was chosen to
ensure participants’ safety. Ratings were assessed using a
numerical rating scale (0 = no pain and anxiety;
100 = maximum imaginable pain and anxiety).
Choi et al. | Modulation of pain sensation by stress-related testosterone and cortisol Anaesthesia 2012

To evaluate the activity of the hypothalamic-
pituitary-adrenal and hypothalamic-pituitary-gonadal
axes, saliva samples were used as this is non-invasive
and saliva has been shown to be valid for the measurement
of bio-active steroid hormones [4, 17, 18]. Before applying
pain stimulation, saliva samples were collected in the
afternoon (14:30 – 16:30) on the rest and stress days. On
the stress day, the saliva sample was collected between the
venepuncture and local anaesthesia tests. To minimise the
effects of diurnal variations, saliva samples were collected
within the same time period for both the resting and
stressful conditions. Participants were instructed to avoid
eating, drinking, chewing gum or brushing their teeth one
hour before sampling. After collection, samples were kept
in the deep freezer overnight.
For salivary cortisol and testosterone analysis, the
samples were thawed and spun at 3000 rpm for 5 min to
obtain a clear, watery supernatant with low viscosity. For
cortisol determination, 100 ll saliva was removed for
duplicate analysis of cortisol levels using a commercial
kit (DRG Instruments Gmbh, Germany). The lower
detection limit of the assay was 0.1 ng.ml)1 with intra-
assay and inter-assay coefficients of variance < 5%
across the expected range of cortisol levels. For testos-
terone determination, 100 ll saliva was used for dupli-
cate analysis of testosterone levels using a commercial
Table 2 Participants’ characteristics (n = 43). Values
are mean (SD).
Age; years 26.2
Weight; kg 71.4
Height; cm 176.6
Body mass index; kg.m)2 22.9
kit (IBL-America, Minneapolis, MN, USA). The lower
detection limit of the assay was 10 pg.ml)1. Intra-assay
and inter-assay coefficients of variation for salivary
testosterone ranged from 2.2 to 5.2 and 3.2 to 9.6%
respectively.
Statistical analysis was performed with PASW
Statistics 18 (Predictive Analytics Software; SPSS Inc.,
Chicago, IL, USA). Group data (resting condition vs
stressful condition) were compared using a paired t-test.
Comparisons between the control and stressful condi-
tions in terms of pain and anxiety scores were analysed
using the Wilcoxon signed-rank test. Correlation coef-
ficients were calculated using a partial correlation
analysis (two-sided). All results were considered signif-
icant at a threshold of p < 0.05.
Results
All 46 participants completed the entire resting portion
of the experiment. However, three participants reported
an intolerable pain sensation before reaching 20 mA
(2 Hz) of electrical current during stress and were
withdrawn from the study. Therefore, data from 43
participants were used in the analyses (Table 2). Anxiety
ratings were significantly higher during the stressful
condition compared with the resting condition
(Table 3), confirming that the OSCE was a valid stress
situation.
Salivary testosterone levels were significantly lower,
and salivary cortisol levels significantly higher, during
the stressful condition than during the resting condition
(3.2) (Table 3). Systolic blood pressure and heart rate were
(8.9)
significantly higher in the stressful condition than in the
(4.9)
(2.8) resting condition, suggesting higher sympathetic ner-
vous system activity during stress (Table 3). Pain ratings

Table 3 Comparison of hormone and self-reported measures during the resting and stress conditions. Values are mean
(SD) or median (IQR [range]) for pain and anxiety ratings.

Resting condition Stress condition p value

)1
Testosterone; pg.ml 78.1 (36.9) 66.2 (25) 0.005
Cortisol; ng.ml)1 2.6 (2.4) 4.0 (3.8) 0.027
Pain ratings 25 (15–30 [5–75]) 40 (30–60 [7–85]) < 0.001
Anxiety ratings 30 (10–30 [5–50]) 50 (40–60 [10–90]) < 0.001
Pain threshold; mA 10.3 (5.0) 8.1 (4.3) < 0.001
Systolic blood pressure; mmHg 119.2 (11.5) 122.9 (12.8) 0.012
Diastolic blood pressure; mmHg 75.3 (10.2) 77.1 (9.5) 0.085
Heart rate; beats.min)1 71.3 (9.7) 76.8 (12.5) 0.003

Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland 3

Anaesthesia 2012 Choi et al. | Modulation of pain sensation by stress-related testosterone and cortisol
were higher, while pain thresholds were lower, in the
stressful condition than in the resting condition
(Table 3).
After controlling for cortisol, the correlation coef-
ficient corresponding to the relationship between pain
thresholds and testosterone levels was 0.32 (p = 0.04).
After controlling for testosterone, the correlation coef-
ficient corresponding to the relationship between pain
thresholds and cortisol levels was )0.34 (p = 0.027).
Discussion
The results of the present study indicate that medical
students experience significant stress during a medical
exam, leading to an increase in pain sensation as
indicated by decreased pain thresholds and increased
pain ratings. This suggests that a stimulus perceived as
painless in the resting condition was perceived as painful
in the stressful condition. Decreases in the level of
testosterone, increases in cortisol and activation of the
sympathetic nervous system are possible factors con-
tributing to the increase in pain sensation during the
stressful condition.
In our earlier study, we found that serum ACTH
and cortisol levels during stress (measured immediately
before surgery) were significantly higher than during
rest [19]. In the present study, salivary cortisol levels
were also significantly higher in the stressful condition
than in the resting condition. Furthermore, in the
stressful condition, cortisol levels were negatively corre-
lated with pain thresholds. Therefore, participants with
higher cortisol levels perceived pain at a lower electrical
current than those with lower cortisol levels. Pain ratings
were higher in the stressful condition than in the resting
condition. These findings suggest that people may
perceive more pain when stressed due in part to
activation of the hypothalamic-pituitary-adrenal axis.
It was previously reported that anticipatory stress
increases the mean cortisol level, while anticipatory
stress decreases the mean testosterone level [3]. In the
present study, testosterone levels were significantly lower
in the stressful condition than in the rest condition. In
particular, testosterone levels were positively correlated
with pain thresholds during stress. As stress lowers the
testosterone level [3, 9], low testosterone levels due to
stress may be associated with decreased pain perception.
In the present study, pain ratings were higher in the
4 Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland
stressful condition, during which testosterone levels
were low, than in the resting condition, during which
the testosterone levels were high. In our earlier study of
young adult women [13, 14], we found a significantly
strong negative correlation between testosterone levels
and pain ratings. These results suggest that increased
testosterone levels in men and women may be associated
with decreased pain perception.
The testosterone levels of the women in our
previous study [13, 14] were also strongly positively
correlated with activation of the right middle frontal
cortex. Activation of the bilateral middle frontal cortices
during pain stimulation was significantly negatively
correlated with pain ratings during the follicular and
luteal phases of the menstrual cycle [13, 14], and there
was a significant negative correlation between thalamic
activation (medial dorsal nucleus) and serum testoster-
one levels in the pain stimulation period [13]. The
medial dorsal nucleus of the thalamus receives spino-
thalamic input and is thought to play a role in the
affective and motivational aspects of pain [20]. Collec-
tively, these results indicate that increased testosterone
may be involved in a decrease in pain perception by
activating the middle frontal cortex and suppressing the
thalamus.
The mechanisms underlying increased pain ratings
in the stress condition may involve activation of the
hypothalamic-pituitary-adrenal axis and deactivation of
the hypothalamic-pituitary-gonadal axis. In the present
study, salivary cortisol levels were higher in the stressful
condition than in the resting condition. Although we are
unable to infer a causal relationship between stress and
cortisol level, previous studies have proved that stress is
associated with higher cortisol levels [3, 6–8], and in this
study, pain ratings were higher in the stressful condition,
during which cortisol levels were high, than in the
resting condition, during which cortisol levels were low.
These findings suggest that stress associated with the
examination may activate the hypothalamic-pituitary-
adrenal axis and release salivary cortisol, thereby
resulting in increased pain ratings under stressful
conditions. Another mechanism potentially involved in
the increase in pain sensation during stress may be
deactivation of the hypothalamic-pituitary-gonadal axis,
which was first reported in our earlier studies [13, 14]. In
this study, the stress associated with the exam may have
Choi et al. | Modulation of pain sensation by stress-related testosterone and cortisol
decreased the salivary testosterone levels during stress,
which suggests that stress may deactivate the hypotha-
lamic-pituitary-gonadal axis, resulting in decreased
salivary testosterone levels and increased pain ratings
in the stressful condition. Although we were not able to
infer a causal relationship between hormone levels and
pain, these mechanisms could contribute to a decrease in
pain thresholds and an increase in pain ratings during
stress.
It is known that stress-induced analgesia is a pain
suppression response that occurs during or following
exposure to a stressful or frightening stimulus [21]. It is
generally accepted that the induction of anxiety may
increase sensitivity to pain, whereas the induction of fear
decreases sensitivity to pain [21–25]. In the present
study, participants may have felt anxiety but not fear
sensation. If participants were exposed to a very stressful
situation, such as examination for medical licensure,
they may have experienced fear sensation and a pain
suppression response to the pain stimulus.
The limitation of this study is that we did not
conduct the present experiment using an ideal experi-
mental design, as the experiment was completed based
on the polyclinic schedule in the Department of
Anesthesiology and Pain Medicine. To clarify the
relation between behavioural data and hormone levels,
further research is necessary with a better experimental
model or research in clinical practice.
In conclusion, our results indicate that stress
increases the activities of the hypothalamic-pituitary-
adrenal axis and decreases hypothalamic-pituitary-
gonadal axis activity, thereby contributing to an increase
in pain perception. These results also suggest that acute
clinical pain, similar to electrical pain used in the present
study, may be relieved by managing stress and control-
ling consequent stress-related testosterone and cortisol
levels.
Acknowledgements
The authors thank the participants of Yonsei University,
Wonju College of Medicine for participating voluntarily
in this experiment. The authors also thank Tae Soo Kim,
B.A. (Department of Medical Information Development,
Yonsei University), Gi Bong Um, B.A. (Planning Team,
Yonsei University), Gill Ho Lee, B.A. (Instructional
Department, Yonsei University), Ju Seog Kim, B.A.
Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland
Anaesthesia 2012
(Librarian Section, Yonsei University), Sang Min Lee,
B.A. (Academic Affairs, Yonsei University) and Myung
Ha Kim, B.A. (Librarian Section, Yonsei University) for
their excellent technical assistance. This study was
supported by a research grant from Yonsei University
Wonju College of Medicine (YUWCM-2011-69), and by
the Basic Science Research Program through the
National Research Foundation of Korea (NRF) funded
by the Ministry of Education, Science and Technology
(2012R1A1A2004868).
Competing interests
No external funding or competing interests declared.
References
1. Borsook D, George E, Kussman B, Becerra L. Anesthesia and
perioperative stress: consequences on neural networks and
postoperative behaviors. Progress in Neurobiology 2010; 92:
601–12.
2. Rodrigues SM, LeDoux JE, Sapolsky RM. The influence of stress
hormones on fear circuitry. Annual Review of Neuroscience
2009; 32: 289–313.
3. Schulz P, Walker JP, Peyrin L, Soulier V, Curtin F, Steimer T. Lower
sex hormones in men during anticipatory stress. NeuroReport
1996; 7: 3101–4.
4. Kanegane K, Penha SS, Munhoz CD, Rocha RG. Dental anxiety
and salivary cortisol levels before urgent dental care. Journal of
Oral Science 2009; 51: 515–20.
5. Litt MD. A model of pain and anxiety associated with acute
stressors: distress in dental procedures. Behaviour Research and
Therapy 1996; 34: 459–76.
6. Brand HS, Schoonheim-Klein M. Is the OSCE more stressful?
Examination anxiety and its consequences in different assess-
ment methods in dental education. European Journal of Dental
Education 2009; 13: 147–53.
7. Ng V, Koh D, Mok BY, Chia SE, Lim LP. Salivary biomarkers
associated with academic assessment stress among dental
undergraduates. Journal of Dental Education 2003; 67: 1091–4.
8. Ng V, Koh D, Chia SE. Examination stress, salivary cortisol, and
academic performance. Psychological Reports 2003; 93: 1133–
4.
9. Van den Berghe G. The neuroendocrine response to stress is a
dynamic process. Best Practice & Research Clinical Endocrinol-
ogy and Metabolism 2001; 15: 405–19.
10. Sternbach H. Age-associated testosterone decline in men:
clinical issues for psychiatry. American Journal of Psychiatry
1998; 155: 1310–8.
11. Edinger KL, Frye CA. Testosterone’s anti-anxiety and analgesic
effects may be due in part to actions of its 5alpha-reduced
metabolites in the hippocampus. Psychoneuroendocrinology
2005; 30: 418–30.
12. Craft RM, Mogil JS, Aloisi AM. Sex differences in pain and
analgesia: the role of gonadal hormones. European Journal of
Pain 2004; 8: 397–411.
13. Choi JC, Park SK, Kim YH, et al. Different brain activation patterns
to pain and pain-related unpleasantness during the menstrual
cycle. Anesthesiology 2006; 105: 120–7.
5
Anaesthesia 2012 Choi et al. | Modulation of pain sensation by stress-related testosterone and cortisol
14. Choi JC, Yi DJ, Han BS, Lee PH, Kim JH, Kim BH. Placebo effects on
analgesia related to testosterone and premotor activation.
NeuroReport 2011; 22: 419–23.
15. Benedetti F, Amanzio M, Vighetti S, Asteggiano G. The
biochemical and neuroendocrine bases of the hyperalgesic
nocebo effect. Journal of Neuroscience 2006; 26: 12014–
22.
16. Patacchioli FR, Monnazzi P, Simeoni S, et al. Salivary cortisol,
dehydroepiandrosterone-sulphate (DHEA-S) and testosterone in
women with chronic migraine. Journal of Headache and Pain
2006; 7: 90–4.
17. King SL, Hegadoren KM. Stress hormones: how do they measure
up? Biological Research for Nursing 2002; 4: 92–103.
18. Liening SH, Stanton SJ, Saini EK, Schultheiss OC. Salivary
testosterone, cortisol, and progesterone: two-week stability,
interhormone correlations, and effects of time of day, menstrual
cycle, and oral contraceptive use on steroid hormone levels.
Physiology and Behavior 2010; 99: 8–16.
6 Anaesthesia ª 2012 The Association of Anaesthetists of Great Britain and Ireland
19. Choi JC, Park SK, Kim SY, et al. Oral alprazolam attenuates
preoperative stress responses to regional anesthesia. Korean
Journal of Anesthesiology 2007; 52: 132–7.
20. Martin JH. Neuroanatomy: Text and Atlas, 3rd edn. New York:
McGraw-Hill Companies, 2003: 125–6.
21. Butler RK, Finn DP. Stress-induced analgesia. Progress in
Neurobiology 2009; 88: 184–202.
22. Meagher MW, Arnau RC, Rhudy JL. Pain and emotion: effects of
affective picture modulation. Psychosomatic Medicine 2001; 63:
79–90.
23. Rhudy JL, Meagher MW. Fear and anxiety: divergent effects on
human pain thresholds. Pain 2000; 84: 65–75.
24. Rhudy JL, Meagher MW. Noise stress and human pain thresh-
olds: divergent effects in men and women. Journal of Pain
2001; 2: 57–64.
25. Ploghaus A, Narain C, Beckmann CF, et al. Exacerbation of pain
by anxiety is associated with activity in a hippocampal network.
Journal of Neuroscience 2001; 21: 9896–903.