Temperature and the Cold Pressor Test

Laura A. Mitchell, Raymond A. R. MacDonald, and Eric E. Brodie

Abstract: As a method of experimental pain induction, the cold pressor test is thought to mimic the
effects of chronic conditions effectively. A survey of previous studies using the cold pressor, however,
revealed a lack of standardization and control of water temperature, questioning comparability and
reliability. This study reports the influence of temperature on pain tolerance and intensity by using a
commercially available circulating water bath. Twenty-six participants (12 men, 14 women) under-
went 4 cold pressor trials with temperature order counterbalanced across 1°C, 3°C, 5°C, and 7°C,
temperatures representative of the range used in previous literature. After each cold immersion
participants rated pain intensity on a visual analogue scale and the McGill Pain Questionnaire.
Tolerance times were recorded for each trial. Significant main effects of temperature were found for
tolerance time, with higher temperatures resulting in longer times, and pain intensity, with lower
temperatures resulting in higher intensities. Gender differences were found, with men tolerating the
stimulus for significantly longer than women. It was concluded that small differences in water
temperature have a significant effect on pain intensity and tolerance time. The use of cold pressor
equipment that ensures a precise constant temperature of circulating water is necessary to ensure
comparable and reliable results.
Perspective: The cold pressor method of experimental pain induction has been widely used in the
evaluation of psychological and physiological pain treatments. This article highlights the need for clear
methodologic guidelines for the technique and demonstrates that very minor changes in experimental
protocol can produce significant differences.
© 2004 by the American Pain Society
Key words: Cold pressor, temperature, gender.

T
he ability to administer noxious stimuli under ex-
perimental conditions is widely recognized as in-
valuable in investigations of analgesic manipula-
tions.28,34 Fast and efficient evaluations of pain
treatments are made possible in healthy participants,
making these techniques a useful and less expensive pre-
requisite to clinical studies.14 Since the early work of
Smith et al31 with experimentally induced ischemic pain,
the precise quantitative control and the ability to manip-
ulate the stimuli have come to be acknowledged as a
considerable research advantage.36
Cold pressor pain, induced by the submergence of the
hand in cold water, is suggested to be a method that
mimics the effects of chronic conditions effectively be-
cause of its unpleasantness28, and it has excellent reli-
ability and validity.6 The procedure has been used in
studies investigating a wide range of pain management
Received October 17, 2003; Revised March 29, 2004; Accepted March 29,
2004.
From the Department of Psychology, Glasgow Caledonian University,
Glasgow, United Kingdom.
Supported by postgraduate studentship (L.A.M.) from the Scottish Net-
work for Chronic Pain Research.
Address reprint requests to Laura Mitchell, Lecturer, Department of Psy-
chology, Glasgow Caledonian University, Cowcaddens Road, Glasgow G4
OBA, UK. E-mail: l.b.mitchell@gcal.ac.uk
1526-5900/$30.00
© 2004 by the American Pain Society
doi:10.1016/j.jpain.2004.03.004
techniques such as acupuncture,1 hypnosis,13 neutral dis-
traction,18 and cognitive preparation.35
In surveying studies from the past 2 decades that have
used the cold pressor, are included in the medical or
psychological literature databases, and give a compre-
hensive description of their methodology, however, a
lack of standardization is revealed, with substantial vari-
ation in both equipment and methodology used (Table
1).
Water temperatures used in previous studies can be
seen to range between 0°C and 7°C, and circulation of
the water, which is essential to avoid heat buildup
around the hand, was used only in approximately half of
the studies. Three studies3,25,27 can also be seen to use
water temperatures that varied by 2°C within their indi-
vidual study. In accordance with these temperature dif-
ferences, the number of immersions and the maximum
possible tolerance time set by experimenters, potentially
important safety aspects, also vary considerably.
The average tolerance time achieved by participants
varies quite substantially between studies. As might be
expected, studies using water circulation have a lower
average tolerance range (28.65-119.75 sec) than those
using uncirculated water (37.11-190.3 sec). Small differ-
ences in temperature when all other variables are con-
stant appear to increase tolerance both in studies with
control conditions (40.80 sec at 0°C-1°C up to 107.03 sec

The Journal of Pain, Vol 5, No 4 (May), 2004: pp 233-238 233

234 Temperature and the Cold Pressor Test

Table 1. Previous Cold Pressor Methodology

NUMBER OF
AUTHOR YEAR TEMPERATURE (C) CIRCULATED? ARM/HAND IMMERSIONS

Rosenbaum30 1980 1°-2° No Hand 1

Worthington and Shumate37 1981 0°-1° Yes Arm 2
McCaul and Haugtvedt24 1982 7° No Hand 1
Ashton et al1 1984 0° Not stated Hand 6
Farthing et al8 1984 1°-2° Yes Hand 2
Johnson et al20 1989 0° Not stated Hand 6
Hodes et al18 1990 0°-1° No Arm 2
Levine and deSimone23 1991 0°-1° No Hand 1
Rainville et al28 1992 5° Yes Hand 3
McCaul et al25 1992 1°-3° No Hand 1
Hekmat and Hertel15 1993 0°-1° Yes Hand 2
Weisenberg et al35 1995 1° Yes Arm 6
Johnson and Petrie19 1997 4° Not stated Hand 1
Compton3 1998 0°-2° No Hand 1
Freeman et al13 2000 2° Yes Hand 3
Keogh et al21 2000 3° Yes Hand 1
Stevens et al32 2000 3° Yes Arm 1
Enggaard et al7 2001 4° Yes Hand 1
Myers et al27 2001 1°-3° Yes Hand 1
deWied and Verbaten4 2001 2° Yes Hand 1
Feldner and Hekmat10 2001 0°-1° Not stated Hand 1
Keogh and Hertenfeldt22 2002 1°-2° Yes Hand 2

at 1°C-3°C) and in studies comparing intervention groups
(28.65-60.48 sec at 1°C-2°C up to 65.52-95.76 sec at 3°C).
Although some guidelines do exist from the early work
of Hilgard16 on the systematic relationship between wa-
ter temperature and pain report at 0°C, 5°C, 10°C, and
15°C, no standardized methodology was advocated. A
further study34 then aimed to establish age, sex, and
ethnic group norms for the cold pressor. Although this
study recognized pain severity to be inversely related to
temperature, the experimenters failed to control tem-
perature precisely.
The methodologic inconsistencies within cold pressor
research and their potential implications were high-
lighted recently in a review article by Eccleston.5 This
author indicated that possible differences in pain expe-
rience due to water temperature remain largely unac-
knowledged within the literature, despite being a po-
tential source of variance within the findings of similar
studies. One further study has included the effects of
cold pressor temperature on tolerance and perception of
pain intensity,17 only in female participants however.
Furthermore, the temperatures included, 3°C, 8°C, 13°C,
18°C, and 23°C, varied substantially and did not repre-
sent those used by previous studies. Last, the indepen-
dent groups design of the study might have meant that
variation arose purely as a result of individual differences
in tolerance of pain.
Methodology in cold pressor studies, therefore, can be
seen to vary to such an extent that the reliability and
validity of the test are compromised. Variation in toler-
ance might in fact relate to differences in methodology
rather than the experimental conditions under investiga-
tion.5 Clarifying the influence of water temperature on
pain tolerance and discomfort is therefore of importance
to the design of more robust, easily comparable studies
in future.
The current study also considers gender differences in
responses to experimentally induced pain. Clinical pain
research has generally indicated that women report
more recurrent, severe, and long-lasting pain.33 Recent
reviews of laboratory-induced pain studies are also
mainly in support of increased sensitivity in women, al-
though results have not been wholly consistent.2,11
Method of pain induction itself is proposed to influence
findings, with most consistent results in methods produc-
ing deep, tonic sensations that are closest to the natural
experience of pain.11
The current study aims to investigate the effects of the
observed incongruity within cold pressor methodology,
examining the effects of water temperature and gender
on tolerance and intensity rating. A repeated measures
design was considered essential to control for the sub-
stantial individual differences that exist in pain re-
sponse.12 By using a commercially available circulating
water bath capable of maintaining a precise and con-
stant temperature, this study attempts to show the im-
portance of this methodologic factor in the design and
interpretation of future studies.
Materials and Methods
Design
The study took the form of a repeated measures design
with 26 participants each undergoing 4 cold pressor tri-
ORIGINAL REPORT/Mitchell et al 235

als. The 4 temperature conditions (1°C, 3°C, 5°C, and 7°C)

were presented in a counterbalanced order to prevent
possible order effects. Dependent variables were total
tolerance time (seconds) and pain intensity rating on vi-
sual analogue scale (VAS) and the pain rating index (PRI)
of the McGill Pain Questionnaire.26

Participants
Twenty-six undergraduate participants (12 men, 14
women) with a mean age of 26 years were recruited
through advertisements on university notice boards and
paid for their participation.

Apparatus Figure 1. Mean tolerance times of men and women in each of

A circulating and cooling water bath, Jeiotech model
VTRC 620 (Seoul, Republic of Korea), was used as the cold
pressor stimulus. This model of water bath has a 6-L water
capacity and maintains temperature to within the 0.1°C
level. Hand temperatures were monitored by using a dig-
ital thermometer held tightly in the palm of the hand.
Measurement indexes were (1) tolerance time (sec-
onds), measured with a stopwatch; (2) 100-mm VAS, on
which participants were asked to make a mark between
“no discomfort” and “worst possible discomfort;” and
PRI of the McGill Pain Questionnaire.
Procedure
Ethical approval for the study was granted by Glasgow
Caledonian University Research Ethics Committee. Be-
fore taking part in the study, all participants signed a
consent form and were screened for contraindicative
medical conditions including heart or circulation prob-
lems (especially Reynaud’s syndrome), blood pressure
problems, diabetes, epilepsy, and recent serious injury.
The study took place in a university laboratory, with
participants taking part individually. Hand temperatures
were measured when the participant entered the labo-
ratory, followed by immersion of the nondominant hand
in a tub of warm water at 32°C for 1 minute. Hand tem-
perature was again measured, and participants were asked
to immerse the hand up to the wrist in the cold water and
to keep it there until it felt too uncomfortable to continue.
Participants were made aware that they could move the
hand or keep it still as they wished, and that maximum
possible tolerance time was limited by the experimenters to
5 minutes, at which point they would be asked to remove
the hand from the water. Tolerance time was recorded,
and after the cold immersion participants rated the inten-
sity of the most intense pain felt during the immersion
period on a VAS and the PRI. There was a break of 5 min-
utes between trials during which participants again im-
mersed the hand in the 32°C water for 1 minute, and hand
temperature was monitored. Participants were not given
any cues that the temperature of the water was altered.
Each participant underwent 4 such trials with temperature
order counterbalanced across 1°C, 3°C, 5°C, and 7°C.
the 4 temperature conditions.
Results
Analysis of the tolerance times, visual analogue ratings
of pain intensity, and PRI of the McGill Pain Question-
naire were carried out for each of the 4 temperatures.
Tolerance
Figure 1 shows the mean tolerance times of men and
women for each temperature. Standard deviations are
shown in opposite directions for men and women.
A mixed design analysis of variance found a significant
main effect of temperature on tolerance time (F3,22 ⫽
4.5; P ⬍ .05) and a significant main effect of gender (F1,24
⫽ 7.79; P ⬍ .01). No significant interaction was found
between gender and temperature. Tolerance times in-
creased as temperature increased, and men had signifi-
cantly longer tolerance times than women at all temper-
atures. Standard deviations were found to be greater in
the 2 higher temperature conditions.
Post hoc Bonferroni–adjusted paired comparisons
were then carried out on overall mean tolerance times of
the 4 temperatures. Water temperatures of 7°C and 5°C
were found to result in a significantly longer tolerance
time than a temperature of 1°C, both at the level of P ⬍
.05.
VAS
Figure 2 shows the mean VAS ratings of pain intensity
and standard deviations for men and women for each of
the 4 temperatures.
A mixed design analysis of variance found a significant
main effect of temperature on visual analogue rating
(F3,22 ⫽ 8.73; P ⬍ .001). No significant effect for gender
on VAS rating was found, and there was no interaction
effect between gender and temperature. Standard devi-
ations were again greater in the higher temperature
conditions.
Post hoc Bonferroni–adjusted paired comparisons
found 1°C to result in a significantly greater VAS pain
rating than 3°C at a level of P ⬍ .001, 5°C at a level of P ⬍
.01, and 7°C, P ⬍ .01.
236
Figure 2. Mean VAS ratings of men and women in each of the
4 temperature conditions.
PRI
No significant main effect of temperature or gender
on total PRI scores was found. There was no interaction
between the 2 factors. No significant main effects or
interactions were found when the PRI was separated
into sensory and affective subclasses.
Hand Temperatures
Figure 3 shows the number of participants with a hand
temperature of 32°C or higher after the first baseline im-
mersion in the 32°C water and after the 32°C immersions
that followed each of the first 3 cold-water immersions.
A ␹2 test found significantly less participants to reach
the 32°C hand temperature after the second cold immer-
sion, ␹ (1) ⫽ 5.54, P ⬍ .05, and after the third immersion,
␹ (1) ⫽ 7.54, P ⬍ .01.
Discussion
Small variations in water temperature when using the
cold pressor test result in significant differences in both
pain intensity levels and tolerance times. Tolerance times
Figure 3. Number of participants with hand temperature of
32°C and greater after each immersion.
Temperature and the Cold Pressor Test
were found to be significantly longer with a difference
of 4°C. The ratings given on the VAS were significantly
higher with a difference of 2°C.
These results make clear the difficulties that arise when
comparing the findings of previous studies. Our survey
revealed not only a substantial range of temperatures
between studies, making future methodologic design
unclear, but also within the individual experiments. In
using apparatus unable to precisely maintain tempera-
ture, participants might be subjected to inconsistent lev-
els of the stimulus, resulting in significantly different
pain experiences. This is complicated further by failure to
circulate the water in a number of studies, an important
methodologic feature in achieving consistency throughout
each cold immersion. Without this, heat is able to build
up around the hand, making longer tolerance possible.
A further methodologic aspect investigated in the cur-
rent study was the immersion of the hand in 32°C water,
a technique reported in many previous studies to return
hand temperature to normal after cold immersion. This
was not found to be the case when hand temperatures
were measured, with the number of participants reach-
ing this level dropping after each trial. When reaching
the second and third trials, only 7 of the 26 participants
had the desired hand temperature. This finding high-
lights the importance of counterbalancing order of con-
ditions between participants and also suggests that 1
minute is not enough time to regain normal hand tem-
perature. Therefore, the hand should ideally be left in
the 32°C water until the original hand temperature is
regained.
Gender differences in line with previous research on
experimentally induced pain tolerance11,29 were ob-
served in the current study, with men tolerating the
painful stimulus significantly longer. Previous investiga-
tion of ratings of pain intensity also suggested women
give higher ratings,9,21 specifically when reporting the
sensory aspects of pain.22 Because a female researcher
carried out the current study, a decrease in the pain re-
port given by men might particularly have been ex-
pected, as suggested by Levine and deSimone.23 Al-
though female participants did appear to consistently
rate pain intensity higher on the VAS, no significant dif-
ferences were found. This finding is therefore inconsis-
tent with previous experimental and clinical evidence of
greater negative responses to pain in women.2,33
Building on the previous findings of Hirsch and
Liebert,17 this study concludes that variations in temper-
ature as small as 2°C when using the cold pressor might
result in significantly different pain experience in both
men and women. In the design of future cold pressor
studies, the use of cold pressor equipment that ensures a
precise and constant temperature is therefore necessary
to ensure replicable and reliable results.
Acknowledgment
Many thanks to Dr Christina Knussen for comments on
earlier drafts of this article.
ORIGINAL REPORT/Mitchell et al
References
1. Ashton H, Ebeneezer I, Golding JF, Thomson JW: Effects of
acupuncture and transcutaneous electrical nerve stimula-
tion on cold-induced pain in normal subjects. J Psychosom
Res 28:301-308, 1984
2. Berkley KJ: Sex differences in pain. Behav Brain Sci 20:371-
380, 1997
3. Compton C: Pain tolerance in opioid addicts on and off
naltrexone pharmacotherapy: A pilot study. J Pain Symptom
Manage 16:21-28, 1998
4. de Wied M, Verbaten M: Affective pictures processing,
attention and pain tolerance. Pain 90:163-172, 2001
5. Eccleston C: The attentional control of pain: Methodolog-
ical and theoretical concerns. Pain 63:3-10, 1995
6. Edens JL, Gil KM: Experimental induction of pain: Utility in
the study of clinical pain. Behav Ther 26:197-216, 1995
7. Enggaard TP, Poulsen L, Arendt-Nielson L, Hansen SH,
Bjornsdottir I, Gram LF, Sindrup SH: The analgesic effect of
codeine as compared to imipramine in different human ex-
perimental pain models. Pain 92:277-282, 2001
8. Farthing GW, Venturino M, Brown SW: Suggestion and
distraction in the control of pain: Test of two hypotheses. J
Abnorm Psychol 93:266-276, 1984
9. Feine JS, Bushnell MC, Miron D, Duncan GH: Sex differ-
ences in the perception of noxious heat stimuli. Pain 44:255-
262, 1991
10. Feldner MT, Hekmat H: Perceived control over anxiety-
related events as a predictor of pain behaviours in a cold
pressor task. J Behav Ther Exp Psychiatry 32:191-202, 2001
11. Fillingim RB, Maixner W: Gender differences in response
to noxious stimuli. Pain Forum 4:209-221, 1995
12. Fillingim RB, Ness TJ: Sex-related hormonal influences on
pain and analgesic responses. Neurosci Biobehav Rev 24:
485-501, 2000
13. Freeman R, Barabasz A, Barabasz M, Warner D: Hypnosis
and distraction differ in their effects on cold pressor pain.
Am J Clin Hypn 43:137-148, 2000
14. Gracely RH: Experimental pain models. In: Portenoy
MMR, Laska E (eds): Advances in Pain Research and Therapy.
vol 18. New York, NY, Raven Press Ltd, 1991, pp 33-53
15. Hekmat H, Hertel JB: Pain attenuating effects of pre-
ferred versus non-preferred music interventions. Psychol
Music 21:163-173, 1993
16. Hilgard ER: Pain as a puzzle for psychology and physiol-
ogy. Am Psychol 24:103-113, 1969
17. Hirsch MS, Liebert RM: The physical and psychological
experience of pain: The effects of labelling and cold pressor
temperature on three pain measures in college women. Pain
77:41-48, 1998
18. Hodes RL, Howland EW, Lightfoot N, Cleeland CS: The
effect of distraction on responses to cold pressor pain. Pain
41:109-114, 1990
19. Johnson MH, Petrie SM: The effects of distraction on
237
exercise and cold pressor tolerance for chronic low back pain
sufferers. Pain 69:43-48, 1997
20. Johnson MI, Ashton CH, Bousfield DR, Thompson JW:
Analgesic effects of different frequencies of TENS on cold-
induced pain in normal subjects. Pain 39:231-236, 1989
21. Keogh E, Hatton K, Ellery D: Avoidance versus focused
attention and the perception of pain: Differential effects for
men and women. Pain 85:225-230, 2000
22. Keogh E, Herdenfeldt M: Gender, coping and the per-
ception of pain. Pain 97:195-201, 2002
23. Levine FM, deSimone LL: The effects of experimenter
gender on pain report in male and female subjects. Pain
44:69-72, 1991
24. McCaul KD, Haugtvedt C: Attention, distraction and cold
pressor pain. J Pers Soc Psychol 43:154-162, 1982
25. McCaul KD, Monson N, Maki RH: Does distraction reduce
pain produced distress among college students? Health Psy-
chol 11:210-217, 1992
26. Melzack R: The McGill Pain Questionnaire: Major prop-
erties and scoring methods. Pain 1:277-299, 1975
27. Myers CD, Robinson ME, Riley JL, Sheffield D: Sex, gen-
der, blood pressure: Contributions to experimental pain re-
port. Psychosom Med 63:545-550, 2001
28. Rainville P, Feine JS, Bushnell MC, Duncan GH: A psycho-
physical comparison of sensory and affective responses to
four modalities of experimental pain. Somatosens Mot Res
9:265-277, 1992
29. Riley JL, Robinson ME, Wise EA, Myers, CD, Fillingim RB:
Sex differences in the perception of noxious experimental
stimuli: A meta analysis. Pain 74:181-187, 1998
30. Rosenbaum M: Individual differences in self control be-
haviours and tolerance of painful stimulation. J Abnorm Psy-
chol 89:581-590, 1980
31. Smith GM, Egbert LD, Markowitz RA, Mosteller F,
Beecher HK: An experimental pain method sensitive to mor-
phine in man: The Submaximum Effect Tourniquet Tech-
nique. J Pharmacol Exp Ther 154:324-332, 1966
32. Stevens A, Batra A, Kotter I, Bartels M, Schwarz J: Both
pain and EEG response to cold pressor stimulation occurs
faster in fibromyalgia patients than in control subjects. Psy-
chiatry Res 97:237-247, 2000
33. Unruh AM: Gender differences in clinical pain experi-
ence. Pain 65:123-167, 1996
34. Walsh NE, Schoenfeld L, Ramamurthy S, Hoffman J: Nor-
mative model for cold pressor test. Am J Phys Med Rehabil
68:6-11, 1989
35. Weisenberg M, Scharzwald J, Tepper I: The influence of
warning signal timing and cognitive preparation on the
aversiveness of cold-pressor pain. Pain 64:379-385, 1995
36. Woolf CJ: Transcutaneous electrical nerve stimulation
and the reaction to experimental pain in human subjects.
Pain 7:115-127, 1979
37. Worthington EL, Shumate M: Imagery and verbal coun-
selling methods in stress inoculation training for pain con-
trol. J Counsel Psychol 28:1-6, 1981