gender, interpersonal transactions, and the perception of pain: an experimental analysis

Awpp

R1FQsHoSF(AJT1©d

2

ORIGINAL REPORTS

Gender, Interpersonal Transactions, and the Perception of Pain:An Experimental Analysis

Todd Jackson,*,† Tony Iezzi,‡ Hong Chen,§ Stephanie Ebnet,† and Karen Eglitis†

Abstract: Two experiments assessed how interpersonal transactions influence responses to coldpressor pain in women versus men. In Experiment 1, 91 young adults (57 women, 34 men) wererandomly assigned to either a no transaction (NT) condition in which they coped alone with the coldpressor test or a transaction opportunity (TO) condition in which they also had the option ofinteracting with an empathetic, reflecting experimenter. Compared to men, women had lower paintolerance and reported more pain and catastrophizing, although there were no gender differences insupport seeking or other ways of coping. Within the TO condition, women were no more likely thanmen to initiate a transaction, but female speakers were more pain-focused than male speakers, andspeaking with the empathetic interaction partner had generally negative effects on pain perceptionand coping. In Experiment 2, 126 young adults (76 women, 50 men) were randomly assigned to NT,TO, or experimenter-directed (1) Distraction (DT), (2) Reinterpretation (RT), or (3) Encouragement (ET)conditions. Although men had similar levels of pain tolerance across the 5 transaction conditions,women in NT and TO conditions exhibited reduced tolerance compared with those in the DT, RT, andET conditions. Pain tolerance times among women in DT, RT, and ET conditions were equal to orexceeded those of men in these conditions. Together, findings suggest the nature of interpersonaltransactions exerts a greater influence on women’s responses to noxious stimulation than those ofmen.Perspective: This study adds to literature indicating that women exhibit reduced tolerance forexperimentally induced pain compared with men. These results suggest that the nature of interper-sonal transactions also affects women’s responses to noxious stimulation, more than those of men.

© 2005 by the American Pain Society
Key words: Gender, interpersonal transactions, pain.
cted

eodpgdssgtbha

s documented in reviews of clinical and labora-tory studies,9,26,27,36 gender is an important influ-ence on the perception of pain. In general,

omen are more likely than men to experience recurrentain, as well as frequent, severe, and/or longer lastingain,27,36 pain-related disability, and unwarranted psy-

eceived June 16, 2004; Revised October 27, 2004; Accepted December6, 2004.rom the *School of Psychology, James Cook University, Townsville,ueensland, Australia, †Department of Psychology, University of Wiscon-

in–Superior, Superior, Wisconsin, ‡Department of Psychology, Londonealth Sciences Centre, London, Ontario, Canada, and §School of Psychol-gy, Southwest China Normal University, Chong Qing, China.upport for this research was provided from Faculty Development andoundation Fellow grants from the University of Wisconsin-SuperiorT.J.).ddress reprint requests to Todd Jackson, PhD, School of Psychology,

ames Cook University, Townsville, Queensland, Australia 4811. E-mail:[email protected]/$30.002005 by the American Pain Society

voi:10.1016/j.jpain.2004.12.004

28 The Journal of Pain, Vol 6, No 4

hogenic attributions for pain by those who treathem.8,36 Women also have comparatively reduced tol-rance and greater sensitivity for experimentally in-uced noxious stimulation.9,26

Coping, constantly changing cognitive and behavioralfforts to manage demands judged to tax or exceedne’s resources,20 might be one factor explaining genderifferences in responses to stressors in general35 and toain in particular.24,31,32 Although evidence is mixed forender differences in use of problem-focused coping,enial, and looking on the bright side,34,36 there are con-istent differences in use of other ways of coping. Theynthesis of animal and human research by Taylor35 sug-ests that compared to males, females engage in rela-ively more “tending and befriending” and elaboratesiologic and psychosocial underpinnings of females’eightened propensity to seek and provide nurturancend social support. Similarly, Mickelson et al24 and Sulli-
an et al31 assert that women engage in more interper-
(April), 2005: pp 228-236

siptpsdcp

sdatetliswnecestha

tptovaritnrmcmtmtitmf

S

piichv

astct

ascwmSepppwTwttpct

M

P

wac42((cccac

P

Biguhmfvt

ot(w

229ORIGINAL REPORT/Jackson et al

onal or communal coping compared to men when deal-ng specifically with pain. Reviews of the general34 andain coping36 literatures have independently concludedhat women seek instrumental support, emotional sup-ort, and spiritual comfort more than men do. Receivingocial support predicts positive outcomes including re-uced pain and use of analgesics during labor,6,7 de-reases in postoperative pain,19 and reductions in cardiacain.21

If women are more likely to seek social support andupport provision is associated with reduced pain andisability, then why do women often have lower toler-nce and report more pain than men do? At least 2 fac-ors might explain this apparent paradox. First, from anarly age, girls and women are more aware than men ofheir own and others’ emotions11,18 and are also moreikely to report feeling better when expressing feel-ngs,37 perhaps, in part, because parents attend to andupport their expression of sadness, anxiety, and painhile meeting similar expressions of boys with neutral oregative responses.8 Even though females might experi-nce more reinforcement for use of emotion-focusedoping, in some contexts, increased use of certain strat-gies such as rumination, venting emotions, and cata-trophizing25,34,36 might exacerbate pain and dysfunc-ion. For example, catastrophizing has accounted forigher pain levels of women than men in both clinical16

nd healthy samples.32

Second, social support per se might be less importanthan kind of support provided as an influence on painerception. Recent experimental studies2,5 bolster con-entions from clinical research that uncomfortable30 orverly solicitous10 interpersonal transactions can ad-ersely influence pain. Brown et al2 found that eitherctive or passive support from a friend or stranger waselated to less cold pressor pain than coping alone ornteracting with a friend or stranger. They speculatedhat an absence of negative transactions with others andot merely the presence of others (ie, passive support)educed pain. Chambers et al5 found that girls whoseothers interacted with them in a “pain-promoting”

ondition (providing reassurance, empathy, apologies,ild criticism) reported more cold pressor pain than

hose whose mothers interacted with them as they nor-ally would (control condition). In turn, control condi-

ion girls reported more pain than girls whose mothersnteracted with them in a “pain-reducing” manner (dis-raction with nonprocedural talk, humor, direct com-ands to use other coping strategies). Effects were not

ound for boys in these conditions.

tudy 1Together, these studies suggest specific kinds of inter-ersonal transactions are pain-promoting or pain-reduc-

ng. For example, as Chambers et al5 observed, supportnvolving empathy, reflection, and/or reassurance canontribute to increased pain. Perhaps when a sufferer isighly pain-focused in communication, empathy inad-
ertently reinforces the sufferer’s focus on pain. Conceiv- w
bly, if women are relatively more likely to (1) seek socialupport and (2) cope by expressing distress, highly empa-hetic responses from others might contribute to in-reased pain and decreased tolerance, albeit experimen-al studies have yet to directly evaluate these claims.Study 1 assessed how interpersonal transactions char-

cterized by experimenter empathy and reflection in re-ponse to participant communications affected pain per-eption for women versus men. It was hypothesized thatomen would show reduced pain tolerance and reportore intense pain than men during the cold pressor test.

econd, women were expected to have higher levels ofmotion-focused coping (ie, catastrophizing, highlyain-focused speech) and interpersonal coping (ie, sup-ort seeking, initiating conversation when given the op-ortunity to do so) than men, albeit gender differencesere not expected for other cognitive coping strategies.hird, participants who opted to initiate transactionsith an empathetic, reflecting experimenter, especially

hose judged to be highly pain-focused, were expectedo show reduced pain tolerance and report increasedain, emotional support seeking, and catastrophizingompared with those in the same experimental condi-ion who chose not to interact with the experimenter.

aterials and Methods

articipantsThe sample was composed of 91 healthy adults (57omen, 34 men) recruited from undergraduate classest University of Wisconsin–Superior and given extraredit for participation. Participants ranged from 18 to7 years of age (mean, 21.11; standard deviation [SD],.89) and were predominantly single (93.8%), white94.2%), and in their first or second year of college85.5%). Exclusion criteria included presence of a painondition, circulatory disorder, high blood pressure, aardiac condition, Raynaud’s disease, a past seriousold temperature injury, problems with blood clotting,skin condition, and/or use of medication for these

onditions.

rocedureThis research was approved by the Institutional Reviewoard at the University of Wisconsin–Superior. On arriv-

ng at their scheduled appointments, participants werereeted by 1 of 3 experimenters (2 female and 1 malendergraduate psychology students) blind to researchypotheses. They read and signed informed consent andedical release forms and completed a demographic in-

ormation sheet. Next, they were instructed how to pro-ide a rating of reported pain (ie, current pain level be-ween 0, No Pain and 10, Severe Pain).Participants were randomly assigned to read 1 of 2rienting passages reflecting a specific transaction op-ion for the upcoming task. Those in the no transactionNT) condition were instructed to do anything theyanted on their own to cope, but that the experimenter
ould not speak with them, other than to occasionally

aoacEctb

wilglmrqebrispitopa

mtfetprh

cptcotaq

M

P

(uwsaasi

C

unpisLDC

C

asra�t

T

tcrtsetphpwpps9t

R

M

l(npeo

P

tfe��o

230 Gender, Interpersonal Transactions, and Pain

sk for a reported pain rating. Those in the transactionpportunity (TO) condition were also instructed to donything they wished on their own to cope, but theyould also speak with the experimenter if they chose.ach orienting passage concluded with a statement en-ouraging participants to leave their hand in the ice wa-er for as long as possible but to remove it if sensationsecame too uncomfortable.Although potential interactions in the TO conditionere relatively open-ended because they were to be

nitiated by research participants, experimenters fol-owed several general guidelines in responding. Ineneral, experimenters were expected to follow the

ead of participants. For example, statements or com-ents were to be met with reflective or empathetic

esponses by the experimenter. If participants askeduestions, the experimenter would answer them hon-stly and directly. Reassurance, advice, and other feed-ack were not provided unless a participant explicitlyequested them. Finally, until and unless participantsnitiated an interaction, the experimenter did notpeak with them, other than to solicit ratings of re-orted pain. Experimenters were trained in the admin-

stration of the TO condition through standardized,yped instructions, role-playing, discussions with eachther and the principal investigator, and supervisedractice runs with pilot subjects. This training wasbout 3 hours in length.Before the cold pressor test (CPT), participants im-ersed their non-writing hand in room temperature wa-

er for 15 seconds to stabilize skin temperature. Just be-ore immersion in water maintained at 1°C to 3°C, thexperimenter reiterated the instructional set of the par-icipant. At 30-second intervals during the CPT, partici-ants gave a rating of current pain. They were told toemove their hand from the water at 240 seconds if theyad not already done so.Immediately after the CPT, measures of coping were

ompleted in another room. As a manipulation check,articipants were also asked to correctly identify the na-ure of the orienting passage they read from the 2hoices reflecting experimental conditions and a thirdption (ie, “I did not read either of these choices prior tohe cold pressor test”). Finally, after debriefing, e-mailddresses were solicited from participants who re-uested a summary of the results.

easures

ain Tolerance and Reported PainPain tolerance was operationalized as the total time

rounded off to the nearest hundredth of a second andp to 4 minutes) a participant’s hand was immersed in iceater. Ratings of reported pain were solicited every 30

econds on a 0- to 10-point scale with 0 as “No pain at all”nd 10 as “Severe pain.” Following from past research,14

rating of average pain was calculated by dividing theum of reported pain ratings by the number of 30-second
ntervals completed during the task. c
oping Strategies QuestionnaireSelf-reported coping during the CPT was assessed bysing 27 items from the Coping Strategies Question-aire29 (CSQ) subscales, Ignoring Pain (5 items), Reinter-reting Pain Sensations (6 items), Diverting Attention (6

tems), Coping Self-Statements (6 items), and Cata-trophizing (4 items) Each item was rated on a 7-pointikert scale between 0 (Never Did That) and 6 (Very Oftenid That). Coefficient alphas ranged between � � .75 foroping Self-Statements and � � .90 for Catastrophizing.

OPECOPE4 items related to Emotional Support (4 items)

nd Instrumental Support (4 items) were used to assessupport seeking during the CPT. These items were alsoated on 7-point Likert scales between 0 (Never Did That)nd 6 (Very Often Did That). Internal consistencies were� .82 for Emotional Support and � � .72 for Instrumen-

al Support.

ranscripts of Participant VerbalizationsWithin the TO condition, all CPT sessions were audio-

aped to facilitate transcription and analysis of verbalontent of transactions of those who spoke. Poor qualityecordings forced us to eliminate transactions of 3 par-icipants, but those of the 26 others who spoke wereuccessfully transcribed. Two observers blind to hypoth-ses and participant data (eg, gender, pain toleranceime) rated degree of pain-focus in typed transcripts ofarticipant verbalizations. Transcripts were judged toave “little or no pain-focus” when less than 25% ofarticipant speech content was focused on pain, “some-hat pain-focused” when between 25% and 60% ofarticipant speech was focused on pain, and “highlyain-focused” when greater than 60% of participantpeech was focused on pain. Inter-rater agreement was6.2%; the one discordant rating was resolved in consul-ation with the principal investigator.

esults

anipulation ChecksParticipants in TO condition were significantly more

ikely than NT group participants to talk during CPT, �2

1) � 27.15, P � .001, hence supporting the distinctive-ess of experimental conditions. Data from 2 partici-ants in the TO condition and 1 in the NT condition wereliminated because they failed to identify correctly therienting passage they read before the CPT.

reliminary AnalysesFemale and male experimenters tested 45 and 43 par-

icipants, respectively. Transaction conditions did not dif-er for gender of participant, �2 (1) � 1.21, P � .27, orxperimenter, �2 (2) � 0.05, P � .83, marital status, �2 (2)1.25, P � .54, ethnicity, �2 (2) � 2.79, P � .25, age, F1,86

0.15, P � .70, years of education, F1,86 � 1.57, P � .21,r water temperature at onset, F1,86 � 0.61, P � .44, and
essation, F1,86 � 0.42, P � .52, of the CPT. A multivariate

acemm�i

M

owtcpfp

nF3imtpien�.

i.sotts

gt0

rgwslemmb

rt0(cFwddppp.

D

Ficemd

T

PRCEIRDIC

*

*

*


nalysis of variance (MANOVA) on measures of pain per-eption and coping obtained nonsignificant effects forxperimenter gender, F9,72 � 0.66, P � .75, and experi-enter gender � gender, F9,72 � 0.81, P � .61, experi-enter gender � group, F9,72 � 1.28, P � .26, and genderexperimenter gender � group, F9,72 � 0.73, P � .68,

nteractions.

ain AnalysesTwo � two (Gender � Transaction Condition) analysesf variance (ANOVAs) supported the hypothesis thatomen would have lower pain tolerance and more pain

han men during the CPT (Table 1). However, transactiononditions did not differ for pain tolerance or reportedain, and gender did not interact with transaction groupor pain tolerance, F3,84 � 0.001, P � .97, or reportedain, F3,84 � 2.23, P � .14.Regarding emotion-focused/interpersonal coping, sig-ificant multivariate effects were obtained for gender,

3,82 � 3.58, P � .02, and transaction condition, F3,82 �.88, P � .01, but not the gender � transaction conditionnteraction, F3,82 � 3.58, P � .02. Women catastrophized

ore than men but did not report more seeking of emo-ional or instrumental support, and the TO group re-orted more catastrophizing and marginally more seek-

ng of emotional support than the NT group (Table 1). Asxpected, on other cognitive coping strategies, nonsig-ificant multivariate effects were found for gender, F4,81

0.87, P � .49, transaction condition, F4,81 � 0.91, P �46, and their interaction, F4,81 � 0.90, P � .47 (Table 1).

Within the entire sample, pain tolerance was related tognoring, r � .31, P � .004, catastrophizing, r � �.72, P �001, seeking emotional support, r � �.32, P � .002, andeeking instrumental support, r � �.35, P � .001, but notther coping strategies. Reported pain was related onlyo catastrophizing, r � .56, P � .001, and seeking emo-ional support, r � .23, P � .03. After controlling for the

able 1. Differences in Pain Perception and Cop

VARIABLES

EXPERIMENTA

NO TRANSACTION

MALE FEMALE

ain Tolerance (sec) 182.83 (80.70) 140.07 (93.18)eported Pain 6.83 (1.82) 7.07 (1.68)atastrophizing 9.58 (7.34) 11.88 (6.45)motional Support 2.32 (5.49) 1.40 (2.22)nstrumental Support 1.37 (3.98) 0.56 (1.69)einterpretation 11.11 (7.34) 14.24 (8.63)iverting Attention 11.42 (7.54) 11.94 (7.65)

gnoring 13.26 (6.77) 14.52 (6.14)oping Self-Statements 23.89 (7.32) 21.96 (7.50)

P � .10.

*P � .05.

**P � .01.

ole gender difference in coping, ie, catastrophizing, t

ender differences were no longer significant for painolerance, F1,85 � 0.46, P � .50, and reported pain, F1,85 �.41, P � .52.Within the TO group, initiating a transaction was not

elated to gender, �2 (1) � 0.02, P � .88, or experimenterender, �2 (1) � 0.29, P � .59. Among speakers, thereere no gender differences in total participant words

poken, F1,24 � 0.51, P � .48, or words spoken, control-ing for pain tolerance time, F1,23 � 0.18, P � .68. How-ver, speakers showed less pain tolerance and reportedore pain, emotional and instrumental support, andarginally more catastrophizing than non-speakers (Ta-

le 2).Finally, among speakers, degree of pain focus was not

elated to experimenter gender, F1,22 � 0.11, P � .74, orhe gender � experimenter gender interaction, F1,22 �.27, P � .61. However, transcripts of female speakersmean, 2.24; SD, .95) were judged to be more pain-fo-used than those of male speakers (mean, 1.44; SD, .53),

1,22 � 4.29, P � .05; all 9 highly pain-focused transcriptsere from women. After controlling for gender, a highegree of pain-focus in transactions was related to re-uced pain tolerance, r � �.79, P �.001, and catastro-hizing, r � .70, and marginally related to increasedain, r � .36, P � .07, but not related to emotional sup-ort, r � .26, P � .21, instrumental support, r � .13, P �

52, or other coping strategies.

iscussionTwo of three hypotheses received support in Study 1.

irst, findings supported past research,9,16,32,36 observ-ng significant gender differences in pain perception andatastrophizing. Women in Study 1 had lower pain tol-rance, reported more pain, and were more likely thanen to catastrophize about pain they experienced. Gen-

er differences in pain perception vanished after con-

: Study 1 (N � 88)DITION (EC)

TRANSACTION OPPORTUNITY UNVARIATE F

MALE FEMALE GENDER EC

201.92 (65.85) 158.05 (79.80) 5.64** 1.035.76 (1.55) 7.10 (1.62) 4.57** 1.974.21 (4.56) 10.33 (6.77) 8.46*** 5.70**2.64 (3.52) 4.73 (6.37) 0.29 2.86*0.93 (1.90) 1.10 (2.26) 0.32 0.019.79 (9.77) 11.00 (7.43) 1.44 1.587.14 (6.55) 10.10 (9.20) 0.95 2.93*

13.43 (8.27) 12.13 (6.38) 0.01 0.5520.36 (7.39) 20.70 (8.00) 0.22 2.01

ingL CON

rolling for catastrophizing.

aeistiatitpplcvpopaeact

iostftbsdstuouapn

S

ebdmsedmSwectfawahc

M

P

wWg2s

M

tcStp

T(

PRCEIRDIC

A


Furthermore, within the TO condition, speaking withn empathetic experimenter was related to less pain tol-rance, more reported pain, catastrophizing, and seek-ng emotional support, and less ignoring pain. Femalepeakers were also more pain-focused in transactionshan men. Although it is not clear whether empatheticnteractions contributed to heightened pain responsesmong women or greater pain sensitivity motivated cer-ain women to seek support, perhaps gender differencesn degree of pain-focus coupled with the empatheticransaction context contributed to gender differences inain perception. Such transactions might parallel oneotential genesis of maladaptive adjustment wherein

oved ones respond empathically to a communicator fo-used on a novel, emerging pain problem, hence inad-ertently perpetuating his or her focus on pain and ex-eriences of pain and suffering. The results parallel thebservation by Chambers et al5 that pain-promoting sup-ort (which included empathy) increased pain in girlsnd not boys, but they add important qualifier vis-à-vismpathy and pain: empathetic, reflective responses aressociated with exacerbations in experiences of pain forommunicators who are highly pain-focused but not forhose who are lower in pain-focus.In contrast, the hypothesis that women would engage

n more interpersonal coping than men was not corrob-rated; there were no gender differences in supporteeking, and women were not more likely to initiateransactions when given the option of doing so. Perhaps,eatures of the experimental situation undermined po-ential gender differences in support seeking. A largeody of evidence suggests that women cope by seekingupport from friends and family more often than meno.34,35 Nonetheless, they might not indiscriminatelyeek support from unfamiliar others (ie, experimentershey do not know) more than men do. Also, highly eval-ative situations can undercut support seeking.2,35 Dem-nstrating the limit of one’s tolerance for a noxious stim-lus while being observed and, in the TO group, beingudiotaped might have added a sense of heightenederformance evaluation, potentially absent from many

able 2. Difference Between Speakers and NonN � 44)

NON-SPEAKERS

VARIABLE

(n � 15)MEAN (SD)

ain Tolerance 219.92 (46.58)eported Pain 5.89 (1.85)atastrophizing 5.80 (5.58)motional, Support 0.87 (2.85)nstrumental Support 0.07 (0.26)einterpretation 12.0 (8.16)iverting Attention 9.80 (8.65)

gnoring 14.27 (6.90)oping Self-Statements 21.60 (7.49)

bbreviation: SD, standard deviation.

aturalistic settings. R

tudy 2In Study 1, women were no more likely than men to

ngage in support seeking, perhaps as a partial result ofeing in an experiment. Regardless, within the TO con-ition, transactions of female speakers were judged asore pain-focused than male speakers. Study 2 was de-

igned not only to replicate these findings but also toxamine whether other kinds of transactions might haveifferential effects on pain tolerance for women versusen. Following from recent research5 and findings from

tudy 1, we expected women in NT and TO conditionsould have lower pain tolerance relative to women inxperimenter-initiated pain-reducing conditions fo-used on (1) distraction away from pain, (2) reinterpre-ation of uncomfortable sensations to neutral or com-ortable sensations, and (3) encouragement and supportimed at continued pain tolerance. In contrast, menere expected to show similar levels of pain tolerancecross all types of transactions. Study 2 also examinedow gender and transaction conditions influencedoping.

ethod

articipantsParticipants were 126 healthy adults (50 men, 76omen) from undergraduate classes at University ofisconsin–Superior. The sample was predominantly sin-

le (93.7%), white (85.8%), and young adult (mean,1.15 years; SD, 4.71). Exclusion criteria were also theame as those used in Study 1.

easuresMeasures of pain tolerance and coping were identical

o those used in Study 1. Adapted CSQ scale internalonsistencies ranged between � � .75 for Coping Self-tatements and � � .88 for Catastrophizing, whereashose for COPE Emotional Support and Instrumental Sup-ort subscales were � � .82 and � � .83, respectively.

eakers in Transaction Opportunity Condition

SPEAKERS

(n � 29)MEAN (SD) UNIVARIATE F P VALUE

47.23 (79.46) 10.59 .0017.08 (1.50) 5.32 .039.72 (6.99) 3.54 .075.72 (6.07) 8.55 .011.55 (2.49) 5.26 .039.86 (8.18) 0.72 .408.83 (8.54) 0.13 .72

11.66 (6.95) 1.40 .2420.07 (7.92) 0.38 .54

-Sp

1

atings of pain were not taken in Study 2 because they

cr

S

pst

GP

litrsrfadw

P

fadt(rtpppptqt(twrarsctm

atomsi

ew

1itfttwwp

R

P

tfteaCsg�idgg�

M

a�atdswm

FtTE


ould potentially disrupt/interrupt experimenter-di-ected transactions.

hyness ScaleThe 9-item Shyness Scale3 (SS) assessed pre-task inter-ersonal discomfort and inhibition on a 5-point Likertcale. Higher scores reflected increased shyness. The in-ernal consistency of the SS was � � .85.

roup Membership and Transcripts ofarticipant VerbalizationsTwo raters blind to group membership of participants

istened independently to all audiotaped sessions involv-ng interpersonal transaction conditions and were askedo correctly identify the group of each respondent. Bothaters correctly identified transaction conditions of allubjects in each group, supporting the treatment integ-ity of each condition. Consonant with Study 1, two dif-erent raters, blind to hypotheses and participant data,ssessed typed transcripts of speakers within the TO con-ition for degree of pain-focus. Their rate of agreementas 100% in Study 2.

rocedureResearch volunteers read and signed informed consent

orms and then completed a demographic data sheetnd the SS. Within each gender, participants were ran-omly assigned to 1 of 5 transaction options so condi-ions would be equal in their relative numbers of womenn � 15) and men (n � 10). Before the CPT, participantsead an orienting passage describing the group to whichhey had been assigned: (1) NT condition (same orientingassage as Study 1); (2) TO condition (same orientingassage and experimenter guidelines as Study 1); (3) Ex-erimenter-Directed Distraction (DT) condition (partici-ants instructed they could do anything they wished onheir own to cope, and experimenter would ask themuestions about their lives throughout the task to dis-ract them); (4) Experimenter-Directed ReinterpretationRT) condition (participants instructed to do anythinghey wished on their own to cope, and experimenterould also talk with them throughout the task to foster

einterpretation of uncomfortable sensations to neutralnd/or comfortable sensations); (5) Experimenter-Di-ected Encouragement (ET) condition (participants in-tructed to do anything they wished on their own toope, and experimenter would also talk with themhroughout the task to offer support and encourage-ent for persisting at the task).To reduce subject attrition from misidentification of

ppropriate transaction condition, participants then hado identify the orienting passage they read on the basisf short descriptions of the above options. Those whoade an error were required to re-read the initial pas-

age and correctly identify their group before engagingn the CPT.

Procedures were identical to those used in Study 1,xcept that (1) all 4 interpersonal transaction conditions
ere audiotaped, (2) in DT, RT, and ET conditions, after m
0 seconds of immersion, the experimenter directed annteraction based on the specific instructions reflected inhat condition, (3) there was 1 new male and 1 newemale experimenter, (4) nearly 7 hours were needed torain experimenters in administrating transaction op-ions, and (5) after completing coping items, participantsere asked a second time to identify the condition tohich they had been assigned. In Study 2, all participantsrovided correct responses to this item.

esults

reliminary AnalysesFemale and male experimenters tested 57 and 69 par-

icipants, respectively. Transaction conditions did not dif-er on gender of participant, �2 (1) � 0.09, P � .99, orester, �2 (2) � 0.69, P � .95, age, F4,120 � 0.46, P � .77,ducation, F4,120 � 0.57, P � .69, P � .68, water temper-ture at onset, F4,120 � 1.94, P � .11, and cessation ofPT, F4,120 � 1.56, P � .19. There were no differences inhyness for transaction condition, F4,120 � 0.30, P � .88,ender, F1,120 � 0.71, P � .40, or their interaction, F4,120

0.99, P � .41. A MANOVA on pain tolerance and cop-ng found nonsignificant effects for experimenter gen-er, F8,97 � 0.83, P � .58, and experimenter gender �ender, F8,97 � 1.03, P � .42, experimenter gender �roup, F32,400 � 1.10, P � .33, and experimenter gendergender � group, F32,400 � 0.70, P � .89, interactions.

ain AnalysesA 2 � 5 (Gender � Transaction) ANOVA on pain toler-

nce found significant effects for gender, F1,116 � 4.07, P.05, transaction, F4,116 � 2.76, P � .03, and their inter-

ction, F4,116 � 2.67, P � .04. Fig 1 illustrates this interac-ion. Men had similar levels of pain tolerance across con-itions, but women in NT and TO groups hadubstantially reduced pain tolerance compared toomen in the 3 experimenter-directed groups whoseean tolerance times equaled or exceeded those of their

igure 1. Differences in pain tolerance by gender and transac-ion condition. Transaction condition: 1, No Transaction; 2,ransaction Opportunity; 3, Distraction; 4, Reinterpretation; 5,ncouragement. Each error bar equals 50 seconds.

ale counterparts.

sg�.ggsotg�

am0tk.lgPw(sAt–ns

D

twphoidb

oosfpmpt

iempdiet

tiscttoadm

hoectirtibt

T

RDIECCI

T tion; 4


A MANOVA on reported support seeking and cata-trophizing obtained a significant effect for transactionroup, F12,345 � 4.87, P � .001, but not for gender, F3,113

1.17, P � .33, or their interaction, F12,345 � 0.52, P �90. Groups did not differ on catastrophizing, but the NTroup reported less emotional support than all otherroups, and the RT group reported more instrumentalupport than all other groups (Table 3). In a MANOVA onther coping strategies, a statistical trend was found forransaction condition, F16,456 � 1.58, P � .07, but not forender, F4,111 � 1.42, P � .23, or their interaction, F16,456

0.60, P � .88.Within the TO group, initiating a conversation was not

ssociated with gender, �2 (1) � 0.03, P � .86, or experi-enter gender, �2 (1) � 1.72, P � .19, or shyness, F1,21 �

.02, P � .90. Male and female speakers did not differ inotal words spoken, F1,15 � 1.48, P � .24, or words spo-en, controlling for pain tolerance time, F1,14 � 0.03, P �

87. Among speakers, degree of pain focus was not re-ated to experimenter gender, F1,13 � 0.82, P � .38, or theender � experimenter gender interaction, F1,13 � 1.91,� .19. However, female speakers (mean, 2.40; SD, .84)ere judged to be more pain focused than male speakers

mean, 1.42; SD, .53), F1,13 � 7.91, P � .02; the 6 tran-cripts judged as highly pain-focused were from women.fter controlling for gender, increased pain-focus in

ransactions was related to reduced pain tolerance, r �.79, P � .001, and catastrophizing, r � .49, P � .05, butot with emotional support, r � .28, P � .26, or othertrategies.

iscussionIn Study 2, women in NT and TO groups had lower pain

olerance than women in DT, RT, and ET conditions,hereas men exhibited relatively consistent levels ofain tolerance, regardless of the condition to which theyad been assigned. In part, this pattern mirrors recentbservation5 that girls’ tolerance for cold pressor pain

ncreased in a condition of pain-reducing support andecreased when support was pain-promoting, whereas

able 3. Between-Condition Differences in Self-

COPING STRATEGY

TRAN

1 2

(n � 25)Mean (SD)

(n � 25)Mean (SD)

e interpretation 10.52 (7.51) 9.62 (7.04)iverting Attention 8.16 (7.04) 9.54 (8.08)

nstrumental Support 0.32 (0.80) 3.35 (4.10)motional Support 0.72 (1.40) 7.42 (6.49)oping Self-Statements 16.32 (6.18) 15.12 (6.36)atastrophizing 5.12 (4.32) 7.92 (6.21)

gnoring 18.28 (8.32) 16.12 (6.19)

ransaction condition: 1, No Transaction; 2, Transaction Opportunity; 3, Distrac

oys’ pain tolerance did not change as a function of type l

f support. Study 2 extends that study (1) from samplesf girls and boys to healthy adults, (2) from use of a closeource of support (mother) to use of a less familiar, moreormal, peer support, and (3) from pain-promoting andain-reducing conditions that included a heterogeneousix of strategies to more highly specific kinds of inter-

ersonal transactions (ie, empathetic, distracting, rein-erpreting, encouraging).Together, these findings suggest that the nature of the

nterpersonal transactions had a greater impact on wom-n’s responses to noxious stimulation than on those ofen. Whereas empathy and reflection were related to

ain-focused communications from a subset of women,istraction, encouraged perseverance, and especially re-

nterpretation of painful sensations were related to tol-rance levels for women that were equal to or exceededhose of men.There were no gender differences in coping, although

ransaction conditions had some differences. Not surpris-ngly, NT respondents reported less emotional supporteeking than those in the 4 interpersonal transactiononditions. Also, the RT group reported more instrumen-al support than other groups, perhaps because theirransactions were characterized by direct confrontationsf pain and specific guidance for transforming pain rel-tive to DT and ET conditions in which pain was neitherirectly acknowledged nor addressed by an experi-enter.Instructions used for transaction conditions mightelp to explain why these conditions did not differ onther coping strategies wherein differences might bexpected. Specifically, participants in all transactiononditions were instructed, in part, to do whateverhey wanted on their own to cope, as well as engagingn a specific kind of transaction. This was done becauseestricting participants to using only strategies relatedo their transaction condition could introduce report-ng biases that might not reflect actual coping, andecause many people use multiple strategies, ratherhan isolated methods, when dealing with pain in real-

orted Coping: Study 2N CONDITION

POST HOC

COMPARISONS

3 4 5

� 25)n (SD)

(n � 25)Mean (SD)

(n � 26)Mean (SD)

6 (7.84) 12.60 (7.25) 10.48 (6.44) No differences0 (6.32) 8.88 (7.21) 11.28 (9.85) No differences0 (3.22) 8.32 (5.86) 3.40 (3.86) 4 �1, 2, 3, 52 (5.66) 11.08 (5.28) 7.16 (5.03) 1 � 2, 3, 4, 5,8 (6.04) 20.12 (5.40) 17.92 (7.11) No differences3 (5.80) 7.60 (6.26) 6.04 (6.72) No differences9 (7.88) 19.32 (7.23) 19.80 (8.93) 3 �2

, Reinterpretation; 5, Encouragement.

RepSACTIO

(nMea

12.411.02.55.4

17.88.3

23.7

ife situations. Thus, participants’ use of both self-gen-

eacRt(fitiswtwiT

G

dmubetwmNpcamcpl

etselwlFwtwaicfa

cetmfm

slletpapo

utatIhwtasctsttspifes

rhrsFrgg

dptaticctdCi

acaph


rated and transaction-specific strategies might havettenuated group differences on coping. Furthermore,ontent of certain transaction conditions (Distraction,einterpretation) might have diverged from item con-ent of certain coping scales having surface similaritiesDiverting Attention, Reinterpretation). For example,act-based autobiographical questions, not specificallynvolving pleasant content, were used in the Distrac-ion condition, whereas several Diverting Attentiontems reflect thinking about something pleasant. As aecond example, in the RT condition, reinterpretationsere based on participant-generated descriptors for

heir sensations that did not necessarily correspondith content of CSQ Reinterpretation items (eg, think-

ng of sensations as numbness rather than pain).herefore, some of these items might not be endorsed.

eneral DiscussionThe present findings add to an extensive literature in-icating women exhibit reduced tolerance for experi-entally induced pain compared to men. More conspic-ously, although biologic variability within women haseen identified as a potential source of gender differ-nces in pain perception,1,9 the current results suggesthat the nature of interpersonal transactions also affectsomen’s responses to noxious stimulation significantlyore than those of men. In 2 experiments, women in anT (coping alone) condition had lower tolerance for coldressor pain than men in the same condition. Moreover,ompared with male speakers, female speakers in trans-ctions characterized by empathetic, reflecting experi-enter responses were also more pain-focused. In this

ontext, women’s awareness and relatively greater pro-ensity toward expressing pain-related distress were re-

ated to lower pain tolerance.In contrast, gender differences in pain tolerance were

liminated in transaction conditions focused on distrac-ion, support for perseverance, and reinterpretation ofensations. Notably, the mean tolerance time for womenncouraged to use reinterpretation was 30 secondsonger than that of men in the same condition. In otherords, transactions that did not foster pain-focus or iso-

ation seemed to increase pain tolerance among women.indings cannot be overextended, given considerableithin gender variability, yet because of their motiva-

ion to make and sustain connections with others,24,35

omen’s pain experiences might be better understoodnd treated when relevant interpersonal influences aredentified. In contrast, a task-oriented, less relational fo-us, favored by many men, possibly immunizes themrom transactions that are potentially pain-promotingnd pain-reducing.Regarding treatment implications, research on the effi-

acy of attention-focusing versus attention-diverting strat-gies has been mixed. An early meta-analysis33 concludedhat attention directed away from pain (avoidance) wasore effective for coping with short-term pain, whereas

ocused attention was more effective for pain persisting for
ore than 2 weeks. These contentions have received some c
upport. For example, pain severity and distress were re-ated to increased attentional diversion among phantomimb patients.12 Other studies have obtained contradictoryffects. In research on cold pressor pain,17 focused atten-ion was associated with less sensory pain than avoidingain, albeit for men but not women. Even more recently,cceptance of pain was consistently found to be a strongerredictor of adjustment to chronic pain compared withther coping approaches.23

Differences in samples and pain outcomes might contrib-te to conflicting evidence on this issue. However, some ofhe present results suggest that attentional focus toward orway from pain is of less import in predicting outcomeshan how attention is directed toward or away from pain.n both experiments, higher degrees of focus on pain andow much it hurt in transactions within the TO conditionere associated with lower pain tolerance and more ca-

astrophizing. In contrast, the RT group in Experiment 2lso attended to pain but attempted to reinterpret painfulensations into less painful ones. These participants, espe-ially women, exhibited significantly more pain tolerancehan TO participants. Thus, merely attending to and cata-trophizing about pain were less adaptive than attendingo pain and attempting to transform the subjective percep-ion of painful sensations. This notion of focusing on sen-ations and reinterpreting them bears semblance to com-onents of potentially effective interventions,15,23 which

ncorporate acknowledgement of pain and perhaps trans-orm its meaning from an experience that produces anxi-ty, fear, and rumination to one that is met more dispas-ionately.These implications aside, the main limitations of this

esearch must be acknowledged. Obviously, becauseealthy young adult samples comprised the experiments,esearch with clinical pain samples is needed to fully as-ess the generalizability and applied value of findings.urthermore, it might be useful to assess how genderole expectations influence responses to transactions,iven that gender roles can mediate relations betweenender and pain perception.28

A limitation of Study 2 was that transaction conditionsiffered not only in content but also in whether partici-ant (TO condition) or experimenter (DT, RT, ET condi-ions) instigated a transaction. This limitation does notlter the conclusion that the nature of interpersonalransactions might influence women’s responses to nox-ous stimulation more than those of men, but it does lendaution to assuming that the content of transactions ac-ounted fully for between-conditions differences in painolerance. In future research, inclusion of experimenter-irected pain-promoting conditions as in the study byhambers et al5 would help to assess the possible role of

nitiating a transaction as an influence on tolerance.Finally, although findings suggest some kinds of trans-

ctions influence tolerance for experimental pain, espe-ially among women, they do not imply such transactionsre more and less adaptive across a wider spectrum ofain experiences. For example, empathy shown forighly pain-focused persons with tissue damage might
ontribute to help-seeking and proper care, whereas dis-

tvs

ct

R

13

2s2

3c

4sc

5tm

6Hct

7lG

8iP

9m1

1I

1U

1p

1m

1Go

1D

1Cis

1am

1p6

1s


raction, reinterpretation, and encouragement to perse-ere might perpetuate underlying nociception in the
ame instance. Consequently, both experimental and v
urgery. Health Psychol 8:221-238, 1989

2N

2cH

2p

2t

2mStJ

2HY

2Ss

2bis4

2It

2cc

3aC

3MoP

3pT

3nc

3ct

3B

3P

3a

linical research is needed to identify interpersonal con-exts that facilitate and hinder adaptation for a wider
ariety of acute and chronic pain experiences.13,22
eferences

. Berkley KJ: Sex differences in pain. Behav Brain Sci 20:71-380, 1997

. Brown JL, Sheffield D, Leary MR, Robinson ME: Socialupport and experimental pain. Psychosom Med 65:276-283,003

. Cheek JM, Buss A: Shyness and sociability. J Pers Soc Psy-hol 41:330-339, 1981

. Carver CS, Scheier MF, Weintraub JK: Assessing copingtrategies: A theoretically-based approach. J Pers Soc Psy-hol 56:267-283, 1989

. Chambers CT, Craig KD, Bennett SM: The impact of ma-ernal behaviour on children’s pain experiences: An experi-ental analysis. J Paediatr Psychol 27:293-301, 2002

. Chalmers B, Wolman W, Nikodem V, Gulmezoglu A,ofmeyer G: Companionship in labour: Do the personalityharacteristics of labour supporters influence their effec-iveness? Curationis 18:77-80, 2000

. Cogan R, Spinnato JA: Social support during prematureabor: Effects on labor and the newborn. J Psychosom Obstetynecol 8:209-216, 1988

. Dunn J, Bretherton I, Munn P: Conversations about feel-ng states between mothers and their young children. Devsychol 23:132-139, 1987

. Fillingim RB: Sex-related influences on pain: A review ofechanisms and clinical implications. Rehabil Psychol 48:

65-174, 2003

0. Fordyce WE: Behavioral Methods in Chronic Pain andllness. St Louis, MO, Mosby, 1976

1. Gilligan C: In a Different Voice. Boston, MA, Harvardniversity Press, 1982

2. Hill A: The use of pain coping strategies by patients withhantom limb pain. Pain 55:347-353, 1993

3. Hill A, Niven CA, Knussen C: The role of coping in adjust-ent to phantom limb pain. Pain 62:79-86, 1995

4. Jackson T, Iezzi T, Gunderson J, Fritch A, Nagasaka T:ender differences in pain perception: The mediating rolef self-efficacy beliefs. Sex Roles 47:561-568, 2002

5. Kabat-Zinn J: Full Catastrophe Living. New York, NY,elacorte Press, 1990

6. Keefe FJ, Lefebyre JC, Egert JF, Affleck G, Sullivan MJ,aldwell DS: The relationship of gender to pain, pain behav-

or and disability in osteoarthritis patients: The role of cata-trophizing. Pain 87:325-334, 2000

7. Keough E, Hatton K, Ellery D: Avoidance versus focusedttention and the perception of pain: Differential effects foren and women. Pain 85:225-230, 2000

8. Kring AM, Gordon AH: Sex differences in emotion: Ex-ression, experience, and physiology. J Pers Soc Psychol 74:86-703, 1998

9. Kulik JA, Mahler HI: Social support and recovery from

0. Lazarus RS, Folkman S: Stress, Appraisal, and Coping.ew York, NY, Springer, 1984

1. Lidderdale JM, Walsh JJ: The effects of social support onardiovascular reactivity and perinatal outcomes. Psycholealth 13:1061-1070, 1998

2. McCracken L: Attention to pain in persons with chronicain: A behavioral approach. Behav Ther 28:271-284, 1997

3. McCracken L, Eccleston C: Coping or acceptance: Whato do about chronic pain? Pain 107:159-167, 2003

4. Mickelson KD, Lyons RF, Sullivan MJL, Coyne JC: Yours,ine, ours: The relational context of communal coping, in

arason BR, Duck S (eds): Personal Relationships: Implica-ions for Clinical and Community Psychology. New York, NY,ohn Wiley & Sons, 2001, pp 181-201

5. Nolen-Hoeksema S: Gender differences in depression, inammen C, Gotlib I (eds): Handbook of Depression. Nework, NY, Guilford Press, 2002, pp 492-509

6. Riley JL, Robinson MF, Wise EA, Myers CD, Fillingim RB:ex differences in the perception of noxious experimentaltimuli: A meta-analysis. Pain 74:181-187, 1998

7. Robinson ME, Riley JFIII, Myers CD: Psychosocial contri-utions to sex-related differences in pain responses, in Fill-

ngim RB (ed): Sex, Gender, and Pain: Progress in Pain Re-earch and Management. Seattle, WA, IASP Press, 2000, pp1-68

8. Robinson ME, Wise EA, Gagnon C, Fillingim RB, Price DD:nfluences of gender role and anxiety on sex differences inemporal summation of pain. J Pain 5:77-82, 2004

9. Rosenstiel A, Keefe F: The use of coping strategies inhronic low back pain patients: Relationships to patientharacteristics and current adjustment. Pain 17:33-44, 1983

0. Schwartz L, Slater MA, Birchler GR: Interpersonal stressnd pain behaviour in patients with chronic pain. J Consultlin Psychol 61:861-864, 1994

1. Sullivan MJL, Thorn B, Haythornthwaite JA, Keefe F,artin M, Bradley LA, Lefebvre JC: Theoretical perspectives

n the relation between catastrophizing and pain. Clin Jain 17:52-64, 2001

2. Sullivan MJL, Tripp DA, Santor D: Gender difference inain and pain behavior: The role of catastrophizing. Cognher Res 24:121-134, 2000

3. Suls J, Fletcher B: The relative efficacy of avoidant andon-avoidant coping strategies: A meta-analysis. Health Psy-hol 4:249-288, 1985

4. Tamres LK, Janicki D, Helgeson VS: Sex differences inoping behavior: A meta-analytic review and an examina-ion of relative coping. Pers Soc Psychol Rev 6:2-30, 2002

5. Taylor SE: The Tending Instinct. New York, NY, Timesooks, 2002

6. Unruh AM: Gender variations in clinical pain experience.ain 65:123-167, 1996

7. Zeman J, Shipman K: Children’s expression of negative
ffect: Reasons and methods. Dev Psychol 32:842-849, 1996

gender, interpersonal transactions, and the perception of pain: an experimental analysis

Documents