Professional Documents
Culture Documents
www.elsevier.com/locate/pain
Received 2 June 2005; received in revised form 30 September 2005; accepted 10 October 2005
Abstract
Although insecure attachment has been associated with a range of variables linked with problematic adjustment to chronic pain, the causal
direction of these relationships remains unclear. Adult attachment style is, theoretically, developmentally antecedent to cognitions, emotions
and behaviours (and might therefore be expected to contribute to maladjustment). It can also be argued, however, that the experience of
chronic pain increases attachment insecurity. This project examined this issue by determining associations between adult attachment
characteristics, collected prior to an acute (coldpressor) pain experience, and a range of emotional, cognitive, pain tolerance, intensity and
threshold variables collected during and after the coldpressor task. A convenience sample of 58 participants with no history of chronic pain
was recruited. Results demonstrated that attachment anxiety was associated with lower pain thresholds; more stress, depression, and
catastrophizing; diminished perceptions of control over pain; and diminished ability to decrease pain. Conversely, secure attachment was
linked with lower levels of depression and catastrophizing, and more control over pain. Of particular interest were findings that attachment
style moderated the effects of pain intensity on the tendency to catastrophize, such that insecurely attached individuals were more likely to
catastrophize when reporting high pain intensity. This is the first study to link attachment with perceptions of pain in a pain-free sample.
These findings cast anxious attachment as a vulnerability factor for chronic pain following acute episodes of pain, while secure attachment
may provide more resilience.
q 2006 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
There has been a great deal of interest in the process by chronic pain (Mikail et al., 1994). A critical assumption of
which acute pain becomes chronic (Aronoff & Feldman, this proposition is that insecure attachment (and associated
2000; Epping-Jordan et al., 1998; Young, 2003). In the maladaptive cognitions, behaviours and emotions) precedes
absence of adequate physiological and demographic chronic pain and thus represents a vulnerability factor for
predictors of chronicity, the contributing roles of a range the development of chronic pain. Nevertheless, it might
of cognitive and behavioural factors have been extensively equally be argued that the experience of chronic pain
investigated (Severeijns et al., 2005). Nevertheless, precise promotes insecure attachment patterns. Hence, determining
mechanisms remain speculative. associations between adult attachment and acute pain are an
Adult attachment theory has been proposed as a important intermediate stage in understanding of the role of
developmental framework for chronic pain (Anderson & attachment in the transition from acute to chronic pain.
Hines, 1994; Kolb, 1982; Mikail et al., 1994). In particular, The adult attachment literature recognizes two dimen-
the negative cognitive dispositions, emotional distress, and sions of attachment reflecting comfort with closeness
self-sabotaging behaviours associated with insecure attach- (comfort) and anxiety over relationships (anxiety) (Feeney
ment, have been proposed to contribute to the etiology of et al., 1994; Fraley & Waller, 1998; Sanford, 1997).
Individuals display predictable patterns of cognitions,
behaviours and emotions consistent with high or low
* Corresponding author. Tel.: C61 7 3365 2652; fax: C61 7 3365 1622. comfort and high or low anxiety; hence, these two
E-mail address: p.meredith@uq.edu.au (P.J. Meredith). dimensions define four attachment styles. These styles
0304-3959/$20.00 q 2006 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.pain.2005.10.008
P.J. Meredith et al. / Pain 120 (2006) 4452 45
Table 2 have been reported (Crawford & Henry, 2003; Lovibond &
Descriptive data Lovibond, 1993), internal consistency, temporal stability,
Variable Mean SD Range and convergent and discriminant validity, are excellent, and
Age 35.69 years 12.24 2169 years a consistent factor structure has been found (Brown et al.,
Secure 4.89 1.42 17 1997; Clara et al., 2001). Sample items include: I found it
Fearful 2.93 1.74 17 hard to wind down (stress), I experienced trembling
Preoccupied 2.39 1.46 17 (anxiety), and I could not seem to experience any positive
Dismissing 4.14 1.89 17
feeling at all (depression). Reliability scores in the present
Attachment 72.27 17.49 27123
avoidance project were .80 for the stress scale, .57 for the anxiety scale,
Attachment 38.23 8.58 1962 and .84 for the depression scale. As a result of the low
anxiety reliability finding for anxiety, this scale was removed from
Threshold 55.88 s 76.25 3240 s further analyses.
Tolerance 173.93 s 86.85 16240 s
The Relationship Questionnaire (RQ, Bartholomew &
Average pain 53.28 21.85 10100
intensity Horowitz, 1991)pre-task. In this self-report questionnaire
Overall pain 5.02 2.59 010 an individual selects one of four paragraphs, representing
intensity secure, preoccupied, fearful-avoidant, or dismissive-avoi-
Minimum VAS 33.22 25.45 0100 dant attachment styles. In addition, participants rate, on a
Maximum VAS 65.67 22.45 11100
seven-point scale (where 1Znot at all like me, and 7Zvery
1 min VAS 47.64 26.15 5100
Stress 10.17 7.52 030 much like me), the extent to which they believe each
Depression 4.83 6.53 034 attachment style reflects their general relationship style. The
Average 1.04 .93 03.5 RQ is a widely used measure of attachment showing
catastrophizing adequate predictive validity and test-retest reliability
Control 3.97 .96 26
(Herzberg et al., 1999; Scharfe & Bartholomew, 1994).
Decrease pain 3.71 .97 26
Average post- .86 1.08 04.2 Attachment Style Questionnaire (ASQ, Feeney et al.,
catastrophizing 1994)pre-task. This self-report measure of adult attach-
Post-control 4.02 1.47 06 ment style identifies two major attachment dimensions:
Post-decrease 3.02 1.59 06 discomfort with closeness (or avoidance) and anxiety over
pain
relationships (anxiety). Participants rate 40 items on a six-
point scale (1Ztotally disagree to 6Ztotally agree). In the
2.3. Apparatus and measures present study, alpha coefficients for the two dimensions
were .92 (discomfort) and .82 (anxiety), with scales
The coldpressor used in this study was an insulated consisting of 27 and 12 items respectively. An example of
container divided into two compartments by wire mesh, a discomfort item is: I prefer to keep to myself, while a
with ice on one side and ice-free chilled water on the other. sample anxiety item is: It is important to me that others like
The tank was tall enough for the entire forearm to be me.
immersed. Water was maintained at 2 8C, and was Coping Strategies QuestionnaireCatastrophizing
circulated by a submerged pump which was switched off Subscale (CSQ-Cat, Rosenstiel & Keefe, 1983)pre-
during testing (Bandura et al., 1987; Geisser et al., 1992). task and post-task. The catastrophizing subscale is a
Instructions for use of the coldpressor were standard for 6-item self-report measure of pain catastrophizing
each participant. Although participants were not informed (Swartzman et al., 1994). Participants rate items on a
prior to the task, there was a four-minute ceiling. seven-point scale (0Znever do that to 6Zalways do that).
Coldpressor measures are described below. A sample item is: It is terrible and I feel it is never going
Participants provided several demographic details to get any better. The pre-treatment catastrophizing alpha
including age, gender, marital status, income, education, coefficient was .87, while post-treatment alpha was .90.
employment status, and country of origin. In addition, Prior to the task, individuals were instructed to indicate
several questionnaires were completed prior to the how much you engage in that activity when you feel
coldpressor task, along with measures during, and a further pain, while after the task participants were asked to
brief questionnaire after, the task. Measures included: indicate how much you engaged in that activity during
Depression Anxiety Stress Scales 21(DASS21, Lovibond the current experimental task. In addition, prior to testing,
& Lovibond, 1993; 1995)pre-task. This 21-item ques- patients rated how much control they believe they would
tionnaire is a brief version of the 42-item DASS. have over the pain (0Zno control, to 6Zcomplete
It comprises three scales of seven items (stress, depression, control) and how successful they would be in decreasing
and anxiety). Questions are scored on a four-point scale their pain (0Zcannot decrease it at all, to 6Zcan decrease
(0Zdid not apply to me at all, to 3Zapplied to me very it completely). After the task they used the same scales to
much or most of the time) based on the previous week. record how much control of, and ability to decrease, the
Adequate reliability, convergent and discriminant validity pain they actually did have.
P.J. Meredith et al. / Pain 120 (2006) 4452 47
Coldpressor ThresholdThis was the amount of time (in the means revealed that women reported being more secure
seconds) after commencing the task that the individual and less avoidant than males in the sample.
reported experiencing pain, recorded with a stopwatch.
Coldpressor ToleranceThis was the total period of 3.3. Correlations among variables
time (in seconds) that the individual tolerated immersion in
the water, recorded with a stopwatch. Correlations among the four continuous RQ attachment
Pain IntensitySeveral measures of pain intensity were scales, two ASQ dimensions, and all other variables are
adopted during and after the task. Pain ratings were presented in Table 3. Secure attachment was linked with
collected at 20-s intervals during the activity (12 scores lower age, more stress and control over pain prior to the
over 4 min), and these results were averaged to determine task, and less catastrophizing (both pre- and post-task).
Average Pain Intensity. Scores representing minimum and Fearful attachment was related to more pre-task catastro-
maximum pain ratings were also documented from these phizing, and lower ratings of pain intensity at one minute.
results. Overall Pain Intensity was indicated on the post-task Preoccupied attachment was related to increased pre-task
questionnaire by requesting the participant to mark, on a 10- catastrophizing, and less control over the pain after the task.
point scale (0Zno pain, to 10Zpain as bad as it could be), Dismissing attachment was not associated with any of the
how painful they rated the coldpressor task overall. variables.
ASQ discomfort with closeness was positively correlated
only with age. ASQ anxiety over relationships, on the other
hand, was linked with more pre-task catastrophizing, stress,
3. Results
and depression, and lower pre-task control over pain, pre-
task ability to decrease pain, and pain threshold. Age was
3.1. Statistics negatively associated with anxiety over relationships.
Attachment variables were predictably linked together.
Analyses were undertaken using SPSS 13.0. Missing data The low correlation between pre- and post-catastrophizing
resulted in a slightly decreased N for some analyses. is particularly noteworthy.
A power plot showing power as a function of effect size was
generated (Cohen, 1988). This suggests that, with 58 3.4. Associations between RQ attachment categories
participants and a power of 0.8, a moderate effect size and other variables
should be detected. Data checking strategies revealed no
outliers, and 11 skewed variables. Ten variables were A series of one-way ANOVAs were conducted with the
positively skewed (age, pre- and post-catastrophizing, categorical RQ attachment as the independent variable and
fearful and preoccupied attachment, stress and depression, each of the other variables as dependent variables.
threshold, minimum pain intensity, and average pain Attachment categories were significantly associated
intensity), and these were transformed using logarithmic with pre-task catastrophizing (F(3, 51)Z3.8, PZ.02), age
transformations. One variable (secure attachment) was (F(3, 51)Z2.77, PZ.05), and depression (F(3, 51)Z3.4,
negatively skewed, and was therefore reflected before PZ.03). Post hoc analyses revealed that those reporting
being transformed logarithmically. Because results using dismissing attachment were older than other participants,
the transformed variables did not differ substantially from and were significantly more depressed (PZ.03) than those
those based on untransformed variables, only original reporting secure attachment styles. Although not reaching
analyses are reported in this paper. Unless otherwise significance, the extent of catastrophizing reported by
stated, the words secure, fearful, preoccupied, and dismissing participants was lower compared to those
dismissing refer to continuous RQ variables. Categorical indicating fearful or preoccupied attachment. In general,
RQ variables are designated as such. secure individuals were less depressed and catastrophized
less than the other styles, with two findings reaching
3.2. Demographic and descriptive details significance (see Table 4). There were no significant
findings for pain intensity, threshold or tolerance variables.
Descriptive details for continuous variables are provided With age included as a covariate in these analyses, and
in Table 2. It is interesting to note the wide range in the incorporating interaction terms, the associations between
threshold and tolerance variables, highlighting the relatively depression and attachment (PZ.63) and catastrophizing and
subjective nature of pain perceptions. Age was included in attachment (PZ.07) were no longer statistically significant.
correlation analyses (see results below) and, because it was Parameter estimates compared three categories of the RQ
linked with both attachment and pain threshold variables, it variable with the fourth category (dismissing), revealing
was included as a covariate in further analyses. A series of that the fearful attachment category (BZ22.54, tZ2.65,
ANOVAs revealed that gender was related to secure PZ.01) was significantly more strongly associated with
attachment (F(1, 55)Z4.61, PZ.04) and attachment catastrophizing relative to the dismissing category. Accord-
avoidance (F(1, 53)Z15.04, P!.001). Inspection of ing to parameter estimates, fearful attachment also
48 P.J. Meredith et al. / Pain 120 (2006) 4452
Table 3
Intercorrelations between the variables
Variable 1 2 3 4 5 6 7 8 9
1 RQ secure 1
2 RQ fearful .10 1
3 RQ preocc K.01 .26* 1
4 RQ dismiss K.23 .04 K.14 1
5 Discomfort K.56*** .28* .12 .26* 1
6 Anxiety K.04 .18 .38** K.41** .002 1
7 Stress .26* .10 K.05 K.01 K.09 .38** 1
8 Depression K.20 .20 .23 .20 .22 .33** .53*** 1
9 Catastroph K.40** .31* .33** .07 .23 .46*** .18 .34** 1
10 Control .27* .01 K.08 .09 K.02 K.27* K.04 K.12 K.38**
11 Y Pain .07 .05 K.04 .19 .02 K.31* K.11 K.11 K.23
12 Post-catast K.30* .14 .09 K.08 .19 .20 .02 .10 .19
13 Post-contr .24 K.17 K.29* K.04 K.06 K.19 .11 K.22 K.28*
14 Post Y pain .20 K.14 K.09 K.09 K.04 .04 .08 K.18 K.07
15 Threshold .08 K.01 K.20 .02 K.05 K.34** K.18 K.19 K.16
16 Tolerance .09 K.06 K.07 K.05 .01 K.13 K.02 K.14 .04
17 VAS one .02 K.32* K.18 .07 .05 K.01 .33* .11 K.17
minute
18 VAS .002 .05 .17 K.12 K.04 .12 K.01 .001 K.04
19 Ave VAS K.001 K.19 .07 K.04 K.06 .11 .19 .07 K.07
20 Min pain .07 K.12 .16 K.02 K.05 .09 .18 .15 K.11
21 Max pain K.11 K.10 .006 K.05 .08 .21 .20 .14 .05
22 Age K.30* K.12 K.15 .19 .30* K.36** .02 .07 K.13
interacted with age in predicting catastrophizing (BZK.48, reporting preoccupied attachment were less able to decrease
tZK2.3, PZ.03) relative to dismissing attachment. Thus, pain relative to older, dismissing individuals.
those who were both younger and fearfully attached were
more inclined to catastrophize relative to older, dismissing 3.5. Continuous attachment variables as predictors
individuals. of pain variables
In addition, several new significant findings emerged
with the inclusion of age as a covariate. First, the categorical In order to explore in more depth the associations
RQ variable approached significance (F(3, 54)Z2.61, PZ between attachment and pain variables, further attention
.06) as a predictor of overall pain intensity; inspection of was given to the two instances in which attachment was
parameter estimates revealed that secure attachment significantly correlated with pain variables: ASQ anxiety
contributed significantly to overall pain intensity (BZ with pain threshold, and RQ fearful with pain intensity.
6.49, tZ2.57, PZ.01) compared with the dismissing Correlated variables were used as predictors of pain
category. The RQ variable and age also interacted to predict threshold and intensity; however, since the notion of
overall pain intensity (F(3, 54)Z2.81, PZ.05), with the prediction presumes that the independent variable precedes
interaction between secure attachment and age (BZK.10, the dependent variable, only measures taken prior to the
tZK2.74, PZ.01) emerging as significantly more strongly coldpressor task were considered as predictors. In the first
associated with pain intensity compared to dismissing case, only ASQ anxiety and pre-control were correlated
attachment. Thus, people who were both older and securely with pain threshold. These two variables were entered as
attached reported less pain relative to younger, dismissing independent variables in a regression analysis with pain
individuals. threshold as the dependent variable, revealing a significant
Second, categorical RQ predicted post-ability to decrease Table 4
pain (F(3, 52)Z2.75, PZ.05), with preoccupied attachment Summary of means of attachment categories for significant ANOVAs
revealed as more strongly associated with ability to decrease
Variable Secure M Fearful M Preoccupied Dismissing
pain (BZK15.50, tZK2.12, PZ.04) compared with M M
dismissing attachment. RQ also interacted with age in
N 29 7 3 16
predicting this variable (F(3, 52)Z3.44, PZ.03), with the Pre-cata- 4.7* 10.6* 12.3 6.4
interaction between the preoccupied category and age (BZ strophizing
.61, tZ2.12, PZ.04) emerging as significantly more Age 33.4 34.7 34.3 41.7
strongly related with post-ability to decrease pain compared Depression 2.4* 7.1 7.3 8.1*
with dismissing attachment. Thus, younger participants *Denotes significant difference P!.05 using Bonferroni Post Hoc analyses.
P.J. Meredith et al. / Pain 120 (2006) 4452 49
overall result (F(2, 52)Z5.4, PZ.007). Attachment anxiety Following from these results, and because it is
contributed uniquely to this relationship (tZK2.12, PZ theoretically plausible that secure attachment may moderate
.04), although control almost reached significance (PZ.07). the effect of pain on catastrophizing, this hypothesis was
Including age as a covariate in this equation slightly explored. Indeed, a regression analysis (incorporating an
decreased the overall significance (F(3, 53)Z4.24, PZ interaction term) revealed a significant interaction between
.009). secure attachment and overall pain intensity in predicting
In the second case, RQ fearful and stress were entered as post-catastrophizing (F(1, 54)Z11.55, PZ.001). These
independent variables to predict pain intensity ratings at three variables were standardized and, using ModGraph
1 min. A significant overall result was obtained (F(2, 43)Z (Jose, 2005), a continuous moderator chart was developed
7.71, PZ.001), with both variables contributing signifi- using standardized variables (see Fig. 1). This chart revealed
cantly to the relationship (tZK2.73, PZ.009 and tZ3.07, that those high in secure attachment consistently catastro-
PZ.004 respectively). Including age as a covariate again phized less than did those lower in secure attachment.
decreased the significance of this equation (F(3, 42)Z5.03, Further, for participants who reported low levels of pain in
PZ.005). response to the coldpressor task, levels of catastrophizing
were similar for all levels of secure attachment. As pain
3.6. Post-catastrophizing increased, however, attachment security gained importance,
with those lowest in security reporting highest catastrophiz-
Because of the different correlation pattern between pre- ing scores, while those high in security reported low
and post-task catastrophizing, and because post-catastro- catastrophizing even for high levels of pain.
phizing was so strongly associated with all pain variables, it
was explored further. With all correlated variables (secure
attachment, tolerance, threshold, post-control, post-ability 4. Discussion
to decrease pain, minimum and maximum pain intensity,
and average and overall pain intensity scores) entered as These findings offer preliminary support for the role of
independent variables and post-catastrophizing as the adult attachment variables in the perception of, and response
dependent variable, a significant result was obtained (F(9, to, experimentally induced pain. Secure participants were
44)Z6.9, P!.001), with secure attachment and overall pain younger and more stressed; however, they reported less
intensity the only contributing variables (tZK2.97, PZ depressive symptoms and less catastrophizing both before
.005, and tZ3.17, PZ.003 respectively). Due to the very and after the coldpressor task. They also perceived more
small degrees of freedom in this analysis, it was repeated control over their pain. These factors are considered
using only the two significant variables (secure attachment adaptive or protective for coping with pain. Fearful
and pain intensity) as independent variables. These two individuals reported more pre-task catastrophizing and,
variables together explained 49% of the variance in post- together with stress, fearful attachment contributed signifi-
catastrophizing (F(2, 52)Z27.06, P!.001). Inclusion of cantly to predictions of pain intensity at one minute.
neither age nor pre-catastrophizing as covariates enhanced Notably, although fearful attachment was hypothesised to
these results, nor emerged as significant predictors. contribute to increased pain sensitivity, this sample
indicated less pain intensity at one minute of the task.
security moderates pain and catastrophizing This finding should be considered with caution, however,
since fearful attachment was not related to any other
security
1.2 intensity variables. This may reflect a chance occurrence, or
High a social desirability bias whereby fearful participants strive
1.0 Medium
Low to appear to be coping well, stoically maintaining the
0.8
post-track catastrophizing
interesting significant interactions between attachment and As noted, pre-task catastrophizing was positively
age emerged. Firstly, those who were both younger and associated with all three anxious attachment variables
fearfully attached were more inclined to catastrophize (fearful, preoccupied, anxious), and negatively with secure
relative to older, dismissing individuals. Secondly, people attachment. This result is consistent with previous research,
who were both older and securely attached reported less and is in accord with predictions. Although pre-task
pain relative to younger, dismissing individuals. Finally, catastrophizing was associated with increased depression
younger participants reporting preoccupied attachment were and perceived inability to control pain, it was not associated
less able to decrease pain relative to older, dismissing with any of the pain intensity variables. This finding is
individuals. Thus, there was a tendency for youth and contrary to initial expectations, and contradicts existing
insecure attachment to be associated with less adaptive evidence that high levels of catastrophizing predict
variables. increased pain intensity (Severeijns et al., 2005). One
Using ASQ attachment dimensions, anxiety over explanation is that measurement of pain catastrophizing
relationships was more strongly linked with pain variables. associated with hypothetical pain (pre-task) may draw more
Anxious individuals reported increased catastrophizing, on tendencies that might more accurately reflect attachment
stress and depression, and perceived less control of, and characteristics, compared with measures of catastrophizing
ability to decrease, pain. Those scoring higher on this associated with real pain. This would also explain the low
dimension also reported experiencing pain earlier in the task correlation between pre- and post-catastrophizing. Further
than those indicating lower attachment anxiety. Indeed, support for this suggestion was provided by the links
attachment anxiety emerged as the sole unique predictor of between post-task catastrophizing (which particularly
pain threshold, although this relationship diminished when referenced the coldpressor pain) and all pain intensity
controlling for age. Thus, attachment anxiety was associated variables.
with maladaptive coping cognitions, emotional distress, and It was revealed that attachment security moderated the
lowered pain thresholds, all of which are associated with relationship between pain intensity and post-catastrophiz-
problematic adjustment to acute pain. ing, such that the impact of attachment security on
Variations in findings between the two attachment catastrophizing was most pronounced for participants who
measures employed in this study, and the relatively low reported high levels of pain in response to the coldpressor
(but conceptually congruent) correlations between them, task. Although this study does not address the direction of
warrant comment. Measures of attachment differ funda- the relationship between pain intensity and pain catastro-
mentally in their approach to attachment issues (Feeney phizing, Dixon et al. (2004) suggested that this relationship
et al., 1994), with ongoing debate regarding the clinical and may be circular (i.e. catastrophic thoughts influence
empirical value of attachment dimensions versus categories perceptions of pain severity which may, in turn, influence
(Fraley & Spieker, 2003a,b; Fraley & Waller, 1998). catastrophic thinking). Findings of the present study, then,
Although secure attachment is theoretically comprised of are particularly interesting as they suggest that those
low discomfort with closeness and low anxiety over reporting insecure attachment tendencies may be at more
relationships, evidence suggests that attachment patterns risk of escalating the catastrophizing/pain intensity spiral.
are largely continuous rather than categorical. While such With catastrophizing known to contribute to chronicity
conceptual and measurement differences presuppose vari- (Severeijns et al., 2005), this finding warrants further
ations in findings, one area of consistency between measures empirical attention.
is noteworthy: catastrophizing was linked with anxiety over Several forms of bias may have been introduced to the
relationships and each of the anxious attachment styles study. Firstly, selection bias may have been introduced as a
(fearful and preoccupied). result of using a convenience sample. Use of students,
Previous research using the coldpressor task has revealed friends and colleagues as participants posed potential threats
associations between depressed mood and both decreased to validity and generalizeability of findings, and confidence
pain tolerance (Gormsen et al., 2004; Zelman et al., 1991), in the relevance of results awaits replication with a more
and increased pain threshold and tolerance (Geisser et al., diverse sample. Nevertheless, convenience samples of
2003). The finding that mood was largely unrelated to the students are typically represented in acute pain research
pain variables in the present study, may reflect a ceiling (e.g. Dixon et al., 2004), and inclusion of friends and
effect due to the relatively low levels of stress and colleagues along with students in the present study may
depression in this healthy sample. The links observed actually have added diversity to this mix.
between mood and attachment warrant further consider- Since the researcher conducting the coldpressor task was
ation, however. Associations between depression and not blind to the purposes of the study, researcher bias may
insecure attachment raise the possibility that, when faced have been introduced. Efforts to minimize this bias were
with chronic, unremitting pain, negative emotional ten- made by not viewing or scoring questionnaires until after
dencies may be activated more strongly in those with the task; hence, researchers were blind to attachment status
insecure attachment patterns, thus increasing the vulner- and catastrophizing scores of participants. In addition, due
ability of these individuals. mainly to the size of the literature and complexity of
P.J. Meredith et al. / Pain 120 (2006) 4452 51
Geisser ME, Casey KL, Brucksch CB, Ribbens CM, Appleton BB, Beck depression and anxiety inventories. Behav Res Ther 1995;33(3):
Crofford LJ. Perception of noxious and innocuous heat stimulation 33553.
among healthy women and women with fibromyalgia: association with Meredith PJ, Strong J, Feeney JA. Evidence of a relationship between adult
mood, somatic focus, and catastrophizing. Pain 2003;102:24350. attachment variables and appraisals of chronic pain. Pain Res Manag in
Gormsen L, Ribe AR, Raun P, Rosenberg R, Videbech P, Vestergaard P, press.
Bach FW, Jensen TS. Pain thresholds during and after treatment for Mikail SF, Henderson PR, Tasca GA. An interpersonally based model of
severe depression with electroconvulsive therapy. Eur J Pain 2004;8: chronic pain: an application of attachment theory. Clin Psychol Rev
48793. 1994;14:116.
Hazan C, Shaver PR. Romantic love conceptualized as an attachment Mikulincer M. Attachment style and the mental representation of the self.
process. J Pers Soc Psychol 1987;52:51124. J Pers Soc Psychol 1995;69:120315.
Herzberg DS, Hammen C, Burge D, Daley SE, Davila J, Lindberg N. Pearce S, Creed P, Cramond T. Attachment style and chronic pain
Attachment cognitions predict perceived and enacted social support syndrome. Aust Pain Soc Newsl 2001;May:24.
during late adolescence. J Adolesc Res 1999;14(4):387404. Rosenstiel AK, Keefe FJ. The use of coping strategies in chronic low back
Hunter JJ, Maunder RG. Using attachment theory to understand illness
pain patients: relationship to patient characteristics and current
behaviour. Gen Hosp Psychiatry 2002;23:17782.
adjustment. Pain 1983;17(1):3344.
Jose PE. Moderation/Mediation Help Centre, Version 1.0. School of
Sanford K. Two dimensions of adult attachment: further validation. J Soc
Psychology, Victoria University of Wellington; June 2004. Available
Pers Relat 1997;14(1):13343.
at: http://www.vuw.ac.nz/psyc/staff/pauljose/files/helpcentre/help1_in-
Scharfe E, Bartholomew K. Reliability and stability of adult attachment
tro.php [Accessed 21 May, 2005].
patterns. Pers Relationships 1994;1:2343.
Kallai I, Barke A, Voss U. The effects of experimenter characteristics on
pain reports in women and men. Pain 2004;112:1427. Schmidt S, Strauss B, Braehler E. Subjective physical complaints and
Kolb LC. Attachment behavior and pain complaints. Psychosomatics 1982; hypochondriacal features from an attachment theoretical perspective.
23:41325. Psychol Psychother 2002;75:31332.
Leak GK, Parsons CJ. The susceptibility of three attachment style measures Severeijns R, van den Hout MA, Vlaeyen YWS. The causal status of pain
to socially desirable responding. Soc Behav Personal 2001;29(1): catastrophizing: an experimental test with healthy participants. Eur
2130. J Pain 2005;9:25765.
Lovibond PF, Lovibond SH. Manual for the Depression Anxiety Stress Swartzman LC, Gwadry FG, Shapiro AP, Teasell RW. The factor structure
Scales (DASS). Psychology foundation monograph; 1993 [Available of the Coping Strategies Questionnaire. Pain 1994;57(3):3116.
from The Psychology Foundation, Room 1005 Mathews Building, Young CW. Transition from acute to chronic pain: a biopsychosocial model
University of New South Wales, 2052, Australia]. of chronicity. Dissert Abstr Int B Sci Eng 2003;63(9-B):4392.
Lovibond PF, Lovibond SH. The structure of negative emotional states: Zelman DC, Howland EW, Nichols SN, Cleeland CS. The effects of
comparison of the Depression Anxiety Stress Scales (DASS) with the induced mood on laboratory pain. Pain 1991;46:10511.