“being” in pain: the role of self-discrepancies in the emotional experience and activity...

7
‘‘Being’’ in pain: The role of self-discrepancies in the emotional experience and activity patterns of patients with chronic low back pain Hanne P.J. Kindermans a,, Ivan P.J. Huijnen b , Marielle E.J.B. Goossens a , Jeffrey Roelofs a , Jeanine A. Verbunt b , Johan W.S. Vlaeyen a,c a Department of Clinical Psychological Science, Maastricht University, Maastricht, The Netherlands b Department of Rehabilitation Medicine, Maastricht University, Maastricht, The Netherlands c Research Center for Health Psychology, University of Leuven, Leuven, Belgium article info Article history: Received 28 January 2010 Received in revised form 15 October 2010 Accepted 9 November 2010 Keywords: Self-discrepancies Activity patterns Chronic low back pain abstract Chronic pain not only interferes with daily activities, it may also have a negative impact on the perceived integrity of one’s self through self-discrepancies. Self-discrepancies are experienced distances between the actual self and self-guides that can exist from 2 perspectives (ie, own and other). Self-discrepancies are associated with negative mood states and incite self-regulatory behavior in order to reduce these dis- crepancies. The present study was aimed at replicating the emotional consequences of self-discrepancies in patients with chronic low back pain, and extending current knowledge of the behavioral consequences of self-discrepancies (ie, behavioral activity patterns such as avoidance and persistence). A cross-sec- tional design was employed with 83 patients who completed a number of self-report measures. We hypothesized that ideal and ought discrepancies, as well as feared congruencies were associated with depressed and anxious mood. On the behavioral level, a U-shaped relationship was hypothesized between ideal and ought self-discrepancies and persistence behavior, whereas feared self-discrepancies were hypothesized to be related to avoidance behavior. Results were partially in line with the hypothe- ses. With respect to the emotional consequences, feared (own and other) self-discrepancies were predic- tive of depressive and anxious mood. With regard to activity patterns, results showed a U-shaped relationship between ideal-other self-discrepancies and persistence behavior and a positive relationship between feared-own self-discrepancies and avoidance behavior. In contrast to expectations, none of the other self-discrepancies was related to activity patterns. Of interest was that avoidance, but not persis- tence behavior, was predictive of higher levels of disability and lower levels of quality of life. Ó 2010 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. 1. Introduction Pain can interrupt ongoing cognitive and physical tasks, and interferes with daily activities [9]. This is especially the case when pain is interpreted catastrophically and when pain-related fear has developed [8,19]. When pain persists, such as in chronic musculo- skeletal pain, the continued interference and associated disability poses a real challenge to patients, which is reflected in low levels of quality of life and negative perception of one’s own identity [18,22]. In an attempt to restore daily functioning, patients can alter their performance. This anticipated change in their activity pattern is, however, often unsuccessful and can even be counterproductive. For example, a body of research has shown that pain catastrophiz- ing and pain-related fear might lead to unnecessary avoidance of activities and, consequently, to disuse and disability in the long term [33]. Besides activity patterns characterized by avoidance, researchers and clinicians have suggested that there is also a group of chronic pain patients with disability complaints who tend to per- sist in performing activities despite their pain [15,31,34]. So far, there is a lack of empirical evidence explaining why some chronic pain patients persist in activities while others avoid activities. A possible mechanism explaining both avoidance and persis- tence behavior might be derived from the Self-Discrepancy Theory (SDT; [16]). SDT explains how self-discrepancies, that is, discrepan- cies between the actual self (ie, the person you are now) and the ideal self (ie, the person you ideally would like to be), the ought self (ie, the person you feel you ought to be), or the feared self (ie, the person you fear to be) lead to negative emotions and distress [6,16]. Studies in chronic pain populations have shown that self- discrepancies are associated with higher levels of depression, anx- iety, distress, and pain [11,35]. Besides emotional consequences, self-discrepancies are also thought to be associated with 0304-3959/$36.00 Ó 2010 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.pain.2010.11.009 Corresponding author. Address: Department of Clinical Psychological Science, Maastricht University, P.O. Box 616, 6200 MD, Maastricht, The Netherlands. Tel.: +31 43 387 2487; fax: +31 43 388 4155. E-mail address: [email protected] (H.P.J. Kindermans). PAIN Ò 152 (2011) 403–409 www.elsevier.com/locate/pain

Upload: hanne-pj-kindermans

Post on 11-Sep-2016

215 views

Category:

Documents


3 download

TRANSCRIPT

PAIN�

152 (2011) 403–409

w w w . e l s e v i e r . c o m / l o c a t e / p a i n

‘‘Being’’ in pain: The role of self-discrepancies in the emotional experienceand activity patterns of patients with chronic low back pain

Hanne P.J. Kindermans a,⇑, Ivan P.J. Huijnen b, Marielle E.J.B. Goossens a, Jeffrey Roelofs a,Jeanine A. Verbunt b, Johan W.S. Vlaeyen a,c

a Department of Clinical Psychological Science, Maastricht University, Maastricht, The Netherlandsb Department of Rehabilitation Medicine, Maastricht University, Maastricht, The Netherlandsc Research Center for Health Psychology, University of Leuven, Leuven, Belgium

a r t i c l e i n f o a b s t r a c t

Article history:Received 28 January 2010Received in revised form 15 October 2010Accepted 9 November 2010

Keywords:Self-discrepanciesActivity patternsChronic low back pain

0304-3959/$36.00 � 2010 International Associationdoi:10.1016/j.pain.2010.11.009

⇑ Corresponding author. Address: Department of CMaastricht University, P.O. Box 616, 6200 MD, Maas+31 43 387 2487; fax: +31 43 388 4155.

E-mail address: h.kindermans@maastrichtuniversi

Chronic pain not only interferes with daily activities, it may also have a negative impact on the perceivedintegrity of one’s self through self-discrepancies. Self-discrepancies are experienced distances betweenthe actual self and self-guides that can exist from 2 perspectives (ie, own and other). Self-discrepanciesare associated with negative mood states and incite self-regulatory behavior in order to reduce these dis-crepancies. The present study was aimed at replicating the emotional consequences of self-discrepanciesin patients with chronic low back pain, and extending current knowledge of the behavioral consequencesof self-discrepancies (ie, behavioral activity patterns such as avoidance and persistence). A cross-sec-tional design was employed with 83 patients who completed a number of self-report measures. Wehypothesized that ideal and ought discrepancies, as well as feared congruencies were associated withdepressed and anxious mood. On the behavioral level, a U-shaped relationship was hypothesizedbetween ideal and ought self-discrepancies and persistence behavior, whereas feared self-discrepancieswere hypothesized to be related to avoidance behavior. Results were partially in line with the hypothe-ses. With respect to the emotional consequences, feared (own and other) self-discrepancies were predic-tive of depressive and anxious mood. With regard to activity patterns, results showed a U-shapedrelationship between ideal-other self-discrepancies and persistence behavior and a positive relationshipbetween feared-own self-discrepancies and avoidance behavior. In contrast to expectations, none of theother self-discrepancies was related to activity patterns. Of interest was that avoidance, but not persis-tence behavior, was predictive of higher levels of disability and lower levels of quality of life.

� 2010 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.

1. Introduction

Pain can interrupt ongoing cognitive and physical tasks, andinterferes with daily activities [9]. This is especially the case whenpain is interpreted catastrophically and when pain-related fear hasdeveloped [8,19]. When pain persists, such as in chronic musculo-skeletal pain, the continued interference and associated disabilityposes a real challenge to patients, which is reflected in low levelsof quality of life and negative perception of one’s own identity[18,22]. In an attempt to restore daily functioning, patients can altertheir performance. This anticipated change in their activity patternis, however, often unsuccessful and can even be counterproductive.For example, a body of research has shown that pain catastrophiz-

for the Study of Pain. Published by

linical Psychological Science,tricht, The Netherlands. Tel.:

ty.nl (H.P.J. Kindermans).

ing and pain-related fear might lead to unnecessary avoidance ofactivities and, consequently, to disuse and disability in the longterm [33]. Besides activity patterns characterized by avoidance,researchers and clinicians have suggested that there is also a groupof chronic pain patients with disability complaints who tend to per-sist in performing activities despite their pain [15,31,34]. So far,there is a lack of empirical evidence explaining why some chronicpain patients persist in activities while others avoid activities.

A possible mechanism explaining both avoidance and persis-tence behavior might be derived from the Self-Discrepancy Theory(SDT; [16]). SDT explains how self-discrepancies, that is, discrepan-cies between the actual self (ie, the person you are now) and theideal self (ie, the person you ideally would like to be), the ought self(ie, the person you feel you ought to be), or the feared self (ie, theperson you fear to be) lead to negative emotions and distress[6,16]. Studies in chronic pain populations have shown that self-discrepancies are associated with higher levels of depression, anx-iety, distress, and pain [11,35]. Besides emotional consequences,self-discrepancies are also thought to be associated with

Elsevier B.V. All rights reserved.

404 H.P.J. Kindermans et al. / PAIN�

152 (2011) 403–409

behavioral consequences. That is, self-discrepancies motivate self-regulatory behaviors in order to reduce discrepancies (or enlargediscrepancies, in the case of the feared self) [3,7].

The first aim of the present study was to replicate earlier find-ings and investigate the emotional consequences of self-discrepan-cies in patients with chronic pain. In line with previous studies[11,35], we hypothesize that ideal and ought discrepancies andfeared congruencies predict depressive and anxious mood. The sec-ond aim was to extend the study of self-discrepancies in chronicpain patients by examining their explanatory function in activitypatterns. We hypothesize that ideal or ought self-discrepanciesare associated with persistence behavior because their behavioris driven by a focus on desired or obliged end states (ie, ideal orought self). Yet, persistence behavior might also be aimed at main-taining an actual-ideal or actual-ought congruency. Therefore, wetested the existence of a U-shaped relationship between idealand ought self-discrepancies and persistence behavior. Fearedself-congruencies are hypothesized to predict avoidance behaviorbecause the focus is on an undesired end state (ie, the feared self).Finally, we hypothesize that both avoidance and persistence areassociated with higher levels of disability and inferior quality oflife.

2. Methods

2.1. Recruitment of participants

The present study took place in the Southern area of the Nether-lands. Participants were either referred to the study by their consul-tant in rehabilitation medicine or responded to the advertisementthat appeared in local newspapers. Consultants in rehabilitationmedicine were asked to refer patients meeting the inclusion andexclusion criteria for the present study. In cases of patientsresponding to the advertisement, selection criteria were checkedby a consultant in rehabilitation medicine who performed a medi-cal screening according to the clinical guidelines for low back painof the Dutch College of General Practitioners [10]. Participants wereeligible for inclusion in the present study if they suffered from non-specific low back pain for at least 3 months, were aged between 18and 65 years, and had sufficient knowledge of the Dutch language.Participants for whom primary psychiatric treatment was indicatedwere excluded. For female participants, pregnancy was also anexclusion criterion.

2.2. Measures

Besides sociodemographics (ie, gender, age, education level,occupation) and pain-related information (ie, pain duration), data

Table 1Descriptive statistics (Cronbach’s alpha, mean, SD) and Pearson correlations for independe

a M SD 1 2 3 4

1 Ideal-own (HSQ) .60 3.26 .822 Ideal-other (HSQ) .74 3.23 .93 .54**

3 Ought-own (HSQ) .75 3.63 .90 .40** .35**

4 Ought-other (HSQ) .70 3.37 .90 .37** .65** .49**

5 Feared-own (HSQ) .75 2.20 .93 �.20 �.14 �.24* �6 Feared-other (HSQ) .65 2.38 .85 �.04 �.25* �.35** �7 Depression (BDI-II) .93 13.63 10.77 �.25* �.15 �.28** �8 Anxiety (HADS) .86 6.65 4.41 �.25* �.12 �.23* �9 Avoidance (POAM-P) .88 19.34 8.20 �.24* �.09 �.10 .010 Overdoing (POAM-P) .75 23.42 6.31 .02 .15 .08 .111 Disability (RDQ) .81 11.73 4.46 �.13 �.11 �.06 �12 QOL-physical (RAND-36) .76 197.97 74.74 .16 .04 .12 �13 QOL-mental (RAND-36) .84 248.35 84.47 .16 .03 .12 .0

* P 6 .05.** P 6 .01.

were obtained for the measurement instruments to follow. Reli-ability coefficients (Cronbach alpha) of all questionnaires are pre-sented in Table 1.

2.2.1. Self-discrepanciesSelf-discrepancies were measured by assessing Hardin’s Selves

Questionnaire (HSQ; [12]) in the form of a semi-structured inter-view. Participants were asked to generate attributes for each self-guide (ie, ideal, ought, and feared) from 2 perspectives, their ownperspective and the internalized perspective of someone who isimportant to them (eg, spouse). (Originally, the HSQ assessed theundesired self instead of the feared self. For the purpose of ourhypotheses, we revised the description of the undesired self and,in concordance with Carver et al. [6], modified it to the feared self).More specifically, participants listed 5 attributes they ideallywould like to possess (Ideal-own), 5 attributes they feel the impor-tant other would ideally like them to possess (Ideal-other), 5 attri-butes they feel they ought to possess (Ought-own), 5 attributesthey think the other feels they ought to possess (Ought-other), 5attributes they fear to possess (Feared-own), and 5 attributes theother fears they possess (Feared-other). After listing all 30 attri-butes, participants were asked to rate to which extent each attri-bute currently describes themselves on a scale from 1 (does notdescribe me at all) to 5 (completely describes me). Self-discrepancyscores were obtained by summing the discrepancy scores of the 5attributes. For the ideal and the ought self, lower scores indicatedlarger discrepancies, reflecting a negative state, while for thefeared self, lower scores indicated smaller congruencies and there-fore reflect a positive state. The HSQ is a valid and reliable instru-ment to measure self-discrepancies [12]. To permit comparisonswith previous studies in which the HSQ was used [11,12], thescores presented in Table 1 are the scores per self-guide dividedby 5 (ie, the number of attributes).

2.2.2. DepressionThe Dutch version of the Beck Depression Inventory II (BDI-II;

[2]) was utilized to measure depression. Depressive symptomatol-ogy is questioned in 21 items with several possible responses andscores ranging from 0 to 3. Total scores are obtained by summingthe item scores, resulting in total scores ranging from 0 (not de-pressed at all) to 63 (severely depressed). The BDI-II has sound psy-chometric properties [1] and has been found suitable foradministration in chronic pain populations [13].

2.2.3. AnxietyThe anxiety subscale of the Hospital Anxiety and Depression

Scale [36] was used to measure anxiety. The anxiety subscale existsout of 7 items about the frequency of experiencing anxious states

nt and dependent variables.

5 6 7 8 9 10 11 12

.12

.30** .64**

.13 .40** .34**

.17 .38** .35** .75**

2 .22* .02 .34** .29**

4 .23* .21 .35** .33** �.05.00 .20 .15 .51** .25* .47** .19.08 �.20 �.05 �.58** �.45** �.42** �.24* �.73**

7 �.28** �.29** �.74** �.67** �.42** �.22* �.52** .68**

H.P.J. Kindermans et al. / PAIN�

152 (2011) 403–409 405

on a 4-point scale. The Dutch version of the Hospital Anxiety andDepression Scale has been found to be a reliable and valid instru-ment [27].

2.2.4. Activity patternsTo measure avoidance and persistence behavior, a recently

developed instrument, the Patterns of Activity Measure-Pain(POAM-P; [5]) was administered. The POAM-P is a 30-item self-re-port measure developed to identify 3 activity patterns in patientswith chronic pain: avoidance (10 items), overdoing (ie, persistencebehavior; 10 items), and pacing (10 items). Yet, because the focusof the present study is on avoidance and persistence, no resultsregarding the pacing scale will be reported. An example of anavoidance item is ‘‘I avoid activities that I know will make my painworse,’’ while a contrasting item of the overdoing scale is ‘‘WhenI’m doing an activity, I don’t stop until it is finished.’’ Participantshave to indicate to which extent the item applies to them on a 5-point scale ranging from 0 (not at all) to 4 (always). Total scoresper subscale range from 0 to 40. The 3-factor structure of the Dutchversion of the POAM-P has been confirmed and the 3 subscaleswere found to have high internal consistency (Cronbach alphasranging from .80 to .94). Significant and meaningful associationswith related constructs such as fear of movement and (re)injury,pain catastrophizing, depression, and disability confirmed thevalidity of the subscales (Kindermans HPJ, Roelofs J, Goossens M,Huijnen IP, Verbunt JA, Vlaeyen J., unpublished manuscript).

2.2.5. DisabilityDisability was measured with the Roland Disability Question-

naire (RDQ; [26]). The RDQ presents 24 daily activities; partici-pants have to indicate whether their low back pain interfereswith the activity (score 1) or not (score 0). The item scores aresummed, leaving total score ranges between 0 and 24, with higherscores reflecting higher levels of disability. The Dutch version ofthe RDQ is a reliable and valid instrument assessing disability inpatients with chronic low back pain [4].

2.2.6. Health-related quality of lifeThe RAND 36-item Health survey (RAND-36, [29]) was adminis-

tered to measure health-related quality of life. The questionnairecomprises 9 subscales measuring both physical and mental health,therefore enabling the computation of a physical health (physicalfunctioning, physical role limitations, general health perception,and pain) and a mental health (social functioning, emotional rolelimitations, mental health, and vitality) scale, which will be usedin the present study. The ninth, 1-item subscale, ‘‘health change,’’does not belong to one of the aforementioned scales and will notbe reported on. Responses to the items are given on a 3- or 6-pointLikert scale, or with ‘‘yes’’ or ‘‘no,’’ depending on the subscale. Psy-chometric properties of the RAND-36 are satisfactory [29,30].

2.2.7. Pain intensityPain intensity was measured on 3 100-mm visual analogue

scales [25] on which participants had to indicate their ‘‘worst painlast week,’’ ‘‘least intense pain last week,’’ and ‘‘current pain inten-sity.’’ The average of the scores on these 3 scales was used as acomposite score of pain intensity.

2.3. Procedure and design

Patients who were interested in participation received a partic-ipant information letter, and if they responded that they were will-ing to participate, they were then contacted by the researchers.Participants who responded to the advertisement also received aparticipant information letter and, if willing to participate, theywere screened by a consultant in rehabilitation medicine to ensure

qualification for the present study. After obtaining written in-formed consent, participants were invited to come to the Universityto have the HSQ interview or were interviewed by telephone.Thereafter, participants completed the self-report questionnaires,either on paper or an electronic version presented via Internet.The present study was part of a larger study, which also includeddaily life assessments and follow-up measurements that will be re-ported elsewhere. Participants received a gift voucher for their par-ticipation and were reimbursed for their travel expenses.

The present study was approved by the Medical Ethics Commit-tee of the Maastricht University Medical Centre (MEC 06-3-046)and by the institutions of the rehabilitation units from which pa-tients were referred to the study.

2.4. Statistical analysis

Data preparation involved missing data analyses and checks fornormality of the variables, for which the Statistical Package for So-cial Sciences (SPSS 15.0; SSPS Inc., Chicago, IL) was used. In cases ofless then 10% missing values per (sub-) scale, the missing valuewas estimated by means of regression techniques. Scales withmore then 10% missing values were dropped from analyses. Twoscales, namely the BDI-II and pain duration, did not meet assump-tions of normality and were square-root transformed prior to anal-yses. After transformation, skewness and kurtosis values of bothscales were in the acceptable range of �1 and +1.

Computations of descriptive information, chi-squared andt-tests, and correlational analyses were carried out in SPSS 15.0.Besides correlational analyses, hierarchical multiple regressionanalyses were carried out in SPSS to determine the contribution ofself-discrepancies to the prediction of negative mood (depressionand anxiety) and activity patterns (avoidance and persistence) andthe contribution of activity patterns to disability and quality of life.To prevent type I errors due to multiple testing, alpha was set to .01.

3. Results

3.1. Participants

3.1.1. Participants flowA total of 116 participants with chronic nonspecific low back

pain participated in the present study. Due to missing data, 33participants (28%) were not included in the analyses. Of these 33participants, one participant did not complete the HSQ interviewand 17 did not complete the self-report questionnaires. Theremaining 15 participants did complete the HSQ interview andthe self-report questionnaires but still had too many missing data(ie, more than 10% on the level of individual items) after runningmissing values analyses. A lack of descriptive information of the17 participants who did not fill in the questionnaires precludedchecking for selective drop-out. The reasons for withdrawal werein most cases unknown, yet some indicated a lack of time due topersonal circumstances or experienced the study as too burden-some. In total, 83 participants were included in the analyses, ofwhich 58 (70%) participants were referred to the study by theirconsultant in rehabilitation medicine and 25 (30%) responded tothe advertisements in local newspapers. Participants who re-sponded to the advertisement were older (M = 52.20 years,SE = 1.94) than those referred by their consultant in rehabilitation(M = 44.41 years, SE = 1.31; t (81) = �3.29, P = 0.001). There wereno significant differences between both samples on the other inde-pendent and dependent variables.

3.1.2. Descriptive informationForty-two women and 41 men were included in the analyses.

There were no gender differences on the self-report questionnaires.

406 H.P.J. Kindermans et al. / PAIN�

152 (2011) 403–409

Age ranged from 22 to 65 years, with a mean of 46.76 years(SD = 10.46). With respect to education level, the majority of par-ticipants received lower secondary professional education or be-low (53%). The rest of the participants received, at maximum,higher general secondary education (31%) or higher education(13%). Fifty-three percent were employed (full- or part-time), 6%were on sick leave, and 21% were recipients of disablement insur-ance benefits. The average duration of the pain complaints was9.81 years (SD = 8.93) and ranged from 5 months to 39 years. Meanpain intensity, measured on 100-mm visual analogue scales, was49.47 (SD = 18.73). Means and standard deviations of the scoreson the self-report questionnaires are presented in Table 1.

3.2. Self-discrepancies and emotional consequences

Prior to the hierarchical multiple regression analyses, correla-tional analyses were carried out to investigate the relationships be-tween the distinct types of self-discrepancies and depression andanxiety (Table 1). Results showed that ought-own and feared-own and feared-other discrepancies were significantly related todepression and feared-own and feared-other to anxiety in the ex-pected direction. Pearson correlations ranged from r = �0.28 tor = 0.40 at P 6 0.01 (Table 1). Thus, the closer participants felt totheir ought-own, the less depressive mood they experienced.Regarding feared selves, both feared-own and feared-other congru-encies were related to depression and anxiety. This means that thecloser participants felt to the person they feared to be, or to theperson that another feared them to be, the more feelings of depres-sion and anxiety the participant experienced. Feeling close to theideal (own or other) or ought-other self was not significantly re-lated to depressive or anxious mood.

Hierarchical multiple regression analyses were performedpredicting depression and anxiety while controlling for sociode-mographic variables (gender and age) in the first block and pain-related variables (pain duration and mean pain intensity) in thesecond block. Each type of self-discrepancy from both perspectives(ideal, ought, and feared) was analyzed separately and was enteredin the third block. Ideal-own and ideal-other self-discrepancieswere not predictive of depression (respectively, b = �0.25,t = �2.08, P = 0.04, and b = �0.06, t = �0.47, P = 0.64) nor anxiety(respectively, b = �0.27, t = �2.11, P = 0.04, and b = 0.00, t = 0.01,P = 0.99). Neither were ought-own or ought-other self-discrepan-cies predictive of depression (respectively, b = �0.25, t = �2.07,

Table 2Summary of hierarchical regression analyses for feared self-discrepancies predicting depre

Dependent Blocks Variables B

Depression (BDI) 1 Gender �0.17Age �0.02

2 Pain duration 0.06Mean pain intensity 0.42

3a Feared-own 0.25Feared-other 0.20

3b Feared-own 0.383c Feared-other

0.35Anxiety (HADS) 1 Gender �0.05

Age �0.08

2 Pain duration �0.01Mean pain intensity 0.27

3a Feared-own 0.25Feared-other 0.22

3b Feared-own 0.393c Feared-other 0.37

BDI, Beck Depression Inventory; HADS, Hospital Anxiety and Depression Scale.* P 6 0.001.

P = 0.04, and b = �0.01, t = �0.12, P = �0.90) or anxiety (respec-tively, b = �0.19, t = �1.48, P = 0.14, and b = �0.07, t = �0.58,P = 0.57). Because ought-own, but not ought-other, self-discrepan-cies were significantly related to depression in the correlationalanalyses, separate regression analyses were run for the ought-own predicting depression. Yet, ought-own self-discrepancies werenot predictive of depression while controlling for sociodemograph-ic and pain-related variables (b = �0.25, t = �2.50, P = 0.02). Fur-ther, as can be seen in Table 2, the addition of feared-own andfeared-other self-discrepancies (block 3a) to the prediction ofdepression and anxiety resulted in a significant F change, whilethe predictors themselves did not reach statistical significance. Be-cause this might indicate a type II error at the predictor level,regression analyses were conducted for the feared-own and thefeared-other self-discrepancy separately (block 3b and 3c). Resultsdemonstrated that both feared-own and feared-other self-discrep-ancies were significant predictors of depression and anxiety overand beyond the contribution of sociodemographic and pain-relatedvariables (Table 2). So, feeling far away from the feared self waspredictive of lower levels of depression and anxiety.

3.3. Self-discrepancies and activity patterns, disability, and quality oflife

Results from the correlational analyses (Table 1) did not showsignificant associations between self-discrepancies and activitypatterns, yet they did show associations between avoidance behav-ior on the one hand, and disability and quality of life (physical andmental) on the other hand. In contrast, persistence behavior wasnot significantly related to disability or quality of life.

To further test the contribution of self-discrepancies to activitypatterns, 3 hierarchical multiple regression analyses were carriedout. In all regression analyses, sociodemographic variables (gen-der and age) were entered in the first block and pain-related vari-ables (pain duration and pain intensity) in the second block. Toexplain persistence behavior, ideal-own and ideal-other self-dis-crepancies were entered in the third block and, in a secondregression analysis, replaced by ought-own and ought-otherself-discrepancies. In the fourth block, the quadratic terms ofthe self-discrepancies were added to test the U-shaped relation-ship. There were no significant main effects of self-discrepancieson persistence behavior. However, the quadratic term of theideal-other significantly predicted persistence behavior (Table 3).

ssion and anxiety while controlling for sociodemographic and pain-related variables.

SE B b DR2 DF P DF

0.23 �0.08 0.01 .29 ns0.12 �0.010.11 0.06 0.16 7.68 60.0010.11 0.40*

0.14 0.24 0.14 7.95 60.0010.13 0.190.11 0.36* 0.12 13.32 60.001

0.10 0.34* .11 12.08 60.0010.23 �0.02 0.01 0.23 ns0.12 �0.07

0.12 �0.01 0.07 2.90 ns0.11 0.260.14 0.23 0.15 7.57 60.0010.14 0.210.11 0.37* 0.13 12.38 60.0010.11 0.36* 0.12 11.88 60.001

Table 3Summary of hierarchical regression analyses for self-discrepancies predicting overdoing and avoidance while controlling for sociodemographic and pain-related variables.

Dependent Blocks Variables B SE B b DR2 DF P DF

Overdoing (POAM-P) 1 Gender �0.02 0.21 �0.01 0.01 0.29 nsAge �0.08 0.11 �0.08

2 Pain duration �0.00 0.11 �0.00 0.05 2.04 nsMean pain intensity 0.22 0.11 0.22

3 Ideal-own �0.06 0.12 �0.06 0.03 1.11 nsIdeal-other 0.24 0.12 0.25

4 Ideal-own2 0.00 0.01 0.09 0.10 4.69 60.01Ideal-other2 0.01 0.01 0.30*

Avoidance (POAM-P) 1 Gender 0.16 0.22 0.08 0.01 0.36 nsAge �0.06 0.12 �0.06

2 Pain duration 0.19 0.12 0.19 0.05 1.88 nsMean pain intensity 0.13 0.11 0.13

3a Feared-own 0.40 0.15 0.39* 0.08 3.69 nsFeared-other �0.20 0.14 �0.20

3b Feared-own 0.26 0.11 0.26 0.06 5.29 ns3c Feared-other 0.04 0.11 0.04 0.00 0.14 ns

POAM-P, Patterns of Activity Measure-Pain.* P 6 0.01.

H.P.J. Kindermans et al. / PAIN�

152 (2011) 403–409 407

This means that participants who felt close to their ideal-other aswell as those who felt far away from their ideal-other, bothshowed higher levels of persistence behavior (Fig. 1). With regardto ought self-discrepancies, neither the quadratic terms of theought-own self-discrepancy, nor ought-other self-discrepancywere significant predictors of persistence behavior (respectively,b = 0.05, t = 0.39, P = 0.70, and b = 0.14, t = 1.08, P = 0.29). In theregression analysis with avoidance behavior as the dependentvariable, the feared-own and feared-other were entered in thethird block. Results showed that the closer participants felt tothe person they feared to be, the more likely they were to exertavoidance behavior (Table 3). In line with aforementioned analy-ses, the feared-own (block 3b) and the feared-other (block 3c)were also tested separately. As presented in Table 3, the feared-own was no longer a significant predictor of avoidance whenthe effect of the feared-other was not controlled for.

Hypotheses concerning the prediction of disability and qualityof life from activity patterns were also tested with hierarchicalregression analyses for each activity pattern separately (Table 4).In line with aforementioned analyses, sociodemographic andpain-related variables were controlled for in the first 2 blocks. Inthe third block, either persistence or avoidance behavior was en-tered. Results showed that avoidance but not persistence behavior

Fig. 1. The relationship between overdoing and ideal-other self-discrepancies.

was a significant predictor of disability and mental and physicalquality of life.

4. Discussion

The aim of the present study was twofold. The first aim was toinspect the emotional consequences of self-discrepancies in pa-tients with chronic pain. The second aim was to extend the studyof self-discrepancies in patients with chronic pain by testing self-discrepancies’ value in predicting self-reported activity patternsand, as a consequence of these activity patterns, disability levelsand quality of life. With regard to the first aim of the study, resultsshowed that feared (own and other) self-discrepancies were in-deed related to emotional distress. Yet, ought and ideal (own orother) self-discrepancies were not. Regarding the second aim, re-sults showed that the feared-own, but not the feared-other, wasassociated with avoidance. Furthermore, a U-shaped relationshipbetween the ideal-other and persistence behavior was demon-strated. Finally, avoidance, but not persistence, behavior contrib-uted to higher levels of disability and diminished mental andphysical health.

Based on the present results, emotional distress appearsstrongly related to the experienced proximity to the feared self, be-cause this was associated with and predictive of both anxiety anddepression. This is also in line with findings from Goossens and col-leagues [11], who studied patients with chronic pain in the upperextremities and found that patients who felt close to their fearedself reported higher levels of depression and anxiety. In contrastwith Waters and colleagues [35], who studied chronic low backpain patients, ideal (own and other) self-discrepancies were not re-lated to depression nor anxiety in the present study. Further,ought-own self-discrepancies were related to depression but thisrelationship was no longer significant after controlling for sociode-mographic and pain-related variables. The finding that ought-otherself-discrepancies were not related to anxiety or depression is incontrast with previous studies, demonstrating detrimental conse-quences of ought self-discrepancies, including depressed and anx-ious mood, in chronic pain patients [11,35]. The present resultsmight originate from the dynamics between different types ofself-discrepancies. Carver and colleagues [6] already suggestedthat proximity to the feared self might minimize the importanceof approaching desired end states, such as the ideal and ought self,because of the salience of the feared self. This is also supported byother studies, showing that the undesired self, which is closely re-lated to the feared self, is a better predictor of well-being-related

Table 4Summary of hierarchical regression analyses for activity patterns predicting disability and quality of life (QOL) while controlling for sociodemographic and pain-related variables.

Dependent Blocks Variables B SE B b DR2 DF P DF

Disability (RDQ) 1 Gender �0.09 0.22 �0.05 .00 .10 nsAge �0.01 0.11 �0.01

2 Pain duration �0.00 0.10 �0.00 .29 16.00 P < 0.001Mean pain intensity 0.52 0.09 0.54*

3a Overdoing 0.07 0.10 0.07 .00 .55 ns3b Avoidance 0.41 0.08 0.43* .17 25.11 P < 0.001

Physical QOL (RAND-36) 1 Gender 0.18 0.21 0.09 .07 3.01 nsAge 0.24 0.11 0.24

2 Pain duration �0.03 0.09 �0.03 .40 29.39 P < 0.001Mean pain intensity �0.63 0.08 �0.64*

3a Overdoing �0.09 0.09 �0.09 .01 1.04 ns3b Avoidance �0.34 0.07 �0.35* .12 21.30 P < 0.001

Mental QOL (RAND-36) 1 Gender 0.19 0.23 0.09 .03 1.01 nsAge 0.12 0.12 0.12

2 Pain duration �0.09 0.11 �0.09 .23 11.90 P < 0.001Mean pain intensity �0.49 0.10 �0.47*

3a Overdoing �0.15 0.10 �0.14 .01 1.23 ns3b Avoidance �0.38 0.09 �0.37* .13 15.89 P < 0.001

RDQ, Roland Disability Questionnaire.* P < 0.001.

408 H.P.J. Kindermans et al. / PAIN�

152 (2011) 403–409

constructs [23,24]. It might be that the feared self was also moresalient to participants in the present study.

With respect to behavioral consequences of self-discrepancies,ideal and ought self-discrepancies were hypothesized to be relatedto persistence behavior. Results showed that patients who feltclose to their ideal-other as well as those who experienced greatdiscrepancies between their actual and ideal-other, both showedhigher levels of persistence behavior. For the participants who feeldiscrepant from their ideal-other self, persistence behavior mightbe an (unsuccessful) attempt to get closer to their ideal self. Forthose who feel close to their ideal self, persistence behavior mayreflect the possibility to continue their daily activities despite pain,and in that case, persistence behavior may be inversely related todisability. This ambiguity may be one of the reasons why there isno straightforward relationship between persistence behaviorand disability levels [17]. While our hypothesis on the U-shapedrelationship between self-discrepancies and persistence behaviorwas restricted to the ideal-other, this finding sheds new light onthe functional meaning of persistence behavior when consideringself-discrepancies.

With regard to avoidance behavior, results demonstratingfeared-own self-discrepancies were related to avoidance. It is note-worthy that feared-other self-discrepancies acted as a suppressorvariable in the regression analysis because the effect of thefeared-own was significant only while controlling for the feared-other. Nevertheless, the present results showed that participantswho felt close to their feared self from their own perspective alsoreported more avoidance behavior. In a similar vein, avoidancebehavior as predicted in the fear-avoidance model is also the con-sequence of fears, yet on a more basic ‘‘do-level.’’ The object of fearin the present study, namely becoming someone you do not wantto be, acts on a higher-order ‘‘be-level’’ but seems to have similarconsequences. This means that the feared self may be understoodin terms of an antigoal and, as antigoals are likely to instigatesafety behavior [20], in patients with chronic pain it is likely thatthis leads to a tendency to avoid activities. In other words, this isthe first study showing that the validity of Fear-Avoidance modelsappears to generalize to the level of perceived self and identity.

In general, the level of measurement of self-discrepancies mightplay a role with regard to the limited results between self-discrep-ancies and activity patterns. Self-discrepancies are higher-orderconstructs, meaning that they are the result of cognitive higher-or-der processes, in which an individual reflects on his or her self. A

content analysis of self-discrepancies has also shown that the ideal,ought, and feared selves concern general attributes mostly relatedto the interpersonal domain (eg, honest) [17]. Activity levels, onthe other hand, are lower-order constructs measured at abehavioral level, and relate to the intrapersonal domain (eg, doinghousehold chores). In order to overcome this conceptual distance,self-discrepancies might be measured on a lower level by, forexample, focusing on activities. It might be discussed, however,to what extent this operationalization justifies the use of the termself-discrepancies instead of activity-discrepancies.

Partly in line with our hypotheses, the current results demon-strated that avoidance behavior was predictive of higher levels ofdisability and diminished quality of life. Yet persistence behaviorwas not. The results are, however, in line with findings of previousstudies on activity patterns [14,21] in which no detrimental effectsof persistence behavior were found. In the study of McCracken andSamuel [21], persistence behavior was unrelated to disability.Moreover, Hasenbring and colleagues [14] found that persistencebehavior was associated with lower levels of disability. It is note-worthy that in these 3 studies, 3 different measurement instru-ments were used to assess persistence behavior. It is possiblethat, despite the agreement in the theoretical conceptualizationof persistence behavior, different items might tap into different as-pects of persistence behavior. For example, an item focusing on‘‘being highly active’’ versus an item focusing on ‘‘experiencing in-creased pain as a consequence of being highly active’’ might be dif-ferentially related to disability. The finding of the present studythat persistence behavior was related to being close to, as well asbeing far away from, the ideal-other self, adds up to the impor-tance of taking a closer look at the concept of persistence behavior.

Future studies should be attentive to the dynamics betweentypes of self-discrepancies, but might also consider focusing onone type of self-discrepancy in order to gain more insight into di-rect effects on emotional or behavioral outcomes. Furthermore, theperspective on the self should not be neglected as different resultsemerge from self-discrepancies from the ‘‘own’’ perspective versusself-discrepancies from the ‘‘other’’ perspective. It might be inter-esting for future studies to also consider the relationship betweenthe patient and his or her significant other, as listed in the self-dis-crepancy interview, because this might be of importance for thestrength of the self-discrepancy. For the study of the role of self-discrepancies in activity patterns, future studies might considerpossible mediating or moderating variables. For example, flexible

H.P.J. Kindermans et al. / PAIN�

152 (2011) 403–409 409

goal adjustment has been shown to be protective against negativeemotions despite the presence of self-discrepancies [11].

The integration of self-discrepancies in the explanatory mecha-nism behind emotional and behavioral consequences of chronicpain might have clinical implications. Clinicians should keep inmind that patients may be burdened by fears not only at the levelof activities but also at the level of being. Chronic pain patientsmight benefit from clinical interventions targeting the way pa-tients see themselves and working with the personal goals theyset for themselves, which are defined in the ideal and ought self.Self-system therapy, which is derived from SDT, has previouslybeen employed in the treatment of depression and has shownpromising results [28,32].

The shortcomings of the present study, being that the resultsrely on self-report and as the design was cross-sectional, no causalrelationships could be confirmed, imply that future studies shouldfocus on longitudinal designs to examine associations in time andexperimentally testing self-discrepancies in patients with chronicpain. It would be interesting to experimentally activate specificself-discrepancies and determine under controlled circumstanceswhether and how they influence pain behavior. Furthermore, cau-tion is needed with respect to generalizing present findings toother pain populations because the present study involved a rela-tively homogeneous group of chronic low back pain patients only.

Considering that this is the first study to investigate this rela-tionship, it is difficult to be conclusive about the role of self-dis-crepancies in the activity patterns of chronic pain patients.Nevertheless, the present results add to the importance of studyingthe feared self in patients with chronic pain and to consider bothemotional and behavioral consequences, as the feared-own was re-lated to avoidance. On the other hand, ideal self-discrepanciesmight be important in the light of explaining persistence behavior.Yet, based on present results, the ought self might be less relevantin explaining emotional or behavioral consequences. Importantly,avoidance but not persistence behavior appeared to predict dis-ability and diminished quality of life.

Acknowledgments

The authors wish to thank Stephen Morley for the interestingand fruitful discussions on the study of self-discrepancies in pa-tients with chronic pain. We are also grateful to Erik Schoutenand Gerard van Breukelen for their statistical assistance and toAstrid Dello, Truus Uijen, and Sita van Riet for their help in datacollection and organization. The study was supported by the Coun-cil for Medical and Health Research of the Netherlands (ZON-MW),Grant No. 14350042. The contribution of Johan W.S. Vlaeyen wassupported by the NWO Social Sciences Research Council of TheNetherlands, Grant No. 453-04-003. Participation of Dr. Roelofswas supported by the EFIC-Grünenthal Grant (EGG)The authorshave no conflict of interest with regard to the present study.

References

[1] Beck AT, Steer RA, Brown GK. Manual for the beck depression inventory. 2nded. San Antonio, TX: The Psychological Corporation; 1996.

[2] Beck AT, Steer RA, Brown GK, Van der Does AJW. BDI-II-NL Handleiding (BDI-II-Dutch Manual). Lisse, The Netherlands: Psychological Corporation; 2002.

[3] Boldero J, Francis J. Ideals, oughts, and self-regulation: are there qualitativelydistinct self-guides? Asian J Soc Psychol 1999;2:343–55.

[4] Brouwer S, Kuijer W, Dijkstra PU, Goeken LN, Groothoff JW, Geertzen JH.Reliability and stability of the Roland morris disability questionnaire: intraclass correlation and limits of agreement. Disabil Rehabil 2004;26:162–5.

[5] Cane D, Nielson WR, Ivey B, Ivey R, McCarthy M, Mazmanian D. Developmentand preliminary evaluation of a measure of three pain-related patterns ofactivity. In: 27th Annual Canadian Pain Society Meeting, Ottowa, Ontaria;2007.

[6] Carver CS, Lawrence JW, Scheier MF. Self-discrepancies and affect:incorporating the role of feared selves. Pers Soc Psychol Bull 1999;25:783–92.

[7] Carver CS, Scheier FM. On the self-regulation of behavior. Cambridge,UK: Cambridge University Press; 1998.

[8] Crombez G, Eccleston C, Baeyens F, Eelen P. Attentional disruption is enhancedby the threat of pain. Behav Res Ther 1998;36:195–204.

[9] Eccleston C, Crombez G. Pain demands attention: a cognitive-affective modelof the interruptive function of pain. Psychol Bull 1999;125:356–66.

[10] Faas A, Chavannes AW, Koes BW, van den Hoogen JMM, Mens JMA, SmeeleLJM, Romeijnders ACM, van der Laan JR. NHG-standaard M54 lage-rugpijn[Dutch]. Huisarts Wet 1996;39:18–31.

[11] Goossens M, Kindermans HPJ, Morley S, Roelofs J, Verbunt JA, Vlaeyen J. Self-discrepancies in work related upper extremity pain: relation to emotions andflexible goal adjustment. Eur J Pain 2010;14:764–70.

[12] Hardin EE, Leong TF. Optimism and pessimism as mediators of the relationsbetween self-discrepancies and distress among Asian and EuropeanAmericans. J Couns Psychol 2005;52:25.

[13] Harris CA, D’Eon JL. Psychometric properties of the beck depression inventory-second edition (BDI-II) in individuals with chronic pain. Pain2008;137:609–22.

[14] Hasenbring M, Hallner D, Rusu AC. Fear-avoidance- and endurance-relatedresponses to pain: development and validation of the avoidance-endurancequestionnaire (AEQ). Eur J Pain 2009;13:620–8.

[15] Hasenbring MI, Plaas H, Fischbein B, Willburger R. The relationship betweenactivity and pain in patients 6 months after lumbar disc surgery: do pain-related coping modes act as moderator variables? Eur J Pain 2006;10:701–9.

[16] Higgins E. Self-discrepancy: a theory relating self and affect. Psychol Rev1987;94:319–40.

[17] Kindermans HPJ, Goossens M, Roelofs J, Huijnen IPJ, Verbunt JA, Morley S,Vlaeyen J. A content analysis of ideal, ought, and feared selves in patients withchronic low back pain. Eur J Pain 2010;14:648–53.

[18] Lamé IE, Peters ML, Vlaeyen JWS, Patijn J, Kleef Mv. Quality of life in chronicpain is more associated with beliefs about pain, than with pain intensity. Eur JPain 2005;9:15–24.

[19] Leeuw M, Goossens ME, Linton SJ, Crombez G, Boersma K, Vlaeyen JW. Thefear-avoidance model of musculoskeletal pain: current state of scientificevidence. J Behav Med 2007;30:77–94.

[20] Mansell W. Control theory and psychopathology: an integrative approach.Psychol Psychother 2005;78:141–78.

[21] McCracken LM, Samuel VM. The role of avoidance, pacing, and other activitypatterns in chronic pain. Pain 2007;130:119–25.

[22] Morley S, Eccleston C. The object of fear in pain. In: Asmundson GJ, Vlaeyen JW,Crombez G, editors. Understanding and treating fear of pain. Oxford,UK: Oxford University Press; 2004. p. 163–88.

[23] Ogilvie DM. The undesired self: a neglected variable in personality research. JPers Soc Psychol 1987;52:379–85.

[24] Phillips AG, Silvia PJ, Paradise MJ. The undesired self and emotionalexperience: a latent variable analysis. J Soc Clin Psychol 2007;26:1035–47.

[25] Price DD, McGrath PA, Rafii A, Buckingham B. The validation of visual analoguescales as ratio scale measures for chronic and experimental pain. Pain1983;17:45–56.

[26] Roland M, Morris R. A study of the natural history of back pain. Part I:development of a reliable and sensitive measure of disability in low-back pain.Spine 1983;8:141–4.

[27] Spinhoven P, Ormel J, Sloekers PP, Kempen GI, Speckens AE, Van Hemert AM. Avalidation study of the hospital anxiety and depression scale (HADS) indifferent groups of Dutch subjects. Psychol Med 1997;27:363–70.

[28] Strauman TJ, Vieth AZ, Merrill KA, Kolden GG, Woods TE, Klein MH, PapadakisAA, Schneider KL, Kwapil L. Self-system therapy as an intervention for self-regulatory dysfunction in depression: a randomized comparison withcognitive therapy. J Consult Clin Psychol 2006;74:367–76.

[29] van der Zee K, Sanderman R. Het meten van de algemene gezondheidstoestandmet de RAND-36: een handleiding Groningen, The Netherlands: NCG; 1993.[Dutch].

[30] van der Zee K, Sanderman R, Heyink J, De Haes H. Psychometric qualities of theRAND 36-item Health Survey 1.0: a multidimensional measure of generalhealth status. Int J Behav Med 1996;3:104–22.

[31] Van Houdenhove B, Neerinckx E. Is ‘‘ergomania’’ a predisposing factor tochronic pain and fatigue? Psychosomatics 1999;40:529–30.

[32] Vieth A-Z, Strauman T-J, Kolden G-G, Woods T-E, Michels J-L, Klein M-H. Self-system therapy (SST): a theory-based psychotherapy for depression. ClinPsychol Sci Pract 2003;10:245–68.

[33] Vlaeyen JW, Linton SJ. Fear-avoidance and its consequences in chronicmusculoskeletal pain: a state of the art. Pain 2000;85:317–32.

[34] Vlaeyen JW, Morley S. Active despite pain: the putative role of stop-rules andcurrent mood. Pain 2004;110:512–6.

[35] Waters SJ, Keefe FJ, Strauman TJ. Self-discrepancy in chronic low back pain:relation to pain, depression, and psychological distress. J Pain SymptomManage 2004;27:251–9.

[36] Zigmund AS, Snaith RP. The hospital anxiety and depression scale. ActaPsychiatr Scand 1983;67:361–70.