Are Illness Perceptions Associated with Pain and Disability in Complex Regional Pain Syndrome? A Cross-Sectional Study

Abstract

Objective

Complex regional pain syndrome (CRPS) is a complex and often poorly understood condition, and people with CRPS will have diverse beliefs about their symptoms. According to the self-regulation model, these beliefs (termed “illness perceptions”) influence health behaviors and outcomes. Previous studies have found that psychological factors influence CRPS outcomes, but few studies have investigated CRPS patients’ illness perceptions specifically. The present study examined whether illness perceptions were related to pain intensity and other relevant outcomes in people with CRPS.

Methods

In this cross-sectional study, 53 patients with CRPS (type 1 and type 2) completed questionnaires assessing illness perceptions, pain, disability, and psychological factors. Multiple regression analyses were used to determine whether illness perceptions were associated with pain intensity, disability, depression, and kinesiophobia, after controlling for possible covariates (including clinical and demographic factors, pain catastrophizing, and negative affect).

Results

Negative illness perceptions were associated with greater pain, disability, and kinesiophobia, but not depression. Specifically, attributing more symptoms to CRPS (more negative illness identity perceptions) was associated with greater pain intensity, and reporting a poorer understanding of CRPS (lower illness coherence scores) was associated with greater disability and kinesiophobia.

Conclusions

Patients with CRPS with more negative illness perceptions (particularly perceptions indicating a poor understanding of their condition) also experience greater pain, disability, and kinesiophobia. Future research could investigate whether altering CRPS patients’ illness perceptions influences clinical outcomes.

Introduction

Complex regional pain syndrome (CRPS) is a chronic pain condition that occurs predominantly in a limb, often resulting from an accident or trauma. The pathophysiology of CRPS is complex and not completely understood [1], and, like other chronic pain conditions, psychological factors may play a role. In particular, although psychological factors do not appear to play a causal role in the development of CRPS [2], psychological factors appear to predict important CRPS outcomes, including pain, disability, depression [3, 4], anxiety [5, 6], and fear-avoidance [7, 8]. Work examining chronic pain more broadly has found that the beliefs people have about their pain reliably predict pain-related outcomes, with those believing that their condition is more threatening experiencing worse outcomes [9, 10]. It therefore seems likely that beliefs about pain and related symptoms might also influence outcomes in CRPS; however, there has been little investigation into CRPS patients’ idiosyncratic beliefs.

One common way of conceptualizing and measuring patients’ beliefs about their health condition is by assessing their illness perceptions. The term illness perceptions originates from Leventhal’s common-sense model of self-regulation (see Figure 1) [11], which proposes that in order to make sense of any health threat, patients develop a set of perceptions or beliefs about their condition [12]. The model theorizes that these perceptions are determined by the specific health threat and will shape the coping strategies and health behaviors that patients adopt, thereby influencing health outcomes [13]. Illness perceptions have been conceptualized as a set of beliefs that fall along the following dimensions: identity—the label or symptoms that people associate with the condition; timeline—the length of time that people believe their condition will last; consequences—the impact of the condition on the person’s life; control/cure—the degree to which the condition can be managed, resolved, or treated; cause—what the person thinks is responsible for the condition; coherence—an individual’s understanding of the condition; and emotional representations—the emotional impact of the condition [14]. Research has found that illness perceptions predict outcomes in a wide range of patient groups, including diabetic, cardiac, neurological, and cancer patients [15–17]. Studies have also shown that altering illness perceptions can lead to better clinical outcomes [18]. In the field of chronic pain, studies have found that negative illness perceptions are associated with worse outcomes in terms of pain-related disability [19], depression, anxiety [20, 21], physical function [22], and symptom severity [21, 23].

Despite findings that illness perceptions predict outcomes among people with chronic pain and that CRPS is a complex pain condition that is difficult for patients to understand, the illness perceptions held by people with CRPS have received little attention. Illness perceptions have, to our knowledge, only once been investigated in patients with CRPS [24]. Although that study reported illness perceptions scores for a CRPS sample and found that these were typically negative and catastrophic, the study did not investigate the relationship between illness perceptions and pain-related outcomes. Thus, the aim of the present study is to explore whether illness perceptions are associated with pain, function, and psychological status in people with CRPS. We hypothesized that negative illness perceptions would be associated with worse outcomes in terms of pain intensity, disability, depression, and kinesiophobia.

Methods

Participants

Participants were 53 people diagnosed with CRPS (type 1 and type 2) that had been seen at The Auckland Regional Pain Service (TARPS), an Interdisciplinary Pain Centre. Participants were recruited between April 2018 and July 2018 and were invited to participate if they had a current diagnosis of CRPS as per the International Association for the Study of Pain (IASP) “Budapest” diagnostic criteria [25] (made by a pain specialist during a 90-minute consult), were more than 18 years of age, and spoke English. Participants were excluded if they no longer met the Budapest diagnostic criteria for CRPS at the time of screening.

Procedure

Ethical approval was granted by the University of Auckland’s Ethics Committee on January 26, 2018, for three years (reference number 020465). Institutional approval was obtained through the Auckland District Health Board Research Review Committee on April 4, 2018 (A + 7963).

Potential participants were identified by students on clinical placement, who reviewed the TARPS database and contacted potential participants by telephone to request consent to forward their details to the principal investigator. The investigator then contacted potential participants, explained the study, and ensured that they still reported symptoms meeting the CRPS Budapest diagnostic criteria by using a tool designed by the COMPACT (Core Outcome Measurement Set for CRPS) research consortium [26]. Those who met the inclusion criteria were provided with the participant information sheet and consent form and were either sent a link to an online questionnaire or mailed a physical copy of the questionnaire. Participants received a $NZ20 voucher as compensation for their participation.

Measures

Demographic and Clinical Characteristics

Demographic characteristics, including age, gender, ethnicity, relationship status, and employment status, were collected, along with clinical characteristics, such as current medication use, CRPS symptom duration, CRPS trigger, CRPS locality, CRPS symptoms, and prior handedness, which were assessed by using questions designed by the COMPACT research consortium [26].

Illness Perceptions Questionnaire–Revised (IPQ-R)

The IPQ-R [14] is a 42-item measure of illness perceptions with good predictive, discriminant, and known groups validity [14, 27]. The first section measures illness identity perceptions and contains 14 items that measure the number of symptoms attributed to CRPS in a yes/no response format, with higher scores representing more negative illness identity perceptions. The second section contains seven subscales (38 items), assessing: timeline–acute/chronic (how long the patient believes the illness/condition will last), timeline–cyclical (whether the patient believes the illness comes and goes in cycles), consequences (the perceived impact that the illness has on the patient’s life), coherence (how well the patient understands the illness), personal control (how much control the patient feels they have over the illness), treatment control (how much control the patient feels the treatment has on the illness), and emotional representations (the emotional impact of the illness). Items are scored on five-point Likert scales, ranging from strongly disagree to strongly agree. Total scores on each subscale were calculated by reverse-coding negatively worded items before summing the items for each subscale. Higher scores on the timeline, consequence, and emotional representations subscales represent more negative illness perceptions. Lower scores on the coherence, personal control, and treatment control subscales represent more negative illness perceptions. The internal reliability of each subscale ranged from satisfactory to excellent: identity (α=0.80), timeline–acute/chronic (α=0.84), timeline–cyclical (α =0.68), consequences (α=0.76), personal control (α=0.78), treatment control (α=0.87), and emotional representations (α=0.75). The third section of the IPQ-R contains 18 items assessing causal perceptions, again rated on five-point Likert scales, from strongly disagree to strongly agree. The final section asks participants to name and rank their top three perceived causes of their condition.

Pain Numeric Rating Scale (NRS)

This single-item scale was used to assess the participants’ average pain level over the prior 2 days on a scale ranging from 0 (no pain) to 10 (worst pain you can imagine). This scale has been used in a number of chronic pain populations [28].

Short-Form McGill Pain Questionnaire (SF-MPQ-2)

The SF-MPQ-2 [29] is a pain scale that contains subscales for continuous pain, intermittent pain, neuropathic pain, and the affective component of pain. Participants rate 22 single-word descriptors on a scale from 0 (none) to 10 (worst possible), with higher scores on an item indicating a strong identification with that pain descriptor. The scale has been used widely in patient populations, including chronic and acute pain groups [30]. Prior work has reported that the convergent and discriminant validity of this scale is adequate [29], and in the present study the internal reliability was excellent (α= 0.91).

Pain Disability Index (PDI)

The PDI [31] consists of seven items that assess the degree to which chronic pain impacts various daily activities. Each item is scored on a Likert scale ranging from 0 (no disability) to 10 (total disability), and all items are summed to create a disability score, with higher scores indicating greater disability. Prior work has demonstrated that this scale has good concurrent and construct validity [31], and the internal reliability was adequate in the present study (α=0.75).

Tampa Scale for Kinesiophobia 11 (TSK-11)

The TSK-11 [32] is an 11-item measure of pain-related fear of movement (kinesiophobia). Participants score each item from 1 (highly disagree) to 4 (highly agree). Total scores range from 11 to 44, with higher scores indicating greater pain-related fear. Previous research indicates that the TSK-11 has adequate concurrent and predictive validity [32], and it was found to have good internal reliability in the present study (α=0.85).

Depression, Anxiety, and Stress Scale (DASS-21)

The DASS-21 [33] was used to assess depression, anxiety, and stress. This self-report questionnaire consists of three seven-item subscales (one each for depression, anxiety, and stress) that are scored on four-point Likert scales, ranging from 0 (did not apply to me at all) to 3 (applied to me very much or most of the time). Total scores were computed on each subscale, with higher scores indicating greater levels of depression, anxiety, or stress. The DASS-21 has been found to have adequate convergent and discriminant validity [33], and in the present study internal reliability scores were excellent for the depression (α =0.92) and stress (α =0.89) subscales and adequate for the anxiety subscale (α =0.79).

Pain Catastrophizing Scale (PCS)

The PCS [34] is a 13-item scale that assesses the degree to which participants experience catastrophic thoughts about their pain. Each item is rated on a four-point Likert scale, from 0 (not at all) to 4 (all the time). Total scores were calculated by summing the responses to the 13 items (0–52), with higher scores indicating greater catastrophizing. Previous research has demonstrated that the PCS has good concurrent and discriminant validity [35], and it was found to have excellent internal reliability in the present study (α =0.91).

Positive and Negative Affect Scale (PANAS)

The PANAS [36] was used to assess trait negative affect. This self-report scale consists of 20 single-word items (10 items on positive affect and 10 items on negative affect). For each item, participants rate the extent to which they generally experience that emotion, with responses ranging from 1 (very slightly or not at all) to 5 (extremely). Total scores were calculated on the negative subscale. This has been shown to have adequate convergent and discriminant validities [36, 37], and in the present study it had excellent internal reliability (α =0.91). The positive affect scale was not intended to be used in analyses but was administered because it is possible that omitting the positively worded items from the scale would affect responding.

Data Analysis

An a priori power analysis was run to achieve power of 0.80 [38] using G*Power software 3.1 [39]. On the basis of previous research on illness perceptions in a pain population [40], we expected a large effect size. The power analysis demonstrated that 46 participants were required; however, because some participants took longer than expected to complete the study, additional participants had already been contacted, so the final sample size was 53. Data were analyzed with Statistical Package for the Social Sciences (SPSS) Version 25. An alpha level of 0.05 was used. The data were assessed for normality, linearity, and outliers. The majority of the data were normally distributed; however, some scales showed a non-normal trend (timeline, timeline–cyclical, consequences, anxiety, depression, and negative affect), so they were analyzed with nonparametric tests. The results showed no difference from a comparative analysis with parametric tests, so parametric tests were used, given guidelines indicating that normality can be assumed if the sample size is greater than 30 [38].

Univariate Analyses

Pearson r correlations were used to test whether there were relationships between illness perceptions, pain outcomes, and psychological factors. Following this, Pearson r correlations, independent t tests, and a one-way analysis of variance (ANOVA) with Tukey’s post hoc tests were used to test whether demographic and clinical characteristics were related to outcome variables so that these could be included as covariates in multiple regression analyses.

Because of sample size restrictions, a factorial analysis of the causal scale could not be completed. Therefore, the causal component of illness perceptions was analyzed by two independent researchers coding each participant’s top reported causal belief for having CRPS into the following themes: injury, poor medical care, psychological cause, and physiological cause.

Multivariate Analyses

Five multiple linear regressions were conducted to determine whether illness perceptions were associated with each dependent variable: 1) pain intensity (measured by NRS), 2) pain intensity (measured by SF-MPQ-2), 3) disability, 4) depression, and 5) kinesiophobia. For each dependent variable, a stepwise approach was taken, with relevant demographic or clinical variables entered on the first step. Negative affect was then entered (as negative affect could explain associations between illness perceptions and the self-reported outcomes, i.e., negative self-report bias), followed by illness perceptions or psychological variables in the final step. These models were tested to meet assumptions of multicollinearity (variance inflation factor and tolerance scores) and homoscedasticity [38]. Because of the high correlations between illness perception variables and psychological variables (see Table 3) and the potential for multicollinearity, a maximum of five possible predictors were used in each model. Predictors were chosen if they were 1) theoretically plausible and 2) had a strong, significant univariate relationship with the outcome variable (variables with the strongest Pearson r correlation with each dependent variable were prioritized). The specific predictor variables included in each model are shown in the regression tables (Table 4).

Results

The sample consisted of 53 patients with CRPS. The majority were female, and the mean age was 51 years. Most participants were of New Zealand European ethnicity, with a small number of Māori and other ethnicities. See Table 1.

Psychological and Pain Characteristics

Means and standard deviations for scores for illness perceptions, pain, and pain-related outcomes are displayed in Table 2.

Causal Beliefs

Causal beliefs were coded into one of four themes (see Figure 2). More than half of the participants endorsed an injury as the main cause of their CRPS (e.g., “broke my bone”), just under a quarter of participants endorsed poor medical care (e.g., “surgeon”), and the other participants endorsed psychological (e.g., “depression”) and physiological (e.g., “nerves misfiring”) causes. See Figure 2.

Univariate Results: Associations Between Demographic/Clinical Characteristics and Outcome Variables

To determine which covariates to include in multivariate models, correlations were run between demographics, clinical characteristics, and outcome variables. Pearson’s correlations showed that symptom duration (r₍₄₉₎=0.30, P=0.033) and total number of medications (NRS) (r₍₅₁₎=0.28, P=0.045) were positively associated with pain intensity (NRS). One-way ANOVA revealed that employment status was associated with pain intensity (measured by the NRS) (F_(6,46)=3.91; P=0.003), with participants receiving a sickness benefit or Accident Compensation Corporation (ACC) income reporting higher pain intensity than those who were employed full-time. There was also a significant difference between employment groups for disability scores (F(_6,46)=5.29; P=<0.001), with Tukey post hoc tests revealing that those receiving ACC compensation reported higher disability than those who were studying or employed full-time. There were also significant correlations between a number of illness perceptions and all of the outcome variables (See Table 3).

Multivariate Analyses

Pain Intensity

To test whether illness perceptions were associated with pain intensity, two hierarchical multiple linear regression models were created. To demonstrate the reliability of findings, one model used the NRS measure of pain intensity as the dependent variable, and the other used the SF-MPQ-2.

For the NRS model, symptom duration was entered in Step 1. This produced a significant model (F_(1,48) =4.66; P =0.036) and explained 9% of the variance in pain intensity. Longer symptom duration was associated with greater pain intensity (β =0.30; P=0.036). The inclusion of negative affect in Step 2 also produced a significant model (F_(2,47) =4.89; P =0.012) and explained a further 6% of the variance. Longer symptom duration continued to predict greater pain intensity (β=0.28; P =0.039), as did greater negative affect (β=0.29; P =0.034). The inclusion of illness identity perceptions in Step 3 also produced a significant model (F_(3,46) =5.23; P =0.003), explaining a further 11% of the variance of pain intensity. More negative illness identity perceptions (higher scores or attributing more symptoms to CRPS) were associated with greater pain intensity (β=0.30; P=0.029) (see Table 4).

For the second pain intensity model using the SF-MPQ-2, no demographic or clinical covariates were identified, so negative affect was entered in Step 1. This produced a significant model (F_(1,50)=20.30; P <0.001) and explained 29% of the variance in pain intensity. Greater negative affect was associated with greater pain intensity (β=0.54; P <0.001). The inclusion of illness identity perceptions and anxiety in Step 2 also produced a significant model (F_(3,48)=10.33; P < 0.001) and explained a further 10% of the variance in pain intensity. More negative illness identity perceptions (higher scores or attributing more symptoms to CRPS) (β=0.28; P =0.032) and greater negative affect (β=0.39; P =0.018) were associated with greater pain intensity (see Table 4).

Disability

For the disability model, the covariates negative affect, pain intensity (SF-MPQ-2), and income compensation were entered in Step 1. This produced a significant model (F_(3,44)=8.42; P <0.001) and explained 37% of the variance in pain disability scores. Receiving income compensation (β=0.40; P=0.005) and greater pain intensity (β=0.36; P=0.018) were associated with greater disability. The inclusion of illness coherence in Step 2 also produced a significant model (F_(3,49)=7.89; P <0.001) and explained a further 5% of the variance. Lower illness coherence (reporting a poorer understanding of CRPS) (β=–0.25; P =0.042), greater pain intensity (β=0.33; P =0.025), and receiving income compensation (β=0.37; P =0.006) were associated with greater disability (see Table 4).

Depression

For the depression model, the covariates negative affect and pain intensity (NRS) were entered in Step 1. This produced a significant model (F_(2,49) =16.71; P <0.001) and explained 41% of the variance in depression scores. Greater negative affect (β=0.60; P <0.001) was associated with greater depression. The inclusion of illness consequence perceptions and catastrophizing in Step 2 also produced a significant model (F_(4,47) =11.72; P <0.001) and explained a further 9% of the variance of depression, with greater negative affect (β=0.43; P <0.001) and greater catastrophizing (β=0.06; P =0.005) associated with greater depression (see Table 4).

Kinesiophobia

For the kinesiophobia model, the covariates symptom duration and pain intensity (SF-MPQ-2) were entered in Step 1. This produced a significant model (F_(2,48) =7.04; P =0.002) and explained 23% of the variance in kinesiophobia. Longer symptom duration was associated with greater kinesiophobia (β=0.35; P=0.01). The inclusion of catastrophizing, timeline perceptions, and illness coherence perceptions in Step 2 also produced a significant model (F_(5,45) =7.75; P =0.001) and explained a further 23% of the variance of kinesiophobia. Greater catastrophizing (β=0.36; P=0.006) and lower illness coherence perceptions (reporting a poorer understanding of CRPS) (β=–0.27; P=0.019) were associated with greater kinesiophobia (see Table 4).

Discussion

The present study found that the illness perceptions of patients with CRPS are associated with pain and functional status. Specifically, negative illness perceptions were associated with worse outcomes in terms of pain intensity, disability, depression, and fear-avoidance. To reduce the risk that the observed relationships were due to self-report bias or other factors, multivariate analyses were run, controlling for trait negative affect and both clinical and demographic covariates. These showed that higher illness identity perceptions (attributing more symptoms to CRPS) were associated with greater pain intensity. Lower illness coherence perceptions, higher pain, and receiving ACC income compensation were associated with greater disability. Lower illness coherence perceptions and higher levels of catastrophizing were associated with greater kinesiophobia. After controlling for covariates, illness perceptions were not associated with depression, which was most strongly associated with pain catastrophizing.

Pain Intensity

Participants who reported high illness identity perceptions (i.e., attributing more symptoms to CRPS) reported higher pain intensity. Interpretively, one possibility is that people with higher levels of pain may experience more symptoms due to a common underlying factor, such as central sensitization, or higher medication use, which may result in greater side effects [41]. Another possibility is that people who are more hypervigilant may notice more symptoms, including pain, and experience pain at a greater intensity [42]. However, the finding that symptoms that are not considered to be a feature of CRPS (such as a sore throat, breathlessness, etc.) were attributed to CRPS by some participants in this study suggests that patients do not understand which symptoms are caused by CRPS and which are unrelated. Research on symptom reporting shows that people report a median of five symptoms per week, and these symptoms can either be attributed to an illness or interpreted as benign [43]. It therefore seems likely that people with CRPS are attributing a range of benign symptoms to their CRPS, and a lack of understanding about the nature of CRPS may result in the condition being perceived as more threatening, potentially increasing pain in line with the neuromatrix theory of pain [44].

Pain Disability

In the present study, lower illness coherence (poorer understanding of the condition) was associated with greater disability. Although illness perceptions have predicted disability in other patient samples [19, 45, 46], illness coherence (understanding of the condition) was not associated with disability in these studies. Interpretively, illness coherence may be more important in CRPS because CRPS is complex, and several multifaceted physiological processes underlie it (such as neurogenic inflammation, autonomic dysregulation, and structural and functional changes to the central nervous system) [47]. This complexity makes it difficult for patients, medical professionals, and lay people to understand [48]. Patients require a base of understanding to manage their condition well, as their understanding will influence the coping strategies they use [17, 49]. Patients who have a poor understanding of CRPS may engage in more passive coping strategies, such as resting or immobilizing the limb, which could then increase their disability. Research has found that CRPS patients who use more “resting” coping strategies report greater disability [50]. Another likely mechanism is fear-avoidance. Patients who do not understand their condition well may be fearful of damaging their bodies further and therefore are less likely to engage in activities. Thus, the present finding that lower illness coherence perceptions predicted greater disability supports the notion that increasing patients’ understanding of their CRPS could reduce their level of disability, and this is worth exploring in future research.

Depression

The present study found that illness perceptions were not associated with depression in patients with CRPS. This is inconsistent with the illness perception literature, which suggests that illness perceptions predict depression in various patient and pain samples [12, 17, 48, 49]. Even before controlling for covariates, the present investigation found only consequences and emotional representations were correlated with depression. It could be that for patients with CRPS, outwardly negative illness perceptions (such as long timeline beliefs) may actually be realistic and thus indicative of greater pain acceptance, potentially reducing distress [51], but this remains to be explored. Consistent with the literature [52], the present investigation found that catastrophizing was positively associated with depression.

Kinesiophobia

The finding that those with low illness coherence perceptions (poorer understanding) were more kinesiophobic was not surprising. It is likely that those who do not understand their condition may be more fearful that their pain is associated with tissue damage and therefore more concerned about injuring themselves further. In contrast, those with a better understanding of the nature of CRPS may already understand that “hurt” does not equate to “harm” and hence be less fearful of (re)injury. Fear-avoidant patients may also engage in unhelpful coping behaviors, such as immobilization, protecting, or neglecting their limb, which may lead to worse pain outcomes [53]. Such immobilization also means that patients do not have the opportunity to disconfirm their fears, resulting in a vicious cycle of poor understanding, immobilization, and pain.

Theoretical and Clinical Implications

This research demonstrates the applicability of the common-sense model of self-regulation and the fear-avoidance model [54] for conceptualizing the beliefs and perceptions held by patients with CRPS. Interestingly, the link between illness coherence perceptions and kinesiophobia suggests that the common-sense model of self-regulation and the fear-avoidance model overlap and that illness perceptions may influence fear-avoidance. The present study also provides support for the neuromatrix theory of pain as a tool to understand CRPS, as it theorizes that pain is influenced by a number of factors, including physiological (e.g., the central nervous system) and psychological (e.g., perceiving the tissue is under threat) processes [44, 55].

This research may also have clinical implications. Overall, findings suggest that patients who lack an understanding of their condition have the poorest clinical outcomes. Although this observational study cannot demonstrate cause and effect, the results imply that negative illness perceptions, or poor understanding of CRPS, could lead to increased pain and disability or vice versa. Therefore, clinicians may wish to elicit illness perceptions from patients, and this could form a starting point for cognitive therapy (to challenge unhelpful negative illness perceptions) or neurologically based pain education to support greater understanding of CRPS [56]. Future research is needed to assess the efficacy of this approach.

Limitations and Strengths

The present study used a cross-sectional design, which, as mentioned previously, cannot establish cause and effect [38], and thus it is unknown whether holding negative illness perceptions leads to poorer outcomes or vice versa, though the relationship is likely bi-directional. Self-report questionnaires were used for both predictors and outcome variables, and measurement error (such as social desirability) was not controlled for and could have influenced results [57, 58]. However, a measure of negative affect was used to control for a general negative self-report bias, which goes some way toward overcoming this limitation. The present study did not separate CRPS type 1 and CRPS type 2 patients, so it is possible that these patients could have differing illness perceptions because CRPS type 2 is differentiated by a peripheral nerve injury [59].

This study supports the relatively small body of literature showing that psychological factors are important in CRPS. Much of the previous research on psychological factors in CRPS has looked at whether psychological factors cause CRPS, and this view has generally not been supported by the data. However, research does indicate that psychological factors influence outcomes such as pain and disability [60, 61], and the present study provides further evidence to support this possibility. The present study has methodological strengths. It used the Budapest diagnostic criteria [25]; patients had been thoroughly assessed by pain medicine specialists to confirm this diagnosis, and the diagnosis was checked to the extent possible, in line with previous research [53, 62]. The present study used a clinical sample of patients with CRPS with few exclusion criteria, which means results should hold some generalizability for other clinical settings.

Future Directions

The findings from this investigation could be further tested in an experimental intervention to examine whether it is possible to change the illness perceptions of patients with CRPS and whether this results in decreased pain intensity, pain disability, and kinesiophobia. It would also be worthwhile to determine whether there are specific types of illness perceptions that influence catastrophizing and kinesiophobia and that are amenable to change.

Conclusion

This study was the first to investigate whether the illness perceptions of patients with CRPS were associated with pain-related outcomes, and it showed that holding negative illness perceptions (specifically, attributing a greater number of symptoms to CRPS and reporting a poorer understanding of CRPS) are associated with greater pain, disability, and kinesiophobia. Although the study cannot demonstrate causal relationships, the present findings demonstrate a promising area for future research. It is possible that challenging unhelpful illness perceptions may be a useful clinical tool to reduce pain intensity and disability in patients with CRPS. This investigation has provided further support for the importance of psychological factors in CRPS.

Acknowledgments

We thank the team at The Auckland Regional Pain Service and all of the participants for their assistance with the study.

Conflicts of interest: The authors have no conflicts of interest to declare.

References