Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2025 Sep 8;29(9):e70123. doi: 10.1002/ejp.70123

A Model of Body Perception Disturbances in Chronic Limb Pain: The Predictive Role of Kinesiophobia, Depersonalization and Symptom Severity

Hana Karpin 1, Jean‐Jacques Vatine 2, Anatoly Livshitz 3, Irit Weissman‐Fogel 1,✉
PMCID: PMC12417787  PMID: 40922473

ABSTRACT

Background

Body Perception Disturbances (BPD) are common in chronic limb pain conditions characterised by negative feelings toward the limb and a reduced sense of agency. Prior research has focused on isolated associations between psychological factors, pain hypersensitivity and BPD. Therefore, an integrated examination of the interconnections between these variables within a theory‐driven model is necessary.

Methods

The model hypothesises that pain hypersensitivity (hyperalgesia, allodynia), directly linked with BPD (assessed by the Bath‐BPD and Neurobehavioral questionnaires) or indirectly, via symptom severity [assessed by complex regional pain syndrome (CRPS) severity score]; coping strategies (depersonalization, kinesiophobia) and psychological symptoms (somatization, depression) are directly related to BPD; and BPD is associated with pain severity and Quality of Life (QoL).

Results

The model was examined using a path analysis of 92 patients with chronic limb pain. Results indicate that depersonalization was directly linked with the Bath‐BPD (β = 0.50, p < 0.001), and depersonalization and kinesiophobia with the Neurobehavioral (β = 0.24, p = 0.010; β = 0.22, p = 0.020, respectively). CRPS severity score accounts for the associations between hyperalgesia intensity and BPD and is directly related to the Bath‐BPD (β = 0.25, p = 0.014), Neurobehavioral (β = 0.24, p = 0.037), pain (β = 0.28, p = 0.014) and QoL (β = −0.34, p = 0.001). The Bath‐BPD marginally associated with QoL (β = −0.20, p = 0.052) but not with pain severity.

Conclusions

The theory‐driven model fits the data, suggesting that psychological copying strategies play a dominant role in BPD. The symptom severity explains the associations between pain hypersensitivity and BPD and is directly linked to BPD, pain and QoL. The model revealed potential mechanisms underlying BPD and its associated clinical outcomes.

Significance Statement

This study is the first to use path analysis to examine the predictors and effects of Body Perception Disturbances (BPD) in chronic limb pain. Results identified depersonalization and kinesiophobia as key psychological predictors of BPD, while hyperalgesia has no direct effect. The Complex Regional Pain Syndrome (CRPS) severity score is negatively associated with BPD, pain and quality of life. Findings emphasise the role of dysfunctional psychological processes in BPD and suggest that targeting these processes and reducing CRPS symptoms may improve BPD and treatment outcomes.

Keywords: body perception disturbances, chronic limb pain, depersonalization, kinesiophobia, pain hypersensitivity, path analysis model

1. Introduction

Body Perception Disturbances (BPD) are defined as disruptions in the perception of the affected limb (Lewis and McCabe 2010), encompassing negative emotions toward the limb, perceived size distortions, reduced sense of ownership (Acapo et al. 2022; Halicka et al. 2020) and in some cases, even a desire for amputation (Shaikh 2017). BPD are frequently reported in Complex Regional Pain Syndrome (CRPS) (Acapo et al. 2022), a chronic pain condition that often follows limb trauma and is characterised by disproportionate pain along with sensory, motor, autonomic and trophic changes. CRPS includes two subtypes: Type I (without confirmed nerve damage) and Type II (with confirmed nerve damage) (Harden et al. 2022). BPD has also been observed in individuals with chronic limb pain following trauma, not fulfilling the CRPS diagnostic criteria (Hall et al. 2016).

Several factors may contribute to BPD in chronic limb pain. Pain hypersensitivity (i.e., allodynia or hyperalgesia) may lead to a sensory‐motor mismatch (McCabe and Blake 2007), which decreases sensory precision and interrupts motor control (van Rooijen et al. 2013). This process may result in body misperception and BPD (Schütz‐Bosbach et al. 2009). Psychologically, limb appearance dissatisfaction (Sündermann et al. 2020) may lead to depression, which is linked to BPD (Michal et al. 2017; Schulte‐Goecking et al. 2020; Ten Brink et al. 2021). Furthermore, coping strategies such as kinesiophobia to minimise pain (Mibu et al. 2021; Ten Brink et al. 2021) or depersonalization and disassociation (Lewis et al. 2007) are associated with BPD (Michal et al. 2017) alongside anxiety (Schulte‐Goecking et al. 2020) and somatization (Michal et al. 2017).

The relationships between symptom severity and BPD remain unclear. Survey‐based reports indicate that greater CRPS symptoms are associated with higher BPD prevalence (Ten Brink and Bultitude 2021), yet studies using the CRPS Severity index have not found such a link (Bean et al. 2015; Ten Brink et al. 2021). The associations between BPD and pain intensity are inconclusive. Some studies reported correlations between BPD and pain severity both in CRPS (Lewis and Schweinhardt 2012; Schulte‐Goecking et al. 2020; Ten Brink et al. 2021) and other chronic limb pain states (Frettlöh et al. 2006; Michal et al. 2017), while other studies found no significant association (Michal et al. 2017; Wittayer et al. 2018). Furthermore, limited evidence suggests that BPD negatively correlates with Quality of Life (QoL) (Schulte‐Goecking et al. 2020).

The literature suggests that multiple factors may contribute to BPD in chronic limb pain, potentially affecting pain and QoL. However, the evidence is limited and conflicting, leaving the relationships between these variables unclear. This study aimed to (1) examine the predictive role of pain hypersensitivity and psychological factors on self‐reported BPD and (2) explore BPD's link to pain severity and QoL. A four‐layer theoretical model illustrating BPD predictors and BPD associations with pain severity and QoL was developed and tested using path analysis. Clarifying these relationships may enhance understanding of BPD in chronic limb pain, offering a basis for developing targeted interventions to mitigate BPD and improve treatment outcomes.

2. Method

2.1. Participants

The study included 92 patients with chronic limb pain following bone or soft tissue injury. They were referred to the pain rehabilitation services (i.e., hospital rehabilitation unit, day care unit or pain clinic) at Reuth Medical Hospital (Tel‐Aviv, Israel) and were voluntarily recruited during their rehabilitation course from June 2018 to January 2021. The inclusion criteria were age > 18 years old and pain that persisted for 3 months since the primary injury. Exclusion criteria were as follows: (1) bilateral CRPS; (2) a primary psychiatric diagnosis of depression, anxiety or post‐traumatic stress disorder (to minimise the confounding influence of pre‐existing psychiatric conditions on the current clinical state); (3) other primary pain syndromes according to the IASP classification (Nicholas et al. 2019); (4) disorders of the central nervous system according to ICD‐11 (https://icd.who.int/) (e.g., Stroke, Parkinson's disease, Multiple sclerosis, Traumatic brain injury, Spinal cord injury); (5) other diseases with sensory or inflammatory components; (6) pregnancy or breastfeeding; and (7) severe visual deficiency and intellectual disability, based on medical record and clinical observation. Sixty‐one subjects who agreed to participate were diagnosed with CRPS type 1 or 2 in the upper or lower limb, based on the Budapest clinical criteria (Harden 2010). The remaining 31 patients did not reach the threshold criterion required for a CRPS diagnosis but shared common clinical features as previously reported, though at different frequencies and severity levels (Ott and Maihöfner 2018) (See Data S1). They were defined as having non‐CRPS chronic limb pain following trauma. All participants were diagnosed by pain medical specialists.

2.2. Procedure

The study was approved by the institutional review boards of Reuth Medical Hospital (2017‐14) and Haifa University (135/18; File 1822). Participants signed written informed consent forms before being included in the study.

Participants completed four or five 1‐h research sessions conducted within 1 month. Due to high pain levels, some CRPS patients were unable to complete a full 1‐h session and therefore required a fifth meeting to complete the study. The research battery included pain hypersensitivity tests: mechanical allodynia, hyperalgesia and dynamic allodynia area in percentages (allodynograph). In addition, the study included demographic, body perception, psychological, health‐related QoL and pain‐severity self‐administered questionnaires. The study measures were selected based on their relevance and importance as core measures for chronic pain (Dworkin et al. 2005) and CRPS (Grieve et al. 2017) clinical trials. The full description of the study protocol is elaborated elsewhere (Karpin et al. 2022).

2.3. Measures

2.3.1. Pain Hypersensitivity Tests

The tests were performed on the dorsal aspect of the involved hand or foot, in the area defined as having ‘secondary hyperalgesia’ (i.e., increased pain sensitivity in non‐injured skin surrounding a site of tissue damage; Treede and Magerl 2000). This method was chosen to avoid a ceiling effect due to possible unbearable pain in the primary hyperalgesia area. The tests included the following.

2.3.1.1. Mechanical Hyperalgesia Intensity

Mechanical hyperalgesia intensity was assessed by a 256 mN pinprick stimulator (MRC Systems Pin Prick Stimulator, Heidelberg, Germany). The participants rated the pain intensity using the 0–10 Numerical Rating Scale (NRS; ‘no pain’ denoted ‘0’ and 10 denoted ‘maximal pain imaginable’). Each test was conducted 3 times; however, we considered the first test only to avoid missing data, as many patients were unable to tolerate the second (n = 50) and the third test (n = 55).

2.3.1.2. Dynamic Mechanical Allodynia Intensity

Dynamic mechanical allodynia intensity was assessed by moving a cotton stick over a 5 cm skin area for 2 s with a pressure force that was < 15 g (Packham et al. 2018). The pain intensity was rated using the 0–10 NRS. (NRS; ‘no pain’ denoted ‘0’ and 10 denoted ‘maximal pain imaginable’).

2.3.1.3. Dynamic Allodynia Area—Allodynograph

Dynamic allodynia area—allodynograph is a novel method for quantifying the allodynia area, explained in detail elsewhere (Turgeman Dahan et al. 2023). The test involved three stages: definition, delineation and measurement. First, the subject identified the painful and sensory change areas, which were marked with dots. Next, the examiner confirmed the allodynia area by touching each dot and recording pain levels above 3/10 on a numeric rating scale. Finally, the allodynia area was measured and calculated as a percentage of the affected body part using the Lund & Browder chart (Murari and Singh 2019). These indices were inserted into the following equation to calculate the allodynia area in percentages:

Allodynia area%=Length of allodynia areacm÷Length of body partcm×Standard percentages of the body part%

2.3.2. Psychological Self‐Reported Measures

2.3.2.1. Tampa Scale of Kinesiophobia

This questionnaire was developed to test fear of movement, fear of physical activity and fear avoidance (Miller et al. 1991). It includes 17 items ranging from 1 to 4 in intensity. The total score ranges from 17 to 68. Ratings are summed up to yield a total score, where higher values reflect greater fear of (re)injury.

2.3.2.2. Beck Depression Inventory

It comprises 21 symptoms and attitudes, representing first‐order depression symptom dimensions, rated from 0 to 3 in intensity (Beck et al. 1988). The items included mood, pessimism, sense of failure, lack of satisfaction, feelings of guilt, sense of punishment, self‐dislike, self‐accusation, suicidal wishes, crying, irritability, social withdrawal, indecisiveness, distortion of body image, work inhibition, sleep disturbance, fatigability, loss of appetite, weight loss, somatic preoccupation and loss of libido. The total score ranges from 0 to 63; higher scores suggest a greater severity of depression.

2.3.2.3. Cambridge Depersonalization Scale

The scale includes 29 items that measure the frequency and duration of depersonalization symptoms: unreality feelings associated with the sensory modalities, mind emptiness, inability to evoke images, qualitative changes in body image, déjà vu, micropsia, autoscopy and out‐of‐body experiences (Sierra and Berrios 2000). The scale consists of two sub‐scales—a frequency scale ranging from 0 (never) to 4 (always) and a duration scale ranging from 1 (seconds) to 6 (over a week). The total score ranges from 0 to 290, a score above 70 indicates clinically significant depersonalization.

2.3.2.4. Brief Symptom Inventory

The questionnaire consists of 53 items rated from 0 to 4 in intensity and evaluates mental distress in nine dimensions: obsessive‐compulsive, interpersonal sensitivity, depression, anxiety, hostility, somatization, phobic anxiety, paranoid ideation and psychoticism (Derogatis 1993). The overall score (General Severity Index) ranges between 0 and 4 and presents the current overall level of distress; the higher the score, the higher the distress level.

2.3.3. CRPS Severity Score

This is a quantitative index that contains 16 signs and symptoms of CRPS (Harden et al. 2010, 2017) under the sensory (Allodynia or Hyperalgesia), vasomotor (Temperature asymmetry and/or skin colour changes and/or skin colour asymmetry), sudomotor/edema (Edema and/or sweating changes and/or sweating asymmetry), and motor/trophic criteria ([weakness, tremor, dystonia], and/or trophic changes [hair/nail/skin]) (Goebel et al. 2021). The score is summed up to create an overall symptom severity index reflecting the severity of limb condition and its physical appearance. Score based on the presence or absence of signs and symptoms (coded 1/0, respectively). The total score ranges from 0 to 16; a higher CRPS severity score indicates a greater degree of CRPS symptoms.

2.3.4. Body Perception Disturbances Questionnaires

2.3.4.1. The Bath Body Perception Disturbances Questionnaire (Bath‐BPD)

The Bath‐BPD is a self‐reported measure (Halicka et al. 2020) that includes seven items regarding aspects related to the perceptiveness of the affected limb: a sense of ownership; limb position awareness; the degree of attention to the painful limb; the feelings toward it; perceptual disparities in size, temperature, pressure and weight; limb amputation desire; and a mental representation of the affected limb (Lewis et al. 2007; Lewis and McCabe 2010). The total score is based on the summation of the seven items ranging from 0 to 57 points (Lewis and McCabe 2010). A recent study (Ten Brink et al. 2021) found that removing the attention item from the total questionnaire improved its internal reliability and suggested a revised version omitting this item (r‐B‐CRPS‐BPDS) (Ten Brink et al. 2021). We followed this recommendation and removed this item, which improved the questionnaire's internal consistency (Cronbach's α = 0.67). The revised overall score in the current study ranged between 0 and 45 points.

2.3.4.2. The Neurobehavioral Questionnaire

This self‐reported BPD questionnaire (Halicka et al. 2020) was first developed by Galer and Jensen (1999). It contains five items: two items assess the presence of motor neglect symptoms, two assess the presence of cognitive neglect and a fifth item assesses the presence of involuntary movements. The expected answer is dichotomous (true or false). The internal consistency of the questionnaire was found acceptable (Cronbach's α = 0.76). Further, the test's dichotomous method was expanded to a 6‐point Likert scale (1 = never, 6 = always) to indicate the severity of the neglect symptoms (Frettlöh et al. 2006). This adapted questionnaire was used in the current study, calculated by the arithmetic mean of the five items. The total score ranges between 0 and 5 points. Testing the correlation between the revised Bath‐BPD and the Neurobehavioral questionnaire revealed a positive and significant correlation (r = 0.63, p < 0.001), indicating that both measures assess a similar, yet not identical construct (Ten Brink et al. 2021). The Bath‐BPD may reflect the more cognitive‐emotional aspects of BPD (feelings and attitudes toward the limb), and the Neurobehavioral questionnaire may reflect the more motor aspects of BPD (involuntary movement, attention efforts to move the limb and the limb as dead weight without giving it full attention). Thus, both questionnaires are used in clinical settings and present a broad perspective of BPD, which justifies their inclusion within the model.

2.3.5. Outcome Measures

2.3.5.1. Short‐Form McGill Pain Questionnaire (SF‐MPQ)

The SF‐MPQ (Melzack 1987) is a self‐report measure that assesses the multi‐dimensional aspects of pain. It involves assessing subjective pain severity using 11 sensory and four emotional descriptions on a 0 to 3 scale: no, mild, moderate and strong, respectively. The total score ranges from 0 to 45 points; a higher score represents a more severe pain experience (Melzack and Raja 2005).

2.3.5.2. The Short‐Form Health Survey Questionnaire (SF‐36)

The SF‐36 was used to assess the patients' Health‐Related Quality of Life (HRQOL) (Ware Jr. and Sherbourne 1992). It contains eight subscales, including physical functioning, role limitation due to physical health, role limitation due to emotional problems, energy and fatigue/vitality, emotional well‐being/mental health, social functioning, body pain and general health (Ware Jr. and Sherbourne 1992). The score in each subscale is converted into a 0–100 scale ranging from worst to best. Although total scoring is discouraged (Lins and Carvalho 2016), we used the total score in the current study for the following reasons: (i) We found a high and significant correlation between the mental and physical health subscales (r = 0.69, p > 0.001); (ii) we found high internal consistency between all subscales (Cronbach's α = 0.87); and (iii) research has shown there is a cross‐loading between the two final subscales, indicating an inherent overlapping between them (Farivar et al. 2007; Lewin‐Epstein et al. 1998; Taft et al. 2001). Furthermore, Anagnostopoulos et al. (2009) demonstrated that structural equation models allowing correlations between these components fit the data better than models treating them as independent (Anagnostopoulos et al. 2009).

2.4. Sample Size

We used path analysis to test the theoretical model described in the next section. As path analysis is an extension of multiple regression (Streiner 2005), when latent variables are not included, regression‐based power analysis can be used to estimate the required sample size in path models (Miles 2003). We conducted an a priori power analysis using the R 2 increase approach in G*power software (Faul et al. 2007). This method estimates the required sample size to detect whether a set of additional predictors significantly improves the explained variance (R 2) in the outcome variable, beyond an initial set of predictors (Faul et al. 2007). We tested for a medium effect (f 2 = 0.15) with α = 0.05 and power = 0.85. The model included six predictors: three exogenous variables (Hypothesized model: hyperalgesia intensity measured by a pinprick, kinesiophobia evaluated by the Tampa Scale of Kinesiophobia and depersonalization evaluated by the Cambridge Depersonalization Scale; Alternative model: allodynia area measured by Allodynograph, depression evaluated by the Beck Depression Inventory and somatization measured by the Brief Symptom Inventory/Somatization subscale) and three endogenous predictors (symptom severity, evaluated by the CRPS Severity Score, and two BPD measures: the Neurobehavioral and Bath‐BPD scores). The calculation yielded a total sample size of n = 87, which is covered by our total sample size of 92 subjects. We used eight model variables (six predictors and two outcome‐dependent variables: pain and QoL) to maintain a ratio of 10 subjects for each variable (Wolf et al. 2013). To improve model stability and reduce potential error and bias, all variables had < 10% missing values (Wolf et al. 2013). Additionally, we compared two competitive models by testing a set of model fit indices and chose the model that best corresponded to the data, as will be detailed further.

2.5. The Theoretical Model

Based on the existing literature, we generated a theoretical model that was based on 4 hypotheses, described according to the layers underlined in Figure 1a:

Higher pain hypersensitivity would be associated with (a) higher symptom severity, (b) more severe BPD (i.e., Bath and Neurobehavioral questionnaires), (c) greater pain severity and (d) lower QoL (Figure 1a , red lines).

More severe psychological factors would be associated with (a) higher symptom severity, (b) more severe BPD (i.e., Bath and Neurobehavioral scores), (c) greater pain severity and (d) lower QoL (Figure 1a , blue lines).

Higher symptom severity would be associated with (a) more severe BPD (i.e., Bath and Neurobehavioral scores), (b) greater pain severity and (c) lower QoL (Figure 1a , green lines).

Higher BPD (i.e., Bath and Neurobehavioral scores) would be associated with (a) greater pain severity and (b) lower QoL (Figure 1a , purple lines).

FIGURE 1.

FIGURE 1

Open in a new tab

(a) The theoretical model. Each model's hypotheses ([Link], [Link], [Link], [Link]) and their respective paths are presented with different colours. (b) The hypothesized model. The figure illustrates the hypothesized model, detailing the specific assessment tools used at each layer of the model. Each hypothesis ([Link], [Link], [Link], [Link]) is represented by a distinct colour, highlighting the corresponding paths within the model structure. (c) The alternative model. The figure illustrates the alternative model, detailing the specific assessment tools used at each layer of the model. Each hypothesis ([Link], [Link], [Link], [Link]) is represented by a distinct colour, highlighting the corresponding paths within the model structure.

The model included independent variables (i.e., exogenous) that can explain the variation of several dependent variables (endogenous), including mediators located in the intermediate layers (Streiner 2005). Specifically, we (i) tested the direct influence of exogenous predictors: psychological factors (depersonalization, kinesiophobia, depression and somatization) and pain hypersensitivity (hyperalgesia intensity and allodynia area), on symptom severity (CRPS severity score), BPD (Bath‐BPD and Neurobehavioral questionnaires), pain severity (SF‐MPQ total score) and QoL (SF‐36); (ii) explored the role of symptoms severity as a direct and indirect predictor (i.e., mediator) of BPD, pain severity and QoL; and (iii) explored the effect of BPD on pain severity and QoL (see Figure 1a).

Two path analyses were conducted to identify the model that best explains the relationships between psychological factors, pain hypersensitivity and BPD. The first path analysis tested the hypothesised model (Figure 1b). The model suggests that coping strategies may be associated with BPD. Empirical findings suggest that kinesiophobia was related to the Neurobehavioral questionnaire scores (Michal et al. 2017), and depersonalisation was associated with both the Bath‐BPD (Lewis et al. 2007) and the Neurobehavioral questionnaire scores (Michal et al. 2017). The model also considered the potential contribution of hyperalgesia, though empirical data is limited (Gierthmühlen et al. 2014; McCabe and Blake 2007). We hypothesised that hyperalgesia and kinesiophobia might be associated with the symptom severity scale via its sensory and motor criteria and that kinesiophobia, depersonalisation and hyperalgesia would be associated with pain severity and QoL outcomes (i.e., pain and SF‐36).

The second path analysis tested an alternative model (Figure 1c). The model suggests that psychological symptoms may be associated with BPD, specifically depression (Schulte‐Goecking et al. 2020) and somatization (Michal et al. 2017). Accordingly, we included depression and somatization as exogenous predictive variables. Likewise, we tested the dynamic allodynia area as another exogenous predictor reflecting the spatial dimension of pain hypersensitivity. All other variables and directionalities were the same as in the first hypothesised model.

2.6. Statistical Analysis

This is a secondary analysis of data previously analysed and published in the context of a different research question (Karpin et al. 2022). Demographic data were analysed using descriptive statistics and presented as mean ± SD and range for continuous variables and count and percentage for categorical variables. A chi‐squared test was conducted to compare demographics (male/female ratio) and disease characteristics between the chronic limb pain patients with and without CRPS. Additionally, t‐tests were used to compare demographic characteristics, disease duration and CRPS severity score between CRPS and non‐CRPS patients. A Pearson correlation was used to examine bivariate relationships between the Bath‐BPD and Neurobehavioral questionnaires and the study variables. The scales' internal consistency was tested using Cronbach's alpha. Hierarchical regression analysis was conducted to control for potential confounding variables, including sex, age, disease duration, affected side (right/left) and affected limb (upper/lower) (Kahlert et al. 2017). The theoretical model was tested using path analysis, which allows for testing models representing a complex set of theoretical hypotheses (West et al. 2012). Path analysis assesses the effects of a set of variables acting on specified outcomes via multiple pathways (West et al. 2012).

2.6.1. Model Verification and Fit Indices

The following indices were used to evaluate the model fit: chi‐square, which assesses the magnitude of discrepancy between the sample and fitted covariances matrices and is acceptable when the value is not significant; the Goodness of Fit Index (GFI), the Comparative Fit Index (CFI) and the Normed Fit Index (NFI) (adequate values—above 0.90, excellent fit—above 0.95). The Root Mean Square Error of Approximation (RMSEA) and the Standardised Root Mean Square Residual (SRMR) were also used; both indices decrease as model fit improves, with values starting from 0, indicating a perfect fit. Regarding RMSEA, the model closely fits the data if the lower limit of its confidence interval falls below 0.05, or alternatively, its upper limit does not exceed 0.08 or 0.10 (Kline 2015; West et al. 2012).

Furthermore, we used two additional statistical procedures for verifying the selected model: (i) the Akaike Information Criterion (AIC), which measures the expected discrepancy between the true model and the hypothesised model, and the Bayesian Information Criterion (BIC) which aims to select the model that is most likely to have generated the data in the ‘Bayesian sense’ (West et al. 2012) and (ii) model selection indices to compare between the two models and explore which models provide the best fit to the data.

Data were analysed using SPSS version 27 and AMOS 27 software, using the bootstrapping method of resampling that combines the cases in a data set in different ways to estimate statistical precision (Kline 2015). We used standardised path coefficients to estimate effect size; absolute values < 0.10 indicated a small effect, values around 0.30 indicated a medium effect and values > 0.50 indicated a large effect (Kline 2015).

2.6.2. Missing Data

All variables included in the model had < 10% missing values. We did not include missing data in the model. Therefore, any missing data was completed based on the mean imputation technique (Baraldi and Enders 2010).

The statistical significance was defined as p ≤ 0.05. Bonferroni correction was used in the case of multiple comparisons (Abdi 2007).

3. Results

3.1. Patient Characteristics

The study included 92 chronic limb pain patients: 61 CRPS and 31 non‐CRPS chronic limb pain patients. The demographics, type of primary injury, disease duration, CRPS severity score and pain level characteristics of the study sample are shown in Table 1. The CRPS and non‐CRPS patients did not significantly differ in age, years of education and disease duration (p > 0.05). The CRPS patients comprised a female/male ratio of 32/29 (52.5/47.5%), whereas among the non‐CRPS patients, there were significantly more females than males (74.2/25.8%; p = 0.044, Φ = 0.210). Both patient groups also showed similar distributions of injury aetiology (i.e., limb fracture, soft tissue damage, nerve injury involvement) and injury characteristics (Rt/Lt side, upper/lower limb).

TABLE 1.

Demographic and clinical characteristics of the study sample.

Variables CRPS (n = 61) Non‐CRPS (n = 31) p All sample (n = 92)
Age (years) 34.69 (11.52) 39.0 (15.09) 0.164 36.10 (12/91)
Sex (female/male) 32/29 (52.5/47.5) 23/8 (74.2/25.8) 0.044 55/37 (59.8/40.2)
Education (years) 12.93 (2.14) 13.71 (2.25) 0.111 13.20 (2.2)
Disease duration (days) 728.62 (903.96) 775.81 (843.25) 0.809 744.52 (879.68)
Type of injury
Fracture 31 (50.81%) 14 (45.16%) 0.739 45
Soft‐tissue trauma 27 (44.26%) 15 (48.38%) 42
Other 3 (4.91%) 2 (6.45%) 5
Limb involved
Hand 24 (39.35%) 12 (38.7%) 0.953 36
Leg 37 (60.65%) 19 (61.29%) 56
Side involved
Rt. 31 (50.81%) 12 (38.71%) 0.715 43
Lt. 30 (49.19%) 19 (61.29%) 49
Nerve injury
No 38 (62.3) 16 (51.6) 0.325 54 (58.7)
Yes 23 (37.7%) 15 (48.4%) 38 (41.3)
CRPS severity score 12.10 (2.35) 5.72 (2.11) t(85) = 12.31, p < 0.001, Cohen's d: 2.85
Open in a new tab

Note: Disease duration was defined as the time since injury associated with pain in the affected limb; The presence of nerve injury was determined by a pain specialist physician based on clinical tests and/or EMG findings. Other, inflammation, acute disease, spontaneous onset.

Abbreviation: CRPS, Complex Regional Pain Syndrome.

The pharmacological treatment included opiates, antiepileptics, antidepressants, sleep drugs, analgesics and B‐phosphonate medications. In addition, 16 participants had a medical licence for cannabis (see Table 1).

3.2. Correlation Analyses Between the Model's Variables

The results of correlation analyses are presented in Tables 2 and 3. Table 2 presents the Pearson correlations between the study's psychological variables of depression (Beck Depression Inventory), somatization (Brief Symptom Inventory), kinesiophobia (Tampa Scale of Kinesiophobia) and depersonalization (Cambridge Depersonalization Scale), BPD (Bath‐Body Perception Disturbances and Neurobehavioral questionnaires), symptoms severity (CRPS severity index), pain severity (SF‐MPQ) and QoL (The Short‐Form health survey questionnaire‐36 items). Table 3 presents the Pearson correlations between the study's pain hypersensitivity variables (hyperalgesia and allodynia area), symptoms severity, BPD, pain severity and QoL. The model's paths and directionalities were defined based on the literature and each correlation's magnitude. (See Tables 2 and 3).

TABLE 2.

Zero‐order correlations between psychological factors and body perception disturbances questionnaires.

Measure/Scale Mean (SD) Bath‐BPD (n = 91) Mental distress (n = 92) Depersonalization (n = 88) Depression (n = 90) Kinesiophobia (n = 88) SF‐36 (n = 89) SF‐McGill pain questionnaire (n = 92) CRPS severity score (n = 87)
Body perception disturbances/Neurobehavioral questionnaire 1.69 (1.18) 0.63 0.49 0.45 0.56 0.51 −0.44 0.40 0.49
Bath‐Body Perception Disturbances questionnaire 21.56 (10.93) — 0.61 0.59 0.68 0.47 −0.56 0.49 0.47
Mental distress/Brief Symptom Inventory 2.40 (0.64) — — 0.54 0.70 0.50 −0.62 0.48 0.46
Depersonalization/Cambridge Depersonalization Scale 58.91 (49.24) — — — 0.68 0.48 −0.52 0.451 0.34
Depression/Beck Depression Inventory 22.72 (11.66) — — — — 0.61 −0.76 0.54 0.51
Kinesiophobia/Tampa Scale of Kinesiophobia 39.55 (8.37) — −0.62 0.43 0.45
Quality of Life/Short‐Form‐36 item questionnaire 37.92 (17.59) — — — — — — −0.56 −0.59
Pain severity/Short‐form McGill pain questionnaire 22.03 (11.09) — — — — — — — 0.53
Open in a new tab

Note: After Bonferroni correction, the corrected α = 0.005; all correlations were significant (p < 0.005).

Abbreviation: CRPS, Complex Regional Pain Syndrome.

TABLE 3.

Zero‐order correlations between evoked and clinical pain variables, and the body perception disturbances questionnaire.

Scale Mean (SD) Bath‐BPD (n = 91) Hyperalgesia intensity (n = 85) Dynamic allodynia (n = 89) Allodynia area% (n = 88) CRPS severity score (n = 87) SF‐McGill pain questionnaire (n = 92)
Neurobehavioral questionnaire 0.63 0.40 0.47 0.29 0.49 0.40
Bath‐Body Perception Disturbances questionnaire — 0.35 0.43 0.35 0.47 0.49
Hyperalgesia intensity 5.44 (3.27) — — 0.70 0.56 0.62 0.42
Dynamic allodynia intensity 4.89 (3.45) — — 0.64 0.52 0.41
Allodynia area % 3.59 (3.54) — — — 0.53 0.45
CRPS severity score 9.95 (3.74) — — — — — 0.53
Short form‐McGill pain questionnaire 22.03 (11.09) — — — — — —
Open in a new tab

Note: After Bonferroni correction, the corrected α = 0.006; all correlations were significant (p < 0.006).

Abbreviation: CRPS, Complex Regional Pain Syndrome.

3.3. Path Analysis

Table 4 presents the differences between the two path models in fit indices and selection indices. The hypothesised model showed excellent fit indices, which were superior to those of the alternative model.

TABLE 4.

Models' comparisons: fit indices.

Hypothesized model Alternative model
AIC 70.41 86.74
BIC 156.15 172.48
χ 2 model fit index 2.41 (2), p = 0.299 18.74 (2), p = 0.000
CFI model fit index 0.999 0.957
GFI model fit index 0.993 0.952
NFI model fit index 0.993 0.955
RMSEA model fit index 90% CI 0.048 CI [0.000–0.219] 0.30 CI [0.188–0.435]
SRMR model fit index 0.019 0.048
Open in a new tab

Note: The hypothesized model included hyperalgesia, depersonalisation and kinesiophobia as exogenous predictors. The alternative model included dynamic allodynia intensity, allodynia area %, somatisation and depression as exogenous predictors.

Abbreviations: AIC, The Akaike information criterion; BIC, The Bayesian information criterion; CFI, The Comparative Fit Index; GFI, Goodness of fit index; NFI, Normed fit index; RMSEA, Root mean square error of approximation; SRMR, Standardised root mean square residual.

3.3.1. Controlling Covariates

Hierarchical regressions were conducted to control the potential contribution of demographics (e.g., age, sex) and disease characteristics (e.g., injury side; the region involved, disease duration) on the Bath‐BPD and Neurobehavioral questionnaire. The results showed that none of these variables were significant model covariates (see Table 5).

TABLE 5.

Standardised beta coefficients for demographic and disease characteristics covariates on body perception disturbance questionnaire scores.

Bath‐body perception disturbances questionnaire Neurobehavioral questionnaire
Beta weight p Beta weight p
Demographic variables
Age −0.220 0.054 −0.146 0.206
Gender 0.170 0.110 0.086 0.428
Disease characteristics
Side −0.039 0.723 −0.021 0.850
Region 0.063 0.581 −0.116 0.321
Disease duration 0.049 0.654 −0.096 0.388
Open in a new tab

Note: Disease duration was defined as the time since injury associated with pain in the affected limb.

3.3.2. Hypothesis Testing

3.3.2.1. H1: Pain Hypersensitivity, Symptom Severity, BPD, Pain Severity and QoL

( H1a ) Higher pain hypersensitivity would be associated with higher symptom severity: Hyperalgesia was directly associated with a higher CRPS severity score (β = 0.53, p < 0.001). Thus, H1a was supported.

( H1b ) Higher pain hypersensitivity would be associated with more severe BPD: Hyperalgesia intensity was not directly associated with the Neurobehavioral questionnaire scores (β = 0.13, p = 0.21) nor with the Bath‐BPD scores (β = 0.07, p = 0.49). Thus, H1b was not supported.

( H1c ) Higher pain hypersensitivity would be associated with greater pain severity and ( H1d ) lower QoL: Hyperalgesia intensity was not directly associated with higher SF‐MPQ (β = 0.12, p = 0.27) and with lower SF‐36 (β = −0.02, p = 0.80). Thus, H1c and H1d were not supported.

3.3.2.2. H2: Psychological Factors, Symptom Severity, BPD, Pain Severity and QoL

( H2a ) More severe psychological factors would be associated with higher symptom severity: Higher kinesiophobia was directly associated with higher CRPS severity score (β = 0.24, p = 0.002), thus supporting H2a.

( H2b ) More severe psychological factors would be associated with more severe BPD: Higher depersonalisation was directly associated with higher Bath‐BPD scores (β = 0.50, p < 0.001), and higher depersonalisation and kinesiophobia were directly associated with higher Neurobehavioral questionnaire scores (β = 0.24, p = 0.010 and β = 0.22, p = 0.020, respectively). Thus, H2b was supported.

( H2c ) More severe psychological factors would be associated with greater pain severity and ( H2d ) lower QoL: Higher depersonalisation was not directly associated with higher SF‐MPQ scores (β = 0.18, p = 0.087) but was associated with lower SF‐36 scores (β = −0.18, p = 0.05). Higher kinesiophobia was directly associated with lower SF‐36 scores (β = −0.33, p < 0.001), but not with higher SF‐MPQ scores (β = 0.10, p = 0.32). Thus, H2c was partially supported and H2d was supported.

3.3.2.3. H3: Symptom Severity, BPD, Pain Severity and QoL

( H3a ) Higher symptom severity would be associated with more severe BPD: Higher CRPS severity score was positively associated with higher Bath‐BPD scores (β = 0.25, p = 0.014) and higher Neurobehavioral questionnaire scores (β = 0.24, p = 0.037). Thus, H3 was supported. Moreover, there was an indirect effect of hyperalgesia intensity on the Neurobehavioral questionnaire (95% CI [−0.001, 0.288], p = 0.051) and the Bath‐BPD (95% CI [0.021, 0.27], p = 0.02) via the CRPS severity score, thus confirming its role as a significant mediator in the selected model.

( H3b ) Higher symptom severity would be associated with greater pain severity and ( H3c ) lower QoL: Higher CRPS severity score was positively associated with higher SF‐MPQ (β = 0.28, p = 0.014) and lower SF‐36 (β = −0.34, p = 0.001) scores. Thus, H3b and H3c were supported. Furthermore, there was an indirect effect of Hyperalgesia intensity on pain severity and SF‐36 scores via the CRPS severity score (95% CI [0.056, 0.315], p = 0.009 and 95% CI [−0.332, −0.078], p = 0.002, respectively), as well as an indirect effect of Kinesiophobia on pain severity and SF‐36 scores via the CRPS severity score (95% CI [0.026, 0.257], p = 0.008 and 95% CI [−0.459, −0.051], p = 0.003, respectively), thus confirming the role of CRPS severity score as a significant mediator in the selected model.

3.3.2.4. H4: BPD, Pain Severity and QoL

( H4a ) Higher BPD scores would be associated with greater pain severity and ( H4b ) lower QoL: Higher Bath‐BPD scores were not directly associated with SF‐MPQ scores (β = 0.18, p = 0.129). However, a higher Bath‐BPD score was marginally significantly associated with lower SF‐36 scores (β = −0.20, p = 0.052). Higher Neurobehavioral questionnaire scores were not associated with SF‐MPQ (β = −0.036, p = 0.749) nor SF‐36 (β = 0.099, p = 0.311) scores. Thus, H4a was not supported, and H4b was partially supported (See Figure 2).

FIGURE 2.

FIGURE 2

Open in a new tab

The final path model. Although insignificant paths were hypothesized and reported, the presented model depicted significant paths only for clarity. Depersonalization was evaluated using the Cambridge Depersonalization Scale; kinesiophobia was assessed with the Tampa Scale of Kinesiophobia; hyperalgesia was measured with a 256 mN pinprick stimulator; symptom severity was determined by the CRPS severity score; body perception disturbances were assessed using the Bath‐Bath Body Perception questionnaire and the Neurobehavioral questionnaire; pain severity was evaluated with the Short‐Form McGill pain questionnaire; and quality of life was measured by the Short‐Form Health‐Related Quality of Life 36‐Item Questionnaire. **p < 0.001, *p ≤ 0.05.

3.3.3. The Model's Predictions

The model explains 48% of the variance in the CRPS severity score, 43% of the variance in the Bath‐BPD questionnaires, 38% of the variance in the Neurobehavioral questionnaire, 39% of the variance in the SF‐MPQ and 55% of the variance in the SF‐36 scores (See Figure 2).

4. Discussion

The study aimed to enhance the understanding of BPD in chronic limb pain following trauma. Using path analysis, we examined the strength and direction of associations between pain hypersensitivity, psychological factors and BPD, as well as explored the relationship between BPD, pain severity and QoL. Our findings showed that depersonalisation directly predicted Bath‐BPD scores, while depersonalisation and kinesiophobia predicted the Neurobehavioral questionnaire score. Contrary to our expectations, hyperalgesia intensity did not directly predict BPD but had an indirect effect through the CRPS severity score, which was positively associated with both BPD measures. Finally, while the Bath‐BPD scores were not linked to pain severity, the evidence implies a link with lower QoL.

4.1. Kinesiophobia and Depersonalization Are Direct Predictors of BPD Severity

Kinesiophobia is a state of fear of movement or reinjury (French et al. 2007) and may manifest in fear avoidance behaviour (Leeuw et al. 2007). In the acute pain stage, it can be a valuable strategy to decrease aversive stimuli and threats. However, in the chronic stages, avoidance leads to an increased threat perception, fear learning (Linton 2013) and higher pain intensity (Bean et al. 2014). Structural changes in the basal ganglia, particularly in the putamen, were found to correlate with decreased motor control in chronic CRPS, potentially underpinning fear‐avoidance behaviour (Azqueta‐Gavaldon et al. 2020). Accordingly, kinesiophobia has been linked to decreased upper limb range of motion (Duport et al. 2022) and proprioceptive errors in the painful region (Asiri et al. 2021). Since body representation is a dynamic construct that relies on continuous sensory input from multiple modalities, reduced somatosensory input may impair this updating process, leading to increased reliance on top‐down brain prediction (Pitron and de Vignemont 2017). Disruptions in these processes have been linked to body misperception in CRPS (Kuttikat et al. 2016). In line with this, our finding that kinesiophobia predicted the Neurobehavioral questionnaire score suggests a possible contribution of movement avoidance to the motor manifestation of BPD (Galer and Jensen 1999).

Previous studies have also linked kinesiophobia to Bath‐BPD severity, either through correlation in a small sample (n = 22; Mibu et al. 2021), or as a predictor explaining only 9% of the variance (Ten Brink et al. 2021), suggesting that additional unmeasured factors may contribute. In the current study, the Bath‐BPD was predicted by depersonalization, with a large effect size, suggesting that dissociative processes may play a more central role in explaining Bath‐ BPD. This findings may reflect the use of an advanced pathological coping strategy of suppression (Linton 2013). Suppression is an active ‘last defence’ strategy (Eccleston 2018; Tabor et al. 2017) involving the inhibition of emotion and pain (Linton 2013), manifested in disembodiment (Morse 2015) and dissociation (Sierra and David 2011; Tabor et al. 2017). Depersonalization is not just a coping strategy for distancing the dysfunctional limb (Sündermann et al. 2018); it may serve as a defence process from inescapable pain (Eccleston 2018; Morse 2015), operating to protect the self by paradoxically sacrificing agency (Eccleston 2018). Using qualitative methodology, dissociation is acknowledged as a prominent theme in CRPS patients' experiences as a coping strategy to ‘get rid of the pain’ (Lewis et al. 2007).

4.2. Hyperalgesia Is Not a Direct Predictor of BPD Severity

Mechanical hyperalgesia measured in the secondary hyperalgesia area is a clinical manifestation of central sensitisation processes (Latremoliere and Woolf 2009). It is reflected in the CRPS severity score as hyperalgesia to pinprick and allodynia (Harden 2010). Therefore, our findings that hyperalgesia is a direct predictor of this scale are not surprising.

The interesting finding is that hyperalgesia intensity was not directly associated with BPD severity, contrary to our hypothesis. An essential function of pain is to protect the body from actual or potential damage (Haggard et al. 2013). However, when the pain becomes chronic (i.e., > 3 months), it is influenced by central sensitisation processes and is less dependent on the presence, intensity or duration of specific peripheral noxious input than in acute pain (Latremoliere and Woolf 2009; Nijs et al. 2012). Thus, the lack of association between centrally‐mediated hyperalgesia and BPD suggests that maladaptive brain plasticity beyond the pain transmission pathways may contribute to BPD through their involvement in spatial attention and sensory‐motor perception processes (Gierthmühlen et al. 2014; Lee et al. 2022).

4.3. The Association Between Symptom Severity and BPD

The CRPS severity score is an index consisting of sensory, vasomotor, sudomotor and motor trophic signs and symptoms, which all may relate to and affect BPD. In detail, temperature asymmetry has been shown to affect BPD both in healthy participants (Kammers et al. 2011) and CRPS patients (Moseley et al. 2013, 2012). Findings indicated that cold hands decrease the sense of limb ownership while warming the hands increases it (Kammers et al. 2011). In CRPS, this effect was spatially based; the affected hand became warmer when it was crossed with the non‐affected hand over the body's midline (Moseley et al. 2013, 2012), highlighting the influence of limb position on body ownership (Mangalam et al. 2019). Moreover, the CRPS severity score includes sudomotor signs such as edema, sweating and evidence of trophic changes in hair, nails and skin. All are visible and can negatively influence body image, which relies more on visual than sensory‐motor input (Pitron et al. 2018). These signs and symptoms may negatively affect body appearance and functionality, which suggests promoting BPD (Sündermann et al. 2020). Additionally, the motor symptoms, including a decreased range of motion, tremor and weakness, may impair goal‐directed movement, which was found to contribute to intact body perception (Wen et al. 2016) and sense of agency (Tsakiris et al. 2010, 2007).

To conclude, findings suggest that multiple factors may contribute to BPD in chronic limb pain following a tissue injury. Therefore, clinical interventions to alleviate sensory, vasomotor, sudomotor and motor symptoms may improve BPD. For example, limb activation (strength, range of motion and functional task), sensory desensitisation (Kotsougiani‐Fischer et al. 2020) or multisensory training (Batalla and Lewis 2024) could be beneficial and are suggested for testing in future studies.

4.4. The Relationship Between BPD, Pain Severity and QoL

Contrary to our hypothesis, we found that BPD questionnaire scores did not predict pain severity. At first glance, these results contradict previous studies that reported an association between BPD and pain severity based on correlation (Schulte‐Goecking et al. 2020) or regression analyses (Ten Brink et al. 2021). Interestingly, based on correlations, our findings also showed significant moderate associations between the BPD scores and pain severity. Notably, path analysis enables the simultaneous estimation of direct and indirect relationships among variables within a theoretically specified model (Kline 2015). In the current study, the path model did not support a direct relationship between BPD and pain severity. Thus, both our findings and existing literature suggest complex relationships between BPD and pain, highlighting the need for further empirical research to understand this association better.

Finally, we found support for a negative association of the Bath‐BPD on SF‐36 scores. This is supported by a former study (Schulte‐Goecking et al. 2020) and a theoretical model regarding body image and chronic pain (Sündermann et al. 2020). The latter acknowledged the mutual multilayer association between chronic pain, body image disturbances and maladaptive coping strategies and behaviour (Sündermann et al. 2020), all leading to reduced QoL. Findings further showed that kinesiophobia and depersonalisation also negatively affected QoL. These clinical behaviours may reduce interpersonal communication and intimacy (de Haan and Dijkerman 2020), contributing to lower QoL.

4.5. Study Limitations

This study has several limitations. First, the assessment of BPD was based on self‐reported questionnaires without the inclusion of other tests, such as laterality recognition or finger identification (Acapo et al. 2022), which have been associated with BPD in chronic limb pain. Second, we did not test the patients' proprioception sensation pre‐entry, which might affect their BPD. Third, upper and lower limb functional tests were not included as outcome measures, which limits the model's implications in these aspects. Finally, primary depression patients were excluded to reduce potential bias, as depression could confound the clinical state. However, their absence may have affected the alternative model, which included depression as an exogenous variable.

5. Conclusions and Recommendations for Future Research

To our knowledge, this is the first study to apply path analysis to examine BPD predictors and BPD effects on pain and QoL in post‐injury chronic limb pain patients. The findings highlight the contribution of psychological coping strategies, specifically depersonalisation and kinesiophobia, to BPD compared to pain hypersensitivity factors.

The presented model offers a preliminary framework that warrants further empirical and clinical validation. From an empirical perspective, longitudinal studies involving multiple time points are recommended to assess and verify the suggested relationships between predictors, mediators and outcome variables. Additionally, clinical trials are recommended to test whether interventions targeting depersonalisation and kinesiophobia, identified in the model as key contributors to BPD, can effectively reduce symptoms of BPD.

The study also raises questions regarding the association between BPD and pain severity in chronic stages, indicating a need for further research to elucidate these relationships.

Author Contributions

H.K., J.‐J.V. and I.W.‐F. designed this study. The experiments were performed by H.K., A.L. and J.‐J.V. The data were analysed by H.K. and I.W.‐F., and the results were critically examined by all authors. H.K. had a primary role in preparing the manuscript, which was edited by I.W.‐F. All authors have approved the final version of the manuscript and agree to be accountable for all aspects of the work.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: ejp70123‐sup‐0001‐DataS1.docx.

EJP-29-0-s001.docx (487KB, docx)

Acknowledgements

We thank Noy Turgeman‐Dahan, Sharon Shmueli and Adi Atuan for their help with data collection for this article.

Karpin, H. , Vatine J.‐J., Livshitz A., and Weissman‐Fogel I.. 2025. “A Model of Body Perception Disturbances in Chronic Limb Pain: The Predictive Role of Kinesiophobia, Depersonalization and Symptom Severity.” European Journal of Pain 29, no. 9: e70123. 10.1002/ejp.70123.

Funding: This research was funded by the Reuth Rehabilitation Hospital Internal Research Grant Program, Research grant no. 2017‐14.

Data Availability Statement

The data supporting this study's findings are available from the corresponding author upon reasonable request.

References

  1. Abdi, H. 2007. “Bonferroni and Šidák Corrections for Multiple Comparisons.” Encyclopedia of Measurement and Statistics 3, no. 1: 2007. [Google Scholar]
  2. Acapo, S. , Osinski T., Rulleau T., Dupeyron A., and Nizard J.. 2022. “Assessment of Body Perception Disturbances in Complex Regional Pain Syndrome: A Systematic Review Using the COSMIN Guideline.” European Journal of Pain 26, no. 10: 2060–2073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Anagnostopoulos, F. , Niakas D., and Tountas Y.. 2009. “Comparison Between Exploratory Factor‐Analytic and SEM‐Based Approaches to Constructing SF‐36 Summary Scores.” Quality of Life Research 18, no. 1: 53–63. [DOI] [PubMed] [Google Scholar]
  4. Asiri, F. , Reddy R. S., Tedla J. S., et al. 2021. “Kinesiophobia and Its Correlations With Pain, Proprioception, and Functional Performance Among Individuals With Chronic Neck Pain.” PLoS One 16, no. 7: e0254262. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Azqueta‐Gavaldon, M. , Youssef A. M., Storz C., et al. 2020. “Implications of the Putamen in Pain and Motor Deficits in Complex Regional Pain Syndrome.” Pain 161, no. 3: 595–608. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Baraldi, A. N. , and Enders C. K.. 2010. “An Introduction to Modern Missing Data Analyses.” Journal of School Psychology 48, no. 1: 5–37. [DOI] [PubMed] [Google Scholar]
  7. Batalla, M. A. P. , and Lewis J. S.. 2024. “Cognitive Multisensory Rehabilitation, a Novel Approach for Complex Regional Pain Syndrome: Case Series.” Physiotherapy Theory and Practice 41, no. 5: 1–1123. [DOI] [PubMed] [Google Scholar]
  8. Bean, D. J. , Johnson M. H., Heiss‐Dunlop W., Lee A. C., and Kydd R. R.. 2015. “Do Psychological Factors Influence Recovery From Complex Regional Pain Syndrome Type 1? A Prospective Study.” Pain 156, no. 11: 2310–2318. [DOI] [PubMed] [Google Scholar]
  9. Bean, D. J. , Johnson M. H., and Kydd R. R.. 2014. “Relationships Between Psychological Factors, Pain, and Disability in Complex Regional Pain Syndrome and Low Back Pain.” Clinical Journal of Pain 30, no. 8: 647–653. 10.1097/ajp.0000000000000007. [DOI] [PubMed] [Google Scholar]
  10. Beck, A. T. , Steer R. A., and Carbin M. G.. 1988. “Psychometric Properties of the Beck Depression Inventory: Twenty‐Five Years of Evaluation.” Clinical Psychology Review 8, no. 1: 77–100. [Google Scholar]
  11. de Haan, E. H. F. , and Dijkerman H. C.. 2020. “Somatosensation in the Brain: A Theoretical re‐Evaluation and a New Model.” Trends in Cognitive Sciences 24, no. 7: 529–541. 10.1016/j.tics.2020.04.003. [DOI] [PubMed] [Google Scholar]
  12. Derogatis, L. R. 1993. “Brief Symptom Inventory.” European Journal of Psychological Assessment.
  13. Duport, A. , Morel P., Léonard G., and Devanne H.. 2022. “The Influence of Pain and Kinesiophobia on Motor Control of the Upper Limb: How Pointing Task Paradigms Can Point to New Avenues of Understanding.” Pain 165: 2044–2054. 10.1097/j.pain.0000000000003213. [DOI] [PubMed] [Google Scholar]
  14. Dworkin, R. H. , Turk D. C., Farrar J. T., et al. 2005. “Core Outcome Measures for Chronic Pain Clinical Trials: IMMPACT Recommendations.” Pain 113, no. 1: 9–19. [DOI] [PubMed] [Google Scholar]
  15. Eccleston, C. 2018. “Chronic Pain as Embodied Defence: Implications for Current and Future Psychological Treatments.” Pain 159: S17–S23. 10.1097/j.pain.0000000000001286. [DOI] [PubMed] [Google Scholar]
  16. Farivar, S. S. , Cunningham W. E., and Hays R. D.. 2007. “Correlated Physical and Mental Health Summary Scores for the SF‐36 and SF‐12 Health Survey, V.I.” Health and Quality of Life Outcomes 5, no. 1: 1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Faul, F. , Erdfelder E., Lang A.‐G., and Buchner A.. 2007. “G* Power 3: A Flexible Statistical Power Analysis Program for the Social, Behavioral, and Biomedical Sciences.” Behavior Research Methods 39, no. 2: 175–191. [DOI] [PubMed] [Google Scholar]
  18. French, D. J. , France C. R., Vigneau F., French J. A., and Evans R. T.. 2007. “Fear of Movement/(Re)Injury in Chronic Pain: A Psychometric Assessment of the Original English Version of the Tampa Scale for Kinesiophobia (TSK).” Pain 127, no. 1–2: 42–51. [DOI] [PubMed] [Google Scholar]
  19. Frettlöh, J. , Hüppe M., and Maier C.. 2006. “Severity and Specificity of Neglect‐Like Symptoms in Patients With Complex Regional Pain Syndrome (CRPS) Compared to Chronic Limb Pain of Other Origins.” Pain 124, no. 1–2: 184–189. [DOI] [PubMed] [Google Scholar]
  20. Galer, B. S. , and Jensen M.. 1999. “Neglect‐Like Symptoms in Complex Regional Pain Syndrome: Results of a Self‐Administered Survey.” Journal of Pain and Symptom Management 18, no. 3: 213–217. [DOI] [PubMed] [Google Scholar]
  21. Gierthmühlen, J. , Binder A., and Baron R.. 2014. “Mechanism‐Based Treatment in Complex Regional Pain Syndromes.” Nature Reviews Neurology 10, no. 9: 518–528. [DOI] [PubMed] [Google Scholar]
  22. Goebel, A. , Birklein F., Brunner F., et al. 2021. “The Valencia Consensus‐Based Adaptation of the IASP Complex Regional Pain Syndrome Diagnostic Criteria.” Pain 162, no. 9: 2346–2348. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Grieve, S. , Perez R. S., Birklein F., et al. 2017. “Recommendations for a First Core Outcome Measurement Set for Complex Regional PAin Syndrome Clinical sTudies (COMPACT).” Pain 158, no. 6: 1083–1090. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Haggard, P. , Iannetti G. D., and Longo M. R.. 2013. “Spatial Sensory Organization and Body Representation in Pain Perception.” Current Biology 23, no. 4: R164–R176. [DOI] [PubMed] [Google Scholar]
  25. Halicka, M. , Vittersø A. D., Proulx M. J., and Bultitude J. H.. 2020. “Neuropsychological Changes in Complex Regional Pain Syndrome (CRPS).” Behavioural Neurology 2020: 1–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hall, J. , Llewellyn A., Palmer S., Rowett‐Harris J., Atkins R. M., and McCabe C. S.. 2016. “Sensorimotor Dysfunction After Limb Fracture—An Exploratory Study.” European Journal of Pain 20, no. 9: 1402–1412. [DOI] [PubMed] [Google Scholar]
  27. Harden, R. N. 2010. “Objectification of the Diagnostic Criteria for CRPS.” Pain Medicine 11, no. 8: 1212–1215. [DOI] [PubMed] [Google Scholar]
  28. Harden, R. N. , Bruehl S., Perez R. S., et al. 2010. “Development of a Severity Score for CRPS.” Pain 151, no. 3: 870–876. [DOI] [PubMed] [Google Scholar]
  29. Harden, R. N. , Maihofner C., Abousaad E., et al. 2017. “A Prospective, Multisite, International Validation of the Complex Regional Pain Syndrome Severity Score.” Pain 158, no. 8: 1430–1436. [DOI] [PubMed] [Google Scholar]
  30. Harden, R. N. , McCabe C. S., Goebel A., et al. 2022. “Complex Regional Pain Syndrome: Practical Diagnostic and Treatment Guidelines.” Pain Medicine 23, no. Supplement_1: S1–S53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kahlert, J. , Gribsholt S. B., Gammelager H., Dekkers O. M., and Luta G.. 2017. “Control of Confounding in the Analysis Phase–an Overview for Clinicians.” Clinical Epidemiology 9: 195–204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kammers, M. P. , Rose K., and Haggard P.. 2011. “Feeling Numb: Temperature, but Not Thermal Pain, Modulates Feeling of Body Ownership.” Neuropsychologia 49, no. 5: 1316–1321. [DOI] [PubMed] [Google Scholar]
  33. Karpin, H. , Vatine J. J., Bachar Kirshenboim Y., Markezana A., and Weissman‐Fogel I.. 2022. “Central Sensitization and Psychological State Distinguishing Complex Regional Pain Syndrome From Other Chronic Limb Pain Conditions: A Cluster Analysis Model.” Biomedicine 11, no. 1: 89. 10.3390/biomedicines11010089. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kline, R. B. 2015. Principles and Practice of Structural Equation Modeling. Guilford Publications. [Google Scholar]
  35. Kotsougiani‐Fischer, D. , Choi J., Oh‐Fischer J., et al. 2020. “ICF‐Based Multidisciplinary Rehabilitation Program for Complex Regional Pain Syndrome of the Hand: Efficacy, Long‐Term Outcomes, and Impact of Therapy Duration.” BMC Surgery 20: 1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Kuttikat, A. , Noreika V., Shenker N., Chennu S., Bekinschtein T., and Brown C. A.. 2016. “Neurocognitive and Neuroplastic Mechanisms of Novel Clinical Signs in CRPS.” Frontiers in Human Neuroscience 10: 16. 10.3389/fnhum.2016.00016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Latremoliere, A. , and Woolf C. J.. 2009. “Central Sensitization: A Generator of Pain Hypersensitivity by Central Neural Plasticity.” Journal of Pain 10, no. 9: 895–926. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Lee, B. , Di Pietro F., Henderson L. A., and Austin P. J.. 2022. “Altered Basal Ganglia Infraslow Oscillation and Resting Functional Connectivity in Complex Regional Pain Syndrome.” Journal of Neuroscience Research 100, no. 7: 1487–1505. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Leeuw, M. , Goossens M. E., Linton S. J., Crombez G., Boersma K., and Vlaeyen J. W.. 2007. “The Fear‐Avoidance Model of Musculoskeletal Pain: Current State of Scientific Evidence.” Journal of Behavioral Medicine 30: 77–94. [DOI] [PubMed] [Google Scholar]
  40. Lewin‐Epstein, N. , Sagiv‐Schifter T., Shabtai E. L., and Shmueli A.. 1998. “Validation of the 36‐Item Short‐Form Health Survey (Hebrew Version) in the Adult Population of Israel.” Medical Care 36: 1361–1370. [DOI] [PubMed] [Google Scholar]
  41. Lewis, J. , and McCabe C.. 2010. “Body Perception Disturbance (BPD) in CRPS.” In Practical Pain Management, 60–66. PPM Communications, Inc. [Google Scholar]
  42. Lewis, J. , and Schweinhardt P.. 2012. “Perceptions of the Painful Body: The Relationship Between Body Perception Disturbance, Pain and Tactile Discrimination in Complex Regional Pain Syndrome.” European Journal of Pain 16, no. 9: 1320–1330. [DOI] [PubMed] [Google Scholar]
  43. Lewis, J. S. , Kersten P., McCabe C. S., McPherson K. M., and Blake D. R.. 2007. “Body Perception Disturbance: A Contribution to Pain in Complex Regional Pain Syndrome (CRPS).” Pain 133, no. 1–3: 111–119. [DOI] [PubMed] [Google Scholar]
  44. Lins, L. , and Carvalho F. M.. 2016. “SF‐36 Total Score as a Single Measure of Health‐Related Quality of Life: Scoping Review.” SAGE Open Medicine 4: 1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Linton, S. J. 2013. “A Transdiagnostic Approach to Pain and Emotion.” Journal of Applied Biobehavioral Research 18, no. 2: 82–103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Mangalam, M. , Cutts S. A., and Fragaszy D. M.. 2019. “Sense of Ownership and Not the Sense of Agency Is Spatially Bounded Within the Space Reachable With the Unaugmented Hand.” Experimental Brain Research 237: 2911–2924. [DOI] [PubMed] [Google Scholar]
  47. McCabe, C. S. , and Blake D. R.. 2007. “Evidence for a Mismatch Between the Brain's Movement Control System and Sensory System as an Explanation for Some Pain‐Related Disorders.” Current Pain and Headache Reports 11: 104–108. [DOI] [PubMed] [Google Scholar]
  48. Melzack, R. 1987. “The Short‐Form McGill Pain Questionnaire.” Pain 30, no. 2: 191–197. [DOI] [PubMed] [Google Scholar]
  49. Melzack, R. , and Raja S. N.. 2005. “The McGill Pain Questionnaire: From Description to Measurement.” Anesthesiology 103, no. 1: 199–202. [DOI] [PubMed] [Google Scholar]
  50. Mibu, A. , Nishigami T., Uematsu H., et al. 2021. “Validation of the Japanese Version of the Bath CRPS Body Perception Disturbance Scale for CRPS.” Journal of Anesthesia 35: 20–26. [DOI] [PubMed] [Google Scholar]
  51. Michal, M. , Adler J., Reiner I., et al. 2017. “Association of Neglect‐Like Symptoms With Anxiety, Somatization, and Depersonalization in Complex Regional Pain Syndrome.” Pain Medicine 18, no. 4: 764–772. [DOI] [PubMed] [Google Scholar]
  52. Miles, J. 2003. “A Framework for Power Analysis Using a Structural Equation Modelling Procedure.” BMC Medical Research Methodology 3, no. 1: 27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Miller, R. P. , Kori S. H., and Todd D. D.. 1991. “The Tampa Scale: A Measure of Kinisophobia.” Clinical Journal of Pain 7, no. 1: 51. [Google Scholar]
  54. Morse, J. M. 2015. “Using Qualitative Methods to Access the Pain Experience.” British Journal of Pain 9, no. 1: 26–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Moseley, G. L. , Gallace A., Di Pietro F., Spence C., and Iannetti G. D.. 2013. “Limb‐Specific Autonomic Dysfunction in Complex Regional Pain Syndrome Modulated by Wearing Prism Glasses.” Pain 154, no. 11: 2463–2468. [DOI] [PubMed] [Google Scholar]
  56. Moseley, G. L. , Gallace A., and Iannetti G. D.. 2012. “Spatially Defined Modulation of Skin Temperature and Hand Ownership of Both Hands in Patients With Unilateral Complex Regional Pain Syndrome.” Brain 135, no. 12: 3676–3686. [DOI] [PubMed] [Google Scholar]
  57. Murari, A. , and Singh K. N.. 2019. “Lund and Browder Chart‐Modified Versus Original: A Comparative Study.” Acute and Critical Care 34, no. 4: 276–281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Nicholas, M. , Vlaeyen J. W. S., Rief W., et al. 2019. “The IASP Classification of Chronic Pain for ICD‐11: Chronic Primary Pain.” Pain 160, no. 1: 28–37. 10.1097/j.pain.0000000000001390. [DOI] [PubMed] [Google Scholar]
  59. Nijs, J. , Daenen L., Cras P., Struyf F., Roussel N., and Oostendorp R. A.. 2012. “Nociception Affects Motor Output: A Review on Sensory‐Motor Interaction With Focus on Clinical Implications.” Clinical Journal of Pain 28, no. 2: 175–181. [DOI] [PubMed] [Google Scholar]
  60. Ott, S. , and Maihöfner C.. 2018. “Signs and Symptoms in 1,043 Patients With Complex Regional Pain Syndrome.” Journal of Pain 19, no. 6: 599–611. [DOI] [PubMed] [Google Scholar]
  61. Packham, T. L. , Spicher C. J., MacDermid J. C., Michlovitz S., and Buckley D. N.. 2018. “Somatosensory Rehabilitation for Allodynia in Complex Regional Pain Syndrome of the Upper Limb: A Retrospective Cohort Study.” Journal of Hand Therapy 31, no. 1: 10–19. [DOI] [PubMed] [Google Scholar]
  62. Pitron, V. , Alsmith A., and de Vignemont F.. 2018. “How Do the Body Schema and the Body Image Interact?” Consciousness and Cognition 65: 352–358. [DOI] [PubMed] [Google Scholar]
  63. Pitron, V. , and de Vignemont F.. 2017. “Beyond Differences Between the Body Schema and the Body Image: Insights From Body Hallucinations.” Consciousness and Cognition 53: 115–121. [DOI] [PubMed] [Google Scholar]
  64. Schulte‐Goecking, H. , Azqueta‐Gavaldon M., Storz C., et al. 2020. “Psychological, Social and Biological Correlates of Body Perception Disturbance in Complex Regional Pain Syndrome.” Current Psychology 41: 1–11. [Google Scholar]
  65. Schütz‐Bosbach, S. , Musil J. J., and Haggard P.. 2009. “Touchant‐Touché: The Role of Self‐Touch in the Representation of Body Structure.” Consciousness and Cognition 18, no. 1: 2–11. [DOI] [PubMed] [Google Scholar]
  66. Shaikh, M. 2017. “11. Autotomy Wish in Complex Regional Pain Syndrome: An Integral Part of the Condition?” Rheumatology Advances in Practice 1, no. suppl_1: rkx014. [Google Scholar]
  67. Sierra, M. , and Berrios G. E.. 2000. “The Cambridge Depersonalisation Scale: A New Instrument for the Measurement of Depersonalisation.” Psychiatry Research 93, no. 2: 153–164. [DOI] [PubMed] [Google Scholar]
  68. Sierra, M. , and David A. S.. 2011. “Depersonalization: A Selective Impairment of Self‐Awareness.” Consciousness and Cognition 20, no. 1: 99–108. [DOI] [PubMed] [Google Scholar]
  69. Streiner, D. L. 2005. “Finding Our Way: An Introduction to Path Analysis.” Canadian Journal of Psychiatry 50, no. 2: 115–122. [DOI] [PubMed] [Google Scholar]
  70. Sündermann, O. , Flink I., and Linton S. J.. 2020. “My Body Is Not Working Right: A Cognitive Behavioral Model of Body Image and Chronic Pain.” Pain 161, no. 6: 1136–1139. [DOI] [PubMed] [Google Scholar]
  71. Sündermann, O. , Rydberg K., Linder L., and Linton S. J.. 2018. “‘When I Feel the Worst Pain, I Look Like Shit’—Body Image Concerns in Persistent Pain.” Scandinavian Journal of Pain 18, no. 3: 379–388. [DOI] [PubMed] [Google Scholar]
  72. Tabor, A. , Keogh E., and Eccleston C.. 2017. “Embodied Pain—Negotiating the Boundaries of Possible Action.” Pain 158, no. 6: 1007–1011. [DOI] [PubMed] [Google Scholar]
  73. Taft, C. , Karlsson J., and Sullivan M.. 2001. “Do SF‐36 Summary Component Scores Accurately Summarize Subscale Scores?” Quality of Life Research 10, no. 5: 395–404. [DOI] [PubMed] [Google Scholar]
  74. Ten Brink, A. F. , and Bultitude J. H.. 2021. “Predictors of Self‐Reported Neglect‐Like Symptoms and Involuntary Movements in Complex Regional Pain Syndrome Compared to Other Chronic Limb Pain Conditions.” Pain Medicine 22, no. 10: 2337–2349. 10.1093/pm/pnab226. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Ten Brink, A. F. , Halicka M., Vittersø A. D., Jones H. G., Stanton T. R., and Bultitude J. H.. 2021. “Validation of the Bath CRPS Body Perception Disturbance Scale.” Journal of Pain 22, no. 11: 1371–1384. [DOI] [PubMed] [Google Scholar]
  76. Treede, R.‐D. , and Magerl W.. 2000. “Multiple Mechanisms of Secondary Hyperalgesia.” Progress in Brain Research 129: 331–341. [DOI] [PubMed] [Google Scholar]
  77. Tsakiris, M. , Longo M. R., and Haggard P.. 2010. “Having a Body Versus Moving Your Body: Neural Signatures of Agency and Body‐Ownership.” Neuropsychologia 48, no. 9: 2740–2749. [DOI] [PubMed] [Google Scholar]
  78. Tsakiris, M. , Schütz‐Bosbach S., and Gallagher S.. 2007. “On Agency and Body‐Ownership: Phenomenological and Neurocognitive Reflections.” Consciousness and Cognition 16, no. 3: 645–660. [DOI] [PubMed] [Google Scholar]
  79. Turgeman Dahan, N. , Vatine J.‐J., Weissman‐Fogel I., Karpin H., Shmuely S., and Bar‐Shalita T.. 2023. “Quantitative Dynamic Allodynograph—A Standardized Measure for Testing Dynamic Mechanical Allodynia in Chronic Limb Pain.” Sensors 23, no. 18: 7949. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. van Rooijen, D. E. , Marinus J., Schouten A. C., Noldus L. P., and van Hilten J. J.. 2013. “Muscle Hyperalgesia Correlates With Motor Function in Complex Regional Pain Syndrome Type 1.” Journal of Pain 14, no. 5: 446–454. [DOI] [PubMed] [Google Scholar]
  81. Ware, J. E., Jr. , and Sherbourne C. D.. 1992. “The MOS 36‐Item Short‐Form Health Survey (SF‐36): I. Conceptual Framework and Item Selection.” Medical Care 30: 473–483. [PubMed] [Google Scholar]
  82. Wen, W. , Muramatsu K., Hamasaki S., et al. 2016. “Goal‐Directed Movement Enhances Body Representation Updating.” Frontiers in Human Neuroscience 10: 329. [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. West, S. G. , Taylor A. B., and Wu W.. 2012. “Model Fit and Model Selection in Structural Equation Modeling.” In Handbook of Structural Equation Modeling, edited by Hoyle R. H., vol. 1, 209–231. Guilford Press. [Google Scholar]
  84. Wittayer, M. , Dimova V., Birklein F., and Schlereth T.. 2018. “Correlates and Importance of Neglect‐Like Symptoms in Complex Regional Pain Syndrome.” Pain 159, no. 5: 978–986. [DOI] [PubMed] [Google Scholar]
  85. Wolf, E. J. , Harrington K. M., Clark S. L., and Miller M. W.. 2013. “Sample Size Requirements for Structural Equation Models: An Evaluation of Power, Bias, and Solution Propriety.” Educational and Psychological Measurement 73, no. 6: 913–934. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1: ejp70123‐sup‐0001‐DataS1.docx.

EJP-29-0-s001.docx (487KB, docx)

Data Availability Statement

The data supporting this study's findings are available from the corresponding author upon reasonable request.

Articles from European Journal of Pain (London, England) are provided here courtesy of Wiley

RESOURCES