What's known on the subject? and What does the study add?
Interstitial cystitis/bladder pain syndrome (IC/BPS) is a disease that is associated with significant disability in the areas of self-care, sexual functioning, occupation, family and home responsibilities, and social functions. Compared to age- and cohort-matched controls, patients with IC/BPS reported poorer mental and physical quality of life (QoL), as well as greater pain, depression, anxiety, catastrophizing, sexual dysfunction and less social support. In particular, catastrophizing and depression are suggested to be important for understanding the factors promoting a poorer QoL in IC/BPS.
The findings of the present study suggest that psychosocial factors are significant in mediating the relationship between impairments and patient disability, with negative affect (i.e. depression, anxiety) and pain catastrophizing acting as significant mediators. These findings argue that catastrophizing and negative affect may be the mechanisms by which pain and symptoms elevate reports of patient disability. Understanding how pain and pain-associated symptoms may become disabling through cognitive mechanisms represents an importance advance for IC/BPS management. Clinical interventions in women suffering from IC/BPS should target catastrophizing and mood for reduction using cognitive behavioural strategies aiming to decrease pain-related disability.
To examine a biopsychosocial framework of patient disability in patients suffering from interstitial cystitis/bladder pain syndrome (IC/BPS).
To evaluate the impact of psychosocial factors on the relationship between pain and disability within women with IC/BPS.
Patients and Methods
Women with IC/BPS completed questionnaires including demographics, symptoms and problems (IC Symptom and Problem Indices), pain (McGill Pain Questionnaire), quality of life (Medical Outcomes Study – Short Form 12), disability (Pain Disability Index) and psychosocial variables (Center for Epidemiological Studies – Depression Scale; State Trait Anxiety Inventory; Pain Catastrophizing Scale; Multidimensional Scale of Perceived Social Support).
Exploratory and confirmatory factor analyses were employed to determine the factor structure and composition of the measured variables.
Structural equation modelling was used to examine model fit and the mediation effect of the psychosocial factors (negative affect, catastrophizing and social support) on impairments and functional disability.
Questionnaires completed by 196 women with IC/BPS provided data for the present study.
The measurement model showed good fit to the data.
Negative affect (P < 0.001) and catastrophizing (P < 0.001) were significant in explaining the relationship between impairments and functional disability, whereas social support did not.
Disability in patients suffering from IC/BPS is partially explained by the impact of negative affect and catastrophizing.
As a result of the refractory nature of IC/BPS, patient management within a biopsychosocial framework represents an essential area of investigation.
Decreases in negative affect and catastrophizing will probably lead to improvements in pain-related disability.
Center for Epidemiological Studies Depression Scale
confirmatory factor analysis
comparitive fit index
disablement process model
exploratory factor analysis
interstitial cystitis/bladder pain syndrome
Multidimensional Scale of Perceived Social Support
Pain Catastrophizing Scale
quality of life
root mean square error of approximation
structural equation modelling
IC/BPS is a disease that is associated with significant disability in the areas of self-care, sexual functioning, occupation, family and home responsibilities, and social functions. In comparison to other chronic health conditions, the costs associated with IC/BPS are equal to or greater than low back pain, fibromyalgia, rheumatoid arthritis and peripheral neuropathy . IC/BPS disability has a major economic impact in terms of costs to the medical system (e.g. inpatient/outpatient care, medications), costs to society (e.g. lost work productivity, unemployment) and diminished patient quality of life (QoL) . Research comparing patients with IC/BPS to age- and cohort-matched controls not only reported poorer mental and physical QoL for patients, but also highlighted that comorbid QoL issues for patients (e.g. greater pain, depression, anxiety, catastrophizing, sexual dysfunction and less social support) are of concern . In particular, catastrophizing and depression are suggested to be important for understanding the factors promoting a poorer QoL in IC/BPS [4, 5].
Although assessments of economic and functional disability are valuable, the mechanisms underlying reports of disability have not been empirically examined in IC/BPS. The disablement process model (DPM) uses a biopsychosocial framework in which psychosocial factors affecting the relationship between impairments (i.e. pain and symptoms) and disability (i.e. the primary pathway of disablement) can be tested . Current IC/BPS research does not offer a model examining potential mechanisms for patient disability. In providing a theoretical and clinical framework for patient disability, the testing of disability models (i.e. DPM) may inform novel research, as well as appropriate clinical referrals.
The present study aimed to examine the utility of the DPM using validated measures of pain, symptoms and disability. Testing such a model will allow the effects of environmental and psychological factors to be incorporated into a biomedical model where pain and symptoms have been previously associated with disability. Based on the DPM, with support from recent IC/BPS research, it is hypothesized that negative affect (i.e. depression and anxiety), catastrophizing and social support will act as mediators between patient impairments and disability.
Patients and Methods
This sample in the present study comprised 196 women who were suffering from IC/BPS (see Table 1). Patients were recruited from existing IC/BPS patient databases and investigator clinical practices at participating centres in Canada, USA and Germany. All study centres received research ethic board clearance. Potential participants were briefed about the present study after their clinical appointment and, upon verbal consent, were given a package containing an informative letter, a consent form, a postage-paid return envelope, and a battery of questionnaires. Informed consent was considered upon return of the questionnaire, as well as the signed letter of information. All of the questionnaires in Canada and the USA were completed in English, and the questionnaires in Germany were completed in German. The German questionnaires were developed through a validated process of back-translation in which the questionnaires were translated into German, back-translated by a professional translator and then checked once again to ensure consistency of meaning . As with the other centres, informed consent was considered upon completion and return of the questionnaires along with a signed form of consent.
Table 1. Sample demographics.
Total sample (n = 196)
Canada (n = 69)
USA (n = 61)
Germany (n = 66)
F = 1.83, P = 0.16
Ethnicity, n (%)
χ2 = 4.92, P = 0.09
Education, n (%)
χ2 = 46.26, P < 0.01
Less than high school
Married, n (%)
χ2 = 3.01, P = 0.22
Employment, n (%)
χ2 = 28.85, P < 0.01
Time since diagnosis (years)
F = 4.44, P = 0.01
Patients from all centres fulfilled the diagnostic criteria described in the IC Data Base Study , as well as the definition of IC/PBS described at the National Institutes of Health Urologic Chronic Pelvic Pain Consensus Symposium (Baltimore, December 2007). Although not available at the time, patients would have mostly have fulfilled the criteria for diagnosis of IC/BPS outlined in the recently published AUA guidelines . Control women were included if they had no identified diagnosis of IC/PBS.
Demographics and symptoms
Participants completed information on demographics including age, education, employment, ethnicity, partner status and length of illness. The O'Leary-Sant Interstitial Cystitis Symptom and Problem Indices  assessed symptom severity and the problems that patients may experience associated with IC/BPS. Pain was measured using the Short-Form – McGill Pain Questionnaire , in which patients were asked to rate the intensity of 15 verbal descriptors of pain. Physical and mental health QoL was assessed using calculated component scores from the Medical Outcomes Survey – Short Form 12 . The Pain Disability Index  was used to assess disability using seven items rated on a scale from 0 to 10 in the areas of family/home responsibilities, social activities, recreational activities, occupation, sexual behaviour, self-care and life support activities.
Depressive symptoms were assessed using the Center for Epidemiological Studies Depression Scale (CES-D) , a validated scale that has been used in similar chronic pelvic pain studies. The State Trait Anxiety Index  consists of 20 items, and was used to evaluate feelings of apprehension, tension, nervousness and worry. The Pain Catastrophizing Scale (PCS)  used 13 valid and reliable items to assess a patient's cognitions of catastrophizing. The Multidimensional Scale of Perceived Social Support (MSPSS)  was used to assess a patient's perceived level of social support from a significant other, family and friends.
When using a statistics methodology such as structural equation modelling (SEM), multiple assessments are required to validate and support a final structural model. Thus, the general procedures of a structural model building methodology as defined by Mulaik and Millsap  were employed in the present study to provide a framework that was both parsimonious and efficient. The first step of this procedure was to employ an exploratory factor analysis (EFA), conducted in SPSS, version 19 , followed by a confirmatory factor analysis (CFA) using SEM, conducted in AMOS, version 19 , aiming to determine the factor structure and composition of the constructs within the primary pathway of disablement (i.e. the link or association between patient reported impairment and disability). These impairments and functional disability indices were created from the physical measures (i.e. IC/BPS symptoms and bother scores, pain, QoL, disability). A similar model process (EFA then CFA) was used to validate and create the factor structure and composition of the psychosocial factors (i.e. social support, negative affect, pain catastrophizing). Once the factor structure of the variables representing the primary pathway and the psychological variables was completed, the next procedure was to specify the structural portion of the full meditational model and to evaluate its model fit. After examining model fit, the final step was to evaluate the psychosocial factors as mediators between impairments and functional disability using Baron and Kenny's mediation steps  using a bootstrapping technique to determine the significance of the mediators .
A series of ANOVAs were employed to ensure that there were no statistically significant differences in the women between the countries before analyzing them as a homogenous sample. Women in Canada, USA and Germany did not differ in terms of age, partner status, employment, IC symptoms, IC problems, total disability, mental-health related QoL, catastrophizing, anxiety, social support or level of depression. The only difference noted was that German women reported significantly better physical-health-related QoL (F = 5.410, P = 0.003) compared to Canadian women. Thus, the groups were considered homogenous for purposes of the present study.
The EFA to determine the structure and composition of the primary pathway of disablement (i.e. impairments leading to disability) yielded a two-factor model (χ2 [34, n = 196] = 59.559, P = 0.004) (Fig. 1). The factor loadings in factor 1 were in the range 0.949–0.345 and the indicators associated with this factor were family/home responsibilities, social activities, recreation, occupation, sexual behaviour, life-support, self-care, pain and physical health-related QoL. Because all of these variables were related to the ability to functional in every day activities, factor 1 was named functional disability. For the second factor, the factor loadings were in the range 0.943–0.913 and the indicators associated with this factor were IC symptoms and IC problems. These two variables were related to abnormalities as a result of disease-related pathology, and this factor was thus named impairments.
The CFA showed that all of the pathways were significant (P = 0.001) and in the range 0.94–0.42. In addition, impairments and functional disability were significantly correlated (r = 0.51, P < 0.001). Because pain had similar factor loading values on to both impairments and functional disability, it was allowed to freely vary between both impairments and functional disability. After some minor modifications, this model displayed good model fit according to the comparitive fit index (CFI) value and acceptable model fit according to the root mean square error of approximation (RMSEA) value (χ2 [41, n = 196] = 69.097, P = 0.004, CFI = 0.976 and RMSEA = 0.059 [90% CI, 0.034–0.083; P = 0.249]).
The results from the EFA to determine the structure and composition of the psychosocial factors yielded a three-factor model (χ2 [33, n = 196] = 75.746, P < 0.001) (Fig. 2). The factor loadings in factor 1 were in the range 0.930–0.428 and the indicators associated with this factor were depressed affect (CES-D), somatic symptoms (CES-D), positive affect (CES-D), interpersonal difficulties (CES-D), mental health-related QoL (Short Form 12) and anxiety (State Trait Anxiety Index). Because this factor incorporated the constructs of depression, anxiety and psychological distress, it was named negative affect. In the assessment of the second factor, the factor loadings were in the range 0.948–0.470. The measured variables that were indicated in this factor were perceived social support from a significant other (MSPSS), from family (MSPSS) and from friends (MSPSS) and, as such, factor 2 was named social support. Lastly, the factor loadings in factor 3 were in the range 0.896–0.802 and measured variables that were indicated in this factor were magnification (PCS), rumination (PCS) and helplessness (PCS) and this factor was thus named pain catastrophizing.
Results from the EFA were followed up with a CFA, in which all of the pathways were shown to be significant, and the standardized regression coefficients were in the range 0.91–0.47. There was a significant correlation between catastrophizing and negative affect (r = 0.60, P < 0.001) and negative affect and social support (r = −0.46, P < 0.001) but not between social support and catastrophizing (r = −0.144, P = 0.082). This three-factor model showed good fit according to the CFI value and acceptable fit according to the RMSEA value, with no necessary modifications (χ2 [51, n = 196] = 110.007, P < 0.001, CFI = 0.956, RMSEA = 0.077 [90% CI, 0.057–0.097; P = 0.014]).
The results from the measurement models described above were used to specify the structural model (Fig. 3). The final model showed acceptable fit to the data according to both the CFI and RMSEA values (χ2 [217, n = 196] = 420.281, P < 0.001, CFI = 0.925, RMSEA = 0.069 [90% CI, 0.059–0.079; P = 0.001]). Negative affect was significantly correlated with catastrophizing (r = 0.52, P < 0.001) and social support (r = −0.39, P < 0.001). Because catastrophizing was not significantly correlated with social support in previous analyses, it was not specified within the final model.
The final step was to evaluate specific hypotheses regarding the parameters of the mediation model. In the evaluation of negative affect as a mediator, the bootstrapping technique produced estimates such that both the standardized direct (0.368, P < 0.001 [CI, 0.231–0.489]) and indirect (0.117, P < 0.001 [CI, 0.050–0.201]) effects were significant in the model, which was consistent with a partial mediation effect . Similarly, in the evaluation of catastrophizing, the bootstrapping technique produced estimates such that the standardized direct (0.373, P < 0.001 [CI, 0.215–0.513]) and indirect (0.113, P < 0.001 [CI, 0.051–0.208]) effects were significant, which was also consistent with a partial mediation effect . However, social support was not found to be a significant mediator between impairments and functional disability (non-significant indirect effect: 0.010, P = 0.188 [CI, –0.005 to 0.005]) and was thus excluded from the final model. This final model including the mediators of negative affect and catastrophizing, but not social support, showed reasonable fit to the data (χ2 [160, n = 196] = 317.233, P < 0.001, CFI = 0.936, RMSEA = 0.071 [90% CI, 0.059–0.082; P = 0.002]). In comparison with the original proposed model, the model with only catastrophizing and negative affect had significantly better fit to the data once social support was removed (χ2 [57, n = 196] = 103.048, P = 0.001). The combined mediation effect of negative affect and catastrophizing produced estimates such that the standardized direct (0.342, P < 0.001 [CI, 0.194–0.474]) and indirect (0.144, P = 0.001 [CI, 0.069 – 0.238]) effects were both significant, thus indicating a partial mediation effect .
The present study is the first to apply and test a theoretical conceptualization of the factors contributing to disability in women suffering from IC/BPS. The findings obtained in the present study suggest that psychosocial factors are significant in mediating the relationship between impairments and patient disability, with negative affect (i.e. depression, anxiety) and pain catastrophizing acting as significant mediators. The results support previous research suggesting that catastrophizing and negative affect have associations with pain, QoL and symptoms in women suffering from IC/BPS [3, 5]. Catastrophizing and social support did not have a significant correlation in the present study. This is contrary to findings in other urogenital pain samples [23, 24]. There is no method that is able to explain the finding within the present dataset. However, when considering disability and social support, there may be other factors at work (e.g. support from the physician, nurse, etc.). These findings indicate that catastrophizing and negative affect may be the mechanisms by which pain and symptoms elevate reports of patient disability.
Understanding how pain and pain-associated symptoms may become disabling through cognitive mechanisms is an importance advance for IC/BPS management. It is possible that patients reporting greater negative affect will probably show a bias towards negative rumination or magnification about their pain. Such patients have a greater probablity of anticipating a broader range of pain experiences, interpret ambiguous signals as being related to pain, and to attend to pain-related cues and experience interference of pain during cognitive tasks . When depressed, women have a greater probability of becoming internally focused and ruminating, which acts to increase attention to painful symptoms. Increased attentional focus from the depression and catastrophizing may manifest in reports of more severe pain . Indeed, previous studies suggest that pain catastrophizing is associated with a heightened attentional bias to pain-relevant stimuli and an inability to disengage from pain or pain cues . Social support was not a significant mediator in the present model and, generally, the beneficial effects of social support are complex and difficult to conceptualize . Particular types of social support such as distracting spousal responses are reported to moderate the negative effects of pain on QoL in IC/BPS , although this model remains to be examined between impairments and disability.
A potential limitation of the present study was that the women were internationally recruited. A series of analyses showed that women across these countries did not differ with respect to many important variables. Women diagnosed with IC/BPS in Canada, USA and Germany were found to have similar demographic information and to experience similar levels of disease-specific symptoms, disability, catastrophizing, anxiety, social support and depression. Instead of a limitation, these data indicate the generalizability of the experiences of women suffering from IC/BPS internationally and, by proxy, make the current findings relevant to women suffering from IC/BPS outside of North America. Another potential limitation of the present study is the relatively high level of education reported by our patients. IC/BPS has typically been considered as a condition experienced by Caucasian women, with much of the research focusing on those with higher education. This may be the result of a sampling effect. Recent research indicates that IC/BPS is common in women of various ethnicities and that symptoms may be more common in the lowest social economic status category (i.e. according to income and education) . Future research should include samples with greater ethnic minorities and greater proportions of those with lower social economic status. Although SEM is an advanced statistical procedure that is able to comprehensively test meditational relationships, this statistical method does not translate into causality. A true method of inferring causation is through longitudinal or experimental study designs. Yet SEM has many advantages over other statistical methods for establishing the conditions necessary to infer causality and is a useful procedure in model construction .
Negative affect and catastrophizing impact the relationship between patients' impairments and disability, and these factors appear to be heightening this relationship via these cognitive mechanisms. Clinical interventions in women suffering from IC/BPS should target catastrophizing and mood for reduction using cognitive behavioural strategies aiming to decrease pain-related disability. Recent feasibility research using a standardized psycho-educational therapy with men suffering from chronic pelvic pain (i.e. chronic prostatitis) shows that significant reductions are possible in the areas of pain, patient disability and catastrophizing in as little as 8 weeks . Discussion of catastrophizing concerning pain and its ‘meaning’ to patients can be critical for improving patient management. Future research in disability should manipulate other variables associated with patient QoL and could consider longitudinal perspectives.