IACFS/ME Bulletin

  Search

 

 

HOW VALID IS THE MODEL BEHIND COGNITIVE BEHAVIOR THERAPY FOR CHRONIC FATIGUE SYNDROME? AN EVALUATION OF THE ADDITIONAL DATA FROM THE TRIAL BY PRINS ET AL.

 

B. Stouten 1*, PhD
Ellen M. Goudsmit 2, PhD FBPsS

 

 

1. Einsteindreef 67A, Utrecht, The Netherlands

2. University of East London, UK
                                                     



 

ABSTRACT

The cognitive behavior therapy (CBT) program studied by Prins et al. is based on a model of chronic fatigue syndrome that posits that fatigue and functional impairment are perpetuated by physical inactivity, somatic attributions, focusing on bodily symptoms and a low sense of control. A recent analysis of the data from three trials based on a model devised by Vercoulen et al. concluded that the effect of CBT on fatigue could not be attributed to a persistent increase in physical activity. We therefore examined the effect of treatment on the remaining three variables in the model using data from one of the trials, available in the public domain. The results from the groups given CBT, Guided Support and treatment as usual revealed that CBT had no significant impact on somatic attributions and focusing on bodily symptoms, and that in line with established guidelines, these two variables were not mediating factors. The only variable in the model showing an effect of CBT was sense of control. We submit that there is now sufficient evidence to warrant a review of CFS guidelines which advocate interventions aimed particularly at increasing physical activity and challenging somatic attributions, and that more flexible programs which address loss of control deserve further consideration and evaluation.


INTRODUCTION

In the largest randomized controlled trial of cognitive behavior therapy (CBT) conducted to date, Prins et al. found that individuals with chronic fatigue syndrome (CFS) improved significantly more on measures of fatigue and functional impairment than the patients who attended guided support groups to exchange experiences, and patients in the no-treatment 'natural course' arm (1). These results were maintained at follow-up.

The CBT program was based on Vercoulen et al.’s model of CFS (1,2), see Figure 1. This model posits that attributing complaints to a somatic cause such as a viral infection results in reduced levels of physical activity (to avoid post-exertional increases in symptomatology), which in turn leads to increased fatigue severity and functional impairment.
Focusing on bodily symptoms and a loss of perceived control over symptoms also have a direct association with fatigue.

The model appears to oversimplify what is a complex, multi-faceted syndrome (3). It was a study by Wiborg et al. (4) that alerted us to a major flaw. They re-analyzed the data from the first trial by Prins et al. plus two other randomized controlled trials which they had conducted, to assess the efficacy of CBT. Their findings revealed not only that the program had little effect on the level of physical activity, but that across the samples, the mean mediation effect of physical activity averaged only about 1% of the total treatment effect (4). These results challenge the rationale behind the graded activity component in their program as well as the validity of the model on which the program is based.

The significance of the findings and possible implications for the management of CFS led us to re-examine the data from the researchers available in the public domain and to assess the role of three elements in the program which had not been evaluated by Wiborg et al. This information had been published in a report by the Dutch Health Care Insurance Board in 2002 and provides additional data relating to the first trial (5). The main aim of our study was to clarify the influence of CBT on the three other variables of the model, i.e. somatic attributions, focusing on bodily symptoms and sense of control. The second aim was to determine whether any of these variables may act as mediating factors between CBT and both fatigue and functional impairment.

Figure 1. Model of perpetuating factors for CFS by Vercoulen et al. (1998).





METHODS
The means and standard deviations recorded by Prins et al. at baseline, after treatment (8 months) and at follow-up (14 months) were obtained from a report on the study written in the Dutch language (5). We used t-tests for independent samples to compare the groups given CBT, Guided Support and treatment as usual ('Natural Course'). Alpha was set at p < 0.017, following Wiborg et al., to correct for multiple comparisons.

Our exploratory mediation analysis followed the procedure recommended by Baron and Kenny (6). The limited data in the public domain only allowed us to perform the first regression analysis, i.e., by testing the regression of the mediator on the independent variable, we were able to identify variables that were not mediators but it was not possible to confirm that a variable was a mediator. Since we compared two groups in this study, (CBT vs Guided Support, or CBT vs Natural Course), regression on the group scores is equivalent to the two-sample t-test, and therefore one can be substituted for the other (7,8). Analyzing Table 2 from Wiborg et al. (4) with t-tests and comparing the outcomes with those of the regression reported in their Table 3 confirms that the two methods produce the same results, hence supporting the use of t-tests to identify the non-mediating variables.

Since the CBT program was directly based on Vercoulen et al.’s model, our mediation analysis focused on the variables in the model itself rather than the change in these variables between baseline and follow-up.


RESULTS
The results are depicted in Table 1. A comparison of the groups given CBT, Guided Support and treatment as usual ('Natural Course') revealed that CBT had no significant impact on somatic attributions. Moreover, while CBT reduced focusing on bodily symptoms, the differences between the groups did not reach significance. The only significant effect of CBT was on sense of control assessed at 8 and 14 months.


The results can also be interpreted as the outcomes of a regression test and this shows that there is no significant relationship between CBT on the one hand, and focusing on bodily symptoms or somatic attributions, on the other. Thus following the guidelines developed by Baron and Kenny, one can argue that these two variables are not mediating factors.

 

Table 1. Differences in perpetuating factors between cognitive behavior therapy (CBT), Guided Support and Natural Course groups directly before treatment (0 months), directly after treatment (8 months) and at follow-up (14 months).

DISCUSSION
Our analysis of the data from a randomized controlled of CBT found that three of the four variables in the model underpinning the protocol, namely activity levels, somatic attributions and focusing on symptoms, showed no significant changes following treatment or at follow-up. Since 87% of the random sample of audiotaped CBT sessions were rated as adequate or good, it may be assumed that the therapists generally kept to the manual and it is therefore plausible that the lack of significant differences between the treated arm and controls may reflect shortcomings of the underlying model.

Given the program appears to have had little impact on physical activity and focusing on bodily symptoms, the model proposed by Vercoulen et al. cannot explain the previously documented effect on functional impairment. It may be argued that this provides further evidence that the model is incomplete, and that additional variables play a role in determining outcome. Indeed, the results from this limited analysis support the view that the effect of CBT on fatigue may be mediated, at least in part, by an increase in the patients’ sense of control.

Our view that the protocol based on Vercoulen’s model may not reflect the complexity of CFS is consistent with the results from a study which evaluated this model using data from a variety of measures obtained from a large community sample (3). Song and Jason divided the participants into various groups, namely people who met the criteria for CFS, those with psychiatrically explained chronic fatigue, people with medically explained chronic fatigue, a group with idiopathic chronic fatigue and a group where fatigue was related to substance abuse. They were able to replicate the model in the sample with psychiatric conditions but not in the patients with CFS.

In light of our findings and those of Wiborg et al., we submit that alternative theories and interventions which address the perceived loss of control deserve further consideration and evaluation. For example, one alternative intervention that may increase the sense of mastery over symptoms is counseling, which has been shown to be as helpful as CBT, but was found to be less costly (9,10). Similarly, more psycho-educational interventions or versions of CBT that do not attempt to change reasonable somatic attributions and which include pacing rather than graded activity to limit exertion-related symptomatology appear to be as helpful, and may be more appropriate for many individuals with CFS (11,12).

A notable limitation of our study is that we did not have access to the raw data and could therefore not establish whether sense of control is a mediator. Since Wiborg et al. and Prins et al. share our interest in the mediating role of illness-related cognitions and possess the data required (1,4), we hope that they will publish their findings in relation to a variable which may be an important influence on outcome.

Currently, virtually all existing guidelines for the management of CFS advocate CBT-based interventions aimed primarily at changing somatic attributions and increasing physical activity, e.g. NICE 2007 (13). Although a recent systematic review showed that these interventions helped about 4 in 10 patients, this effect was not sustained over time (14). Given that there is little empirical support for the view that changing attributions and increasing activity are essential for successful outcome, and since the failure to address somatic symptoms, except from a cognitive-behavioral perspective, often leads to tensions in the patient-therapist relationship, we submit that there is now sufficient evidence to warrant a review of the treatment models and guidelines at the earliest opportunity.


REFERENCES

(1)  Prins JB, Bleijenberg G, Bazelmans E, Elving LD, de Boo TM, Severens JL, van der Wilt GJ, Spinhoven P, van der Meer JWM (2001). Cognitive behaviour therapy for chronic fatigue syndrome: a multicentre randomised controlled trial. Lancet. 2001; 357:841-847.


(2)  Vercoulen JHMM, Swanink CMA, Galama JMD, Fennis JFM, Jongen PJH, Hommes OR, van der Meer JWM, Bleijenberg G. The persistence of fatigue in chronic fatigue syndrome and multiple sclerosis: development of a model. J Psychosom Res. 1998; 45:507-517.


(3)  Song S, Jason LA. A population-based study of chronic fatigue syndrome (CFS) experienced in differing patient groups: an effort to replicate Vercoulen et al.'s model of CFS. J Ment Health. 2005; 14:277-289.

(4)  Wiborg JF, Knoop H, Stulemeijer M, Prins JB, Bleijenberg G. How does cognitive behaviour therapy reduce fatigue in patients with chronic fatigue syndrome? The role of physical activity. Psychol Med. 2010; 40:1281-1287. 



(5)  Van Essen M, de Winter LJM. Cognitieve gedragstherapie bij het chronisch vermoeidheidssyndroom (cognitive behaviour therapy for chronic fatigue syndrome). Nr.: 02/111. Amstelveen, The Netherlands: College voor Zorgverzekeringen; 2002.



(6)  Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986; 51:1173-1182.



(7)  Huitema BE (1980). The analysis of covariance and alternatives. New York : John Wiley & Sons. 1980: pp 48-54. ISBN: 0-471-42044-1.



(8)  Altman DG. Practical statistics for medical research. London: Chapman & Hall. 1991: p 319. ISBN: 0-412-27630-5.

(9)  Ridsdale L, Godfrey E, Chalder T, Seed P, King M, Wallace P, Wessely S. Chronic fatigue in general practice: is counselling as good as cognitive behaviour therapy? Br J Gen Pract. 2001; 51:19-24.

(10)  Chisholm D, Godfrey E, Ridsdale L, Chalder T, King M, Seed P, Wallace P, Wessely S, Fatigue Trialists' Group. Chronic fatigue in general practice: economic evaluation of counselling versus cognitive behaviour therapy. Br J Gen Pract. 2001; 51:15–18.


(11)  Jason LA, Torres-Harding S, Friedberg F, Corradi K, Njoku MG, Donalek J, Reynolds N, Brown M, Weitner BB, Rademaker A, Papernik M. Non-pharmacologic interventions for CFS: a randomized trial. J Clin Psych Med Settings. 2007; 14:275-296.

(12)  Goudsmit EM, Ho-Yen DO, Dancey CP. Learning to cope with chronic illness. Efficacy of a multi-component treatment for people with chronic fatigue syndrome. Patient Educ Counsel. 2009; 77:231-236.

(13)  National Institute for Health and Clinical Excellence (NICE). NICE clinical guideline 53: Chronic fatigue syndrome / myalgic encephalomyelitis (or encephalopathy): diagnosis and management of CFS/ME in adults and children,  August 2007. Available at: http://www.nice.org.uk/nicemedia/pdf/CG53NICEGuideline.pdf. Accessed March 5, 2010.


(14)  Price JR, Mitchell E, Tidy E, Hunot V. Cognitive therapy for chronic fatigue syndrome in adults. Cochrane Database of Systematic Reviews. 2008;3. Art. No.: CD001027. doi: 10.1002 / 146511858.CD001027.pub2.


Bulletin of the IACFS/ME. 2010;18(2):82-89. © 2010 IACFS/ME

Download a PDF of this paper

(You will need Adobe Reader to open this file. If you do not have Adobe Reader, click here to download.)

Return to Bulletin of the IACFS/ME, Volume 18, Issue 2

HOW VALID IS THE MODEL BEHIND COGNITIVE BEHAVIOR THERAPY FOR CHRONIC FATIGUE SYNDROME? AN EVALUATION OF THE ADDITIONAL DATA FROM THE TRIAL BY PRINS ET AL.

 

B. Stouten 1*, PhD
Ellen M. Goudsmit 2, PhD FBPsS

 

 

1. Einsteindreef 67A, Utrecht, The Netherlands

2. University of East London, UK
                                                     



 

ABSTRACT

The cognitive behavior therapy (CBT) program studied by Prins et al. is based on a model of chronic fatigue syndrome that posits that fatigue and functional impairment are perpetuated by physical inactivity, somatic attributions, focusing on bodily symptoms and a low sense of control. A recent analysis of the data from three trials based on a model devised by Vercoulen et al. concluded that the effect of CBT on fatigue could not be attributed to a persistent increase in physical activity. We therefore examined the effect of treatment on the remaining three variables in the model using data from one of the trials, available in the public domain. The results from the groups given CBT, Guided Support and treatment as usual revealed that CBT had no significant impact on somatic attributions and focusing on bodily symptoms, and that in line with established guidelines, these two variables were not mediating factors. The only variable in the model showing an effect of CBT was sense of control. We submit that there is now sufficient evidence to warrant a review of CFS guidelines which advocate interventions aimed particularly at increasing physical activity and challenging somatic attributions, and that more flexible programs which address loss of control deserve further consideration and evaluation.


INTRODUCTION

In the largest randomized controlled trial of cognitive behavior therapy (CBT) conducted to date, Prins et al. found that individuals with chronic fatigue syndrome (CFS) improved significantly more on measures of fatigue and functional impairment than the patients who attended guided support groups to exchange experiences, and patients in the no-treatment 'natural course' arm (1). These results were maintained at follow-up.

The CBT program was based on Vercoulen et al.’s model of CFS (1,2), see Figure 1. This model posits that attributing complaints to a somatic cause such as a viral infection results in reduced levels of physical activity (to avoid post-exertional increases in symptomatology), which in turn leads to increased fatigue severity and functional impairment.
Focusing on bodily symptoms and a loss of perceived control over symptoms also have a direct association with fatigue.

The model appears to oversimplify what is a complex, multi-faceted syndrome (3). It was a study by Wiborg et al. (4) that alerted us to a major flaw. They re-analyzed the data from the first trial by Prins et al. plus two other randomized controlled trials which they had conducted, to assess the efficacy of CBT. Their findings revealed not only that the program had little effect on the level of physical activity, but that across the samples, the mean mediation effect of physical activity averaged only about 1% of the total treatment effect (4). These results challenge the rationale behind the graded activity component in their program as well as the validity of the model on which the program is based.

The significance of the findings and possible implications for the management of CFS led us to re-examine the data from the researchers available in the public domain and to assess the role of three elements in the program which had not been evaluated by Wiborg et al. This information had been published in a report by the Dutch Health Care Insurance Board in 2002 and provides additional data relating to the first trial (5). The main aim of our study was to clarify the influence of CBT on the three other variables of the model, i.e. somatic attributions, focusing on bodily symptoms and sense of control. The second aim was to determine whether any of these variables may act as mediating factors between CBT and both fatigue and functional impairment.

Figure 1. Model of perpetuating factors for CFS by Vercoulen et al. (1998).





METHODS
The means and standard deviations recorded by Prins et al. at baseline, after treatment (8 months) and at follow-up (14 months) were obtained from a report on the study written in the Dutch language (5). We used t-tests for independent samples to compare the groups given CBT, Guided Support and treatment as usual ('Natural Course'). Alpha was set at p < 0.017, following Wiborg et al., to correct for multiple comparisons.

Our exploratory mediation analysis followed the procedure recommended by Baron and Kenny (6). The limited data in the public domain only allowed us to perform the first regression analysis, i.e., by testing the regression of the mediator on the independent variable, we were able to identify variables that were not mediators but it was not possible to confirm that a variable was a mediator. Since we compared two groups in this study, (CBT vs Guided Support, or CBT vs Natural Course), regression on the group scores is equivalent to the two-sample t-test, and therefore one can be substituted for the other (7,8). Analyzing Table 2 from Wiborg et al. (4) with t-tests and comparing the outcomes with those of the regression reported in their Table 3 confirms that the two methods produce the same results, hence supporting the use of t-tests to identify the non-mediating variables.

Since the CBT program was directly based on Vercoulen et al.’s model, our mediation analysis focused on the variables in the model itself rather than the change in these variables between baseline and follow-up.


RESULTS
The results are depicted in Table 1. A comparison of the groups given CBT, Guided Support and treatment as usual ('Natural Course') revealed that CBT had no significant impact on somatic attributions. Moreover, while CBT reduced focusing on bodily symptoms, the differences between the groups did not reach significance. The only significant effect of CBT was on sense of control assessed at 8 and 14 months.


The results can also be interpreted as the outcomes of a regression test and this shows that there is no significant relationship between CBT on the one hand, and focusing on bodily symptoms or somatic attributions, on the other. Thus following the guidelines developed by Baron and Kenny, one can argue that these two variables are not mediating factors.

 

Table 1. Differences in perpetuating factors between cognitive behavior therapy (CBT), Guided Support and Natural Course groups directly before treatment (0 months), directly after treatment (8 months) and at follow-up (14 months).

DISCUSSION
Our analysis of the data from a randomized controlled of CBT found that three of the four variables in the model underpinning the protocol, namely activity levels, somatic attributions and focusing on symptoms, showed no significant changes following treatment or at follow-up. Since 87% of the random sample of audiotaped CBT sessions were rated as adequate or good, it may be assumed that the therapists generally kept to the manual and it is therefore plausible that the lack of significant differences between the treated arm and controls may reflect shortcomings of the underlying model.

Given the program appears to have had little impact on physical activity and focusing on bodily symptoms, the model proposed by Vercoulen et al. cannot explain the previously documented effect on functional impairment. It may be argued that this provides further evidence that the model is incomplete, and that additional variables play a role in determining outcome. Indeed, the results from this limited analysis support the view that the effect of CBT on fatigue may be mediated, at least in part, by an increase in the patients’ sense of control.

Our view that the protocol based on Vercoulen’s model may not reflect the complexity of CFS is consistent with the results from a study which evaluated this model using data from a variety of measures obtained from a large community sample (3). Song and Jason divided the participants into various groups, namely people who met the criteria for CFS, those with psychiatrically explained chronic fatigue, people with medically explained chronic fatigue, a group with idiopathic chronic fatigue and a group where fatigue was related to substance abuse. They were able to replicate the model in the sample with psychiatric conditions but not in the patients with CFS.

In light of our findings and those of Wiborg et al., we submit that alternative theories and interventions which address the perceived loss of control deserve further consideration and evaluation. For example, one alternative intervention that may increase the sense of mastery over symptoms is counseling, which has been shown to be as helpful as CBT, but was found to be less costly (9,10). Similarly, more psycho-educational interventions or versions of CBT that do not attempt to change reasonable somatic attributions and which include pacing rather than graded activity to limit exertion-related symptomatology appear to be as helpful, and may be more appropriate for many individuals with CFS (11,12).

A notable limitation of our study is that we did not have access to the raw data and could therefore not establish whether sense of control is a mediator. Since Wiborg et al. and Prins et al. share our interest in the mediating role of illness-related cognitions and possess the data required (1,4), we hope that they will publish their findings in relation to a variable which may be an important influence on outcome.

Currently, virtually all existing guidelines for the management of CFS advocate CBT-based interventions aimed primarily at changing somatic attributions and increasing physical activity, e.g. NICE 2007 (13). Although a recent systematic review showed that these interventions helped about 4 in 10 patients, this effect was not sustained over time (14). Given that there is little empirical support for the view that changing attributions and increasing activity are essential for successful outcome, and since the failure to address somatic symptoms, except from a cognitive-behavioral perspective, often leads to tensions in the patient-therapist relationship, we submit that there is now sufficient evidence to warrant a review of the treatment models and guidelines at the earliest opportunity.


REFERENCES

(1)  Prins JB, Bleijenberg G, Bazelmans E, Elving LD, de Boo TM, Severens JL, van der Wilt GJ, Spinhoven P, van der Meer JWM (2001). Cognitive behaviour therapy for chronic fatigue syndrome: a multicentre randomised controlled trial. Lancet. 2001; 357:841-847.


(2)  Vercoulen JHMM, Swanink CMA, Galama JMD, Fennis JFM, Jongen PJH, Hommes OR, van der Meer JWM, Bleijenberg G. The persistence of fatigue in chronic fatigue syndrome and multiple sclerosis: development of a model. J Psychosom Res. 1998; 45:507-517.


(3)  Song S, Jason LA. A population-based study of chronic fatigue syndrome (CFS) experienced in differing patient groups: an effort to replicate Vercoulen et al.'s model of CFS. J Ment Health. 2005; 14:277-289.

(4)  Wiborg JF, Knoop H, Stulemeijer M, Prins JB, Bleijenberg G. How does cognitive behaviour therapy reduce fatigue in patients with chronic fatigue syndrome? The role of physical activity. Psychol Med. 2010; 40:1281-1287. 



(5)  Van Essen M, de Winter LJM. Cognitieve gedragstherapie bij het chronisch vermoeidheidssyndroom (cognitive behaviour therapy for chronic fatigue syndrome). Nr.: 02/111. Amstelveen, The Netherlands: College voor Zorgverzekeringen; 2002.



(6)  Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986; 51:1173-1182.



(7)  Huitema BE (1980). The analysis of covariance and alternatives. New York : John Wiley & Sons. 1980: pp 48-54. ISBN: 0-471-42044-1.



(8)  Altman DG. Practical statistics for medical research. London: Chapman & Hall. 1991: p 319. ISBN: 0-412-27630-5.

(9)  Ridsdale L, Godfrey E, Chalder T, Seed P, King M, Wallace P, Wessely S. Chronic fatigue in general practice: is counselling as good as cognitive behaviour therapy? Br J Gen Pract. 2001; 51:19-24.

(10)  Chisholm D, Godfrey E, Ridsdale L, Chalder T, King M, Seed P, Wallace P, Wessely S, Fatigue Trialists' Group. Chronic fatigue in general practice: economic evaluation of counselling versus cognitive behaviour therapy. Br J Gen Pract. 2001; 51:15–18.


(11)  Jason LA, Torres-Harding S, Friedberg F, Corradi K, Njoku MG, Donalek J, Reynolds N, Brown M, Weitner BB, Rademaker A, Papernik M. Non-pharmacologic interventions for CFS: a randomized trial. J Clin Psych Med Settings. 2007; 14:275-296.

(12)  Goudsmit EM, Ho-Yen DO, Dancey CP. Learning to cope with chronic illness. Efficacy of a multi-component treatment for people with chronic fatigue syndrome. Patient Educ Counsel. 2009; 77:231-236.

(13)  National Institute for Health and Clinical Excellence (NICE). NICE clinical guideline 53: Chronic fatigue syndrome / myalgic encephalomyelitis (or encephalopathy): diagnosis and management of CFS/ME in adults and children,  August 2007. Available at: http://www.nice.org.uk/nicemedia/pdf/CG53NICEGuideline.pdf. Accessed March 5, 2010.


(14)  Price JR, Mitchell E, Tidy E, Hunot V. Cognitive therapy for chronic fatigue syndrome in adults. Cochrane Database of Systematic Reviews. 2008;3. Art. No.: CD001027. doi: 10.1002 / 146511858.CD001027.pub2.


Bulletin of the IACFS/ME. 2010;18(2):82-89. © 2010 IACFS/ME

Download a PDF of this paper

(You will need Adobe Reader to open this file. If you do not have Adobe Reader, click here to download.)

Return to Bulletin of the IACFS/ME, Volume 18, Issue 2

Copyright 2013 by IACFS/ME  · Terms Of Use · Privacy Statement