To the Editors:

Although the causes of chronic pain and disability remain largely unknown, psychological factors may maintain the symptoms (1, 2). Cognitive–behavioral therapy (CBT) has been developed to help patients reattribute their complaints, and reinterpret the relationship between their cognitive appraisals and behavior with the aim of reducing the effect of pain on daily functioning. Although the effects of CBT for patients with chronic pain are promising, several authors have called for further refinement. Indeed, one of the challenges of CBT is its relative lack of specificity; CBT may include self-instructions, relaxation, biofeedback, and goal setting, to name a few examples, and it often is embedded in more comprehensive pain management programs (3). Some authors have pleaded in favor of studies that aim to better match the particular kind of CBT to the specific characteristics of the patients (4, 5).

The article by van Koulil and colleagues published recently in Arthritis Care & Research (6) is an example of such an attempt to tailor treatment to subgroups of fibromyalgia patients. Briefly, the authors selected high-risk patients (high negative mood and/or anxiety) and classified them into either a pain-avoidance or a pain-persistence group. Each patient was then randomized into a corresponding CBT (plus exercise training) or a waiting list condition (WLC). Although the 2 kinds of CBT had a number of features in common, they were also geared towards the characteristic of the subgroup. Generally, CBT for the pain-avoidance group was aimed at reducing fears and avoidance behavior, while CBT for the pain-persistence group included activity pacing techniques and changing pain-persistence cognitions. The results of this novel approach were that both forms of CBT, in combination with exercise training, were superior to WLC, with significant differences in pain, fatigue, disability, and negative mood. The authors also computed reliable change indices for several measures, and showed that a greater proportion of patients in the treated groups achieved reliable changes (7, 8). In passing, we note that the authors regarded a reliable change as synonymous with a clinically significant change, a position with which we disagree. The authors concluded that tailored CBT for high-risk patients are effective in improving physical and psychological functioning.

Although we strongly encourage the kind of research that van Koulil et al have conducted, and appreciate the importance of their study, there are at least 3 methodologic reasons why we think their conclusion that they found empirical evidence for tailored treatment is not justified, and at least premature.

First, the authors made a propositional logic error. One can only conclude that a certain treatment A is the best treatment for subgroup X, and treatment B for subgroup Y, when it is demonstrated not only that A leads to beneficial outcomes in X, and B to beneficial outcomes in Y, but also the reverse, namely that X and Y do less well (minimal improvement, no significant changes, or significant deterioration) when receiving the counterpart treatments B and A, respectively. In fact, van Koulil et al only tested the first part of the proposition, showing that both kinds of CBT are effective for the respective subgroups, which is a necessary but not sufficient condition to conclude that the treatments are tailored.

Second, tailored treatments only make sense when they are applied to an empirically tested and theoretically grounded definition of subgroups. The authors categorized the patients based on a judgment of a trained therapist using a semistructured interview and a cut-off score on a questionnaire assessing self-reported avoidance behavior. This seems like a reasonable way to identify patients who tend to avoid painful activities, and whose avoidance behavior may mediate their levels of functional disability. But how about the selection of the so-called pain-persistence group? Patients scoring below the cut-off score for avoidance behavior were automatically considered as belonging to the pain-persistence subgroup, as if pain persistence was simply the opposite of pain avoidance. The literature on behavioral persistence patterns in chronic pain is rather scarce, but suggests that the underlying mechanisms are largely unknown, and the hypotheses that have been put forward are rather complex (9, 10). Indeed, one version of the pain-persistence model posits that persistence may be a form of active avoidance of other unwanted non–pain-associated outcomes, such as social disapproval (11). Also of interest is that both the pain-avoidance and pain-persistence treatments were aimed at increasing daily activities and physical condition, which is at odds with the authors' assumption that some patients “tend to persist in their activities in spite of the pain, which can lead to overuse and more symptoms in the long run” (6). If the authors in fact adhere to this idea, then one would expect a treatment that is aimed at flexible adjustment of personal goals, and self-regulation or decrease of activities, rather than an increase of activities.

Third, the design of the randomized controlled trial makes it impossible to discern the effects of CBT and the exercise training components on patients receiving active treatment. Given that all patients receiving active treatment improved, 1 parsimonious interpretation is that the improvements were due to the exposure to (painful) physical activities, leading to a reattribution of pain as hampering daily activities, with functional improvement as a result.

To conclude, we admire the authors' diligence in conducting this trial. They represent a select group of researchers who have attempted to broach this difficult problem of tailoring psychological treatment to the specific characteristics of the patients. We would like to take the opportunity to call for a novel methodology that clinical researchers can use when studying tailored treatments for chronic pain patients. The benefits of such an approach can be seen in the moderator–mediator distinction put forward by Baron and Kenny (12), in which the presences of theoretically grounded mechanisms of pain and disability, in combination with empirically tested manipulations, as well as their interactions, are systematically evaluated. Such an approach is complex, but most likely will enhance both our understanding and management of individuals with chronic pain (13, 14).


Dr. Vlaeyen has received consultancies, speaking fees, and/or honoraria (less than $10,000) from Institut UPSA de la Douleur and royalties from Oxford University Press for an edited book on fear of pain (11) but declares no conflicts of interest.

  • 1
    Leeuw M, Goossens ME, Linton SJ, Crombez G, Boersma K, Vlaeyen JW. The fear-avoidance model of musculoskeletal pain: current state of scientific evidence. J Behav Med 2007; 30: 7794.
  • 2
    Heuts PH, Vlaeyen JW, Roelofs J, De Bie RA, Aretz K, Van Weel C, et al. Pain-related fear and daily functioning in patients with osteoarthritis. Pain 2004; 110: 22835.
  • 3
    Morley S. Efficacy and effectiveness of cognitive behaviour therapy for chronic pain: progress and some challenges. Pain 2011; 152 Suppl: S99106.
  • 4
    Turk DC. Customizing treatment for chronic pain patients: who, what, and why. Clin J Pain 1990; 6: 25570.
  • 5
    Vlaeyen JW, Morley S. Cognitive-behavioral treatments for chronic pain: what works for whom? Clin J Pain 2005; 21: 18.
  • 6
    Van Koulil S, van Lankveld W, Kraaimaat FW, van Helmond T, Vedder A, van Hoorn H, et al. Tailored cognitive–behavioral therapy and exercise training for high-risk patients with fibromyalgia. Arthritis Care Res (Hoboken) 2010; 62: 137785.
  • 7
    Jacobson NS, Roberts LJ, Berns SB, McGlinchey JB. Methods for defining and determining the clinical significance of treatment effects: description, application, and alternatives. J Consult Clin Psychol 1999; 67: 3007.
  • 8
    Morley S, Williams A, Hussain S. Estimating the clinical effectiveness of cognitive behavioural therapy in the clinic: evaluation of a CBT informed pain management programme. Pain 2008; 137: 67080.
  • 9
    Hasenbring MI, Hallner D, Rusu AC. Fear-avoidance- and endurance-related responses to pain: development and validation of the Avoidance-Endurance Questionnaire (AEQ). Eur J Pain 2009; 13: 6208.
  • 10
    Vlaeyen JW, Morley S. Active despite pain: the putative role of stop-rules and current mood. Pain 2004; 110: 5126.
  • 11
    Morley JS, Eccleston C. The object of fear in pain. In: AsmundsonGJ, VlaeyenJW, CrombezG, editors. Understanding and treating fear of pain. Oxford: Oxford University Press; 2004.
  • 12
    Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol 1986; 51: 117382.
  • 13
    Morley S, Vlaeyen JW. Epilogue to the special topic series. Clin J Pain 2005; 21: 6972.
  • 14
    Turner JA, Holtzman S, Manci L. Mediators, moderators, and predictors of therapeutic change in cognitive-behavioral therapy for chronic pain. Pain 2007; 127: 27686.

Johan W. S. Vlaeyen*, Stephen J. Morley†, * University of Leuven, Leuven, Belgium, † University of Leeds, Leeds, UK.