Pain, sleep and catastrophizing: The conceptualization matters Comment on Wilt JA et al. “A multilevel path model analysis of the relations between sleep, pain, and pain catastrophizing in chronic pain rehabilitation patients”

In this issue of the Scandinavian Journal of Pain, Joshua Wilt and colleagues present the results from a study examining the relationship between pain, sleep and pain catastrophizing in 50 patients with chronic pain undergoing an interdisciplinary treatment programme at a rehabilitation clinic [1]. The study is notable for several reasons. The interrelation between sleep and pain is an intriguing issue which indeed requires attention in research. Co-occurrence of sleep and pain is strikingly common, with over half of pain patients reporting concurrent sleep disturbances [2,3]. There is growing evidence that treating sleep difficulties may be vital for achieving improvements in patients with co-existing pain and sleep difficulties (e.g., [4,5]). However, it should be noted that purely targeting sleep has fairly small effects on pain outcomes (for a review, see [6]). Consequently, there are still areas to explore concerning the nature of the relationship between pain and sleep, and considering the role of catastrophizing may add valuable knowledge.

The longitudinal study by Wilt and colleagues makes a contribution to the field as it investigates the nature of the interrelation between pain and sleep over time and even includes possible mediating factors (see Fig. 1 for the models that were tested). Indeed, focusing on mediation in this setting is important as it may reveal central targets for intervention. Focusing on catastrophizing as a possible critical factor in this context is highly relevant, because catastrophizing is recognized as a powerful psychological determinant of pain-related disability (e.g., [7, 8, 9, 10, 11, 12]). It is surprising then that so little research has focused on the role of catastrophizing in the sleep-pain relationship, and we therefore welcome the current study. Moreover, scrutinizing the nature of the interrelation between pain, sleep and catastrophizing is of clinical interest. For instance, patients with high levels of catastrophizing are often not helped by current treatments (see e.g., [13,14]), and it is essential to find ways of facilitating reductions in catastrophizing. The results from the current study indicate that targeting sleep may be one approach for reducing catastrophic thinking. Nevertheless, the interdisciplinary programme applied in the study succeeded in reducing levels of catastrophizing without explicitly targeting sleep. Even though the authors do not highlight it, they achieved impressive reductions in catastrophizing (>15 points on the PCS), thus pointing to its clinical significance.

Fig. 1

Mediation models tested in the study by Wilt et al. [1].

Notwithstanding these features, there are several issues concerning the role of catastrophizing in the sleep-pain relationship which are not brought up by the authors. A critical concern is the direction of the association. The authors were interested in looking at mediating factors, focusing on pain and sleep difficulties as possible mediators. However, in all models that were tested, sleep difficulties were proposed to precede pain.This disregards the reciprocity in the sleep-pain relationship; an increase in sleep difficulties may indeed predict pain, but the reverse may also be true, i.e. increases in pain may predict sleep problems [2,15], which is not taken into consideration in their mediation models. Another issue regards the role of catastrophizing as a mediator. In the current study, catastrophizing was namely considered as an outcome and therefore it was measured before and after the intervention. The possibility that catastrophizing might serve as a mediator was not even tested, and the authors ignore the issue altogether. This is quite surprising since catastrophizing is conceptualized as a principal mediator in models, e.g., the fear-avoidance model [16], as well as in the treatment of pain [14]. This raises the question of how the authors conceptualize catastrophizing.

Considering the amount of research confirming the link between catastrophizing and negative outcomes, surprisingly little effort has been put into theorizing about catastrophizing. It is frequently assessed and included in studies examining relations between different concepts, without considering how it relates to other concepts. Moreover, this often appears to be done with little or no reflection upon what catastrophizing really is; catastrophizing becomes what the Pain Catastrophizing Scale (PCS; [17]) measures. Recently however, we proposed a new theoretical framework regarding catastrophizing, which builds on and adds to earlier theories [7]. We conceptualized it as a form of negative repetitive thinking which is abstract, intrusive and difficult to disengage from. When seen from this light, the extensive work on repetitive negative thinking and worry in the psychology literature becomes most relevant and we believe it could be of great assistance in guiding the research in the pain field. In this framework, theories and findings from the contemporary psychological literature are integrated into the pain area, building specifically on theories around repetitive negative thinking.

Repetitive negative thinking has been defined as “a style of thinking about one’s problems (current, past, or future) or negative experiences (past or anticipated) that is repetitive, at least partly intrusive, and is difficult to disengage from” [18]. It has been argued that repetitive negative thinking is an avoidant coping strategy [19], which is characterized by an abstract content [20]. According to this perspective, patients with high levels of pain-related fear and anxiety might engage in catastrophizing as a way of reducing aversive psychological and physiological aspects of the fear response [21].To highlight the similarities between catastrophizing and other forms of repetitive negative thinking, we coined the term “catastrophic worry”, to relate it to both the emotional and cognitive aspects. The term itself is not new, as it has been used to describe more intense forms of worry (e.g., [22] ). However, in the pain area the term catastrophizing is entirely predominant. We suggest a resumption of the term catastrophic worry to link it to contemporary psychological literature, and to emphasize the close links between catastrophizing, anxiety, and emotional processing [23].

With the conceptualization of catastrophizing as repetitive negative thinking in mind, it is understandable that the first model in the current study was not supported: sleep was not a mediator between catastrophizing and pain. If our model of catastrophic worry is correct, then the intrinsic function of catastrophizing is to reduce aversive physiological reactions such as bodily arousal. If that is the case, then catastrophizing might in fact decrease bodily arousal that interferes with sleep, which in turn supports the idea that sleep is not a precursor of pain difficulties. The second model, however, which suggested an indirect path from sleep difficulties to catastrophizing via pain, turned out to be significant, and this is in line with our conceptualization. If catastrophizing is regarded as a behavioural response to a trigger (in this case pain), then pain should indeed precede catastrophizing, just as in the current study. In other words, catastrophizing is an immediate response triggered by a pain stimulus.

In our conceptualization of catastrophic worry, it is put forward as a transdiagnostic construct, identified across disorders (for a review, see [20]). The transdiagnostic perspective claims that one explanation of the high comorbidity between different problems, such as pain and sleep, is that they share essential maintaining processes [24]. According to this view, catastrophic worry is a critical process that contributes to pain, but also to the sleep disturbances; the same process, but with different co-morbid problem descriptions. This provides one explanation of the large overlap between problems with pain and sleep.

Considering catastrophic worry as a transdiagnostic also sheds light on interventions since targeting catastrophizing might have benefits for the pain problem as well as for sleep disturbances. Wilt and colleagues indeed mention the possibility of applying hybrid treatments, in which both pain and sleep are targeted. We agree that this is a promising idea, but would like to underscore the importance of also targeting catastrophizing, as earlier research has shown that high catastrophizing is a predictor of poor treatment outcome (for a review, see [25]). Reducing catastrophizing then, especially for patients with high levels, is a viable approach for treating both pain and sleep simultaneously.

In sum, the study by Wilt and colleagues contributes to the field by shedding light on the fascinating interrelationship between sleep and pain, in particular by pointing out sleep interventions as a potential means of attaining improvements for patients with comorbid pain and sleep disturbances. However, purely targeting sleep has shown small effects on pain outcomes [6], which highlights the need for exploring and targeting common maintaining processes in pain and sleep problems. Furthermore, we claim that a firm theoretical framework may guide the choice of potential mediators, and the authors do not discuss the theoretical underpinnings of their choices. We suggest that considering catastrophic worry as a transdiagnostic process which may explain co-occurring pain and sleep disturbances may inspire future research in this area.