The article by Sullivan et al.1 presents an impressive synthesis of research findings and issues/controversies regarding catastrophizing and pain, and possible conceptual models of this relation. This article is certain to be provocative in stimulating new directions in research and conceptualization. We would like to commend the authors for this contribution to the literature that necessarily drew from studies of disparate patient and nonpatient populations and theoretical orientations. It provides us with an opportunity to consider the origin of the construct of catastrophizing, how research on pain catastrophizing might be broadened by greater focus on construct considerations, and the future research directions most likely to advance theory and to benefit people with pain.
Origins of the concept of catastrophizing
We would like first to ask the reader to ponder the question of what the term "catastrophizing" means. The term catastrophizing was used by Albert Ellis, the founder of rational-emotional therapy, almost four decades ago.2 Ellis2 gave the following example of catastrophizing: "How terrible the situation is; I positively cannot stand it!" Later, Beck et al.3,4 discussed catastrophizing in terms of dwelling on the worst possible outcome of any situation in which there is a possibility for an unpleasant outcome. Examples of catastrophizing given by Beck et al.4 include the following: (1) during an airplane flight, a woman dwells on the possibility of the plane's crashing and her being killed; and (2) a college student taking an examination is preoccupied with the possibility of failing and consequently flunking out of college. Such thoughts are tied to the perception of oneself as vulnerable and as being subject to danger over which one has insufficient control. A slightly different definition of catastrophizing has been used in studies of worry: perceiving progressively worse and worse outcomes to a specific worry.5
Catastrophizing is related to anxiety (e.g., anxiety is associated with the tendency to overemphasize the probability of a catastrophic outcome and the possible consequences of such an outcome). During the past decade, cognitive models of panic disorder have emphasized the role of catastrophizing in panic attacks. According to these models, individuals with panic disorder interpret anxiety-produced bodily sensations (e.g., palpitations, breathlessness) in a catastrophic fashion (e.g., signs of impending insanity, death, loss of control).6 A similar model has been applied to hypochondriasis (an unrealistic interpretation of bodily sensations that leads to preoccupation with the fear of, or belief that one has, a serious disease, despite medical reassurance).7
What do pain catastrophizing scales really measure?
How do these definitions of catastrophizing, developed in cognitive models of anxiety and other psychologic disorders, fit with the available measures of pain catastrophizing? Table 1 lists items on these measures. The first measure developed to assess pain-related catastrophizing was the Cognitive Error Questionnaire.8 The Cognitive Error Questionnaire consists of 24 vignettes, each followed by a dysphoric cognition. Respondents indicate how similar the cognition is to how they would think in that situation. Six items form the catastrophizing scale (the others reflect other cognitive errors: overgeneralization, personalization, and selective abstraction). The Cognitive Error Questionnaire has not been widely used; the majority of research in the area has involved the six-item catastrophizing scale of the Coping Strategies Questionnaire (CSQ).9 However, Sullivan et al.10 sought to develop a broader measure of pain catastrophizing and in 1995 published the Pain Catastrophizing Scale (PCS). This measure has three subscales: rumination, helplessness, and magnification. The helplessness scale includes five of the six CSQ Catastrophizing scale items, plus one additional item.
Although not labeled as a pain catastrophizing measure, the Pain Anxiety Symptoms Scale,11 developed to measure fear and anxiety associated with pain, has two subscales that are closely related to catastrophizing: fear (fearful thoughts and ruminations about the consequences of pain) and cognitive anxiety. These two subscales are highly correlated with one another and with the CSQ Catastrophizing scale (r = 0.66-0.74) and with a measure of trait anxiety.11,12
Examining the various items in Table 1, the question arises as to whether the definition of catastrophizing typically used by psychologists is fully exemplified in the most commonly used pain catastrophizing scales. Although the example given by Ellis2 is similar to items on the CSQ Catastrophizing scale and helplessness scale of the Pain Catastrophizing scale, do items on these scales (reflecting appraisals that the pain is terrible and intolerable) fully encompass the construct of catastrophizing? Should a measure of pain catastrophizing also include items that depict worst-case scenarios for patients with chronically painful conditions; for example: "I might end up paralyzed," "I might become totally disabled," "I will lose my job and not be able to support my family"? Such items might increase the face, content, and construct validity of pain catastrophizing scales by capturing an additional dimension of catastrophizing. As can be seen in Table 1, the Cognitive Errors Questionnaire catastrophizing scale items and some of the items on the Pain Anxiety Symptoms Scale fear subscale come closer than the other scale items to these catastrophic thoughts. At this stage of our knowledge acquisition, it may be useful to broaden the area of inquiry to determine whether our current measures of pain-related catastrophizing adequately reflect the concept of catastrophizing as it was originally conceived and as it has been applied to anxiety and other psychologic disorders. The addition of items reflecting worry about worst possible outcomes to existing or new measures would appear to be a potentially fruitful area for future research.
Catastrophizing: stable personality disposition or situational response?
The review by Sullivan et al.1 indicates that measures of pain-related catastrophizing consistently are associated positively with measures of physical and psychosocial disability among patients with a variety of pain conditions. The evidence also suggests that catastrophizing represents something other than solely a manifestation of depression.13 However, before we can attempt to apply existing theoretical frameworks to explain the relation of catastrophizing to pain and adjustment, it may be useful to consider whether pain-related catastrophizing is a stable personality disposition or a situational response that varies over time, elicited by certain stimuli/conditions. That is, is catastrophizing related to a dispositional (e.g., schema- or personality-based) construct that is present across situations and acts as a filter through which one develops appraisals of pain as a threat with which one cannot cope? Or is catastrophizing a response that varies according to situational circumstances? This question is touched upon by Sullivan et al.1 in the section on "Stability and Situational Specificity of Catastrophizing." We would like to highlight this distinction here because clarification may prove to be heuristic in further research and delineation of the construct of catastrophizing. This may, in turn, result in certain theoretical models emerging as more applicable than others or, alternatively, necessitate the creation of a new model altogether. In the next two sections, we consider each possibility.
Catastrophizing as a stable, dispositional characteristic
In the review by Sullivan et al.1 and more generally in the pain literature, the term "catastrophizer" is used commonly to refer to one who catastrophizes, which suggests that we believe it to be a person-based construct. In support of this argument, a study examining the stability of the CSQ Catastrophizing scale over time reported a high (0.81) 6-month test-retest stability coefficient.14 Similarly, the Pain Catastrophizing Scale has been found to have a high test-retest correlation (r = 0.75) across a 6-week period.10 With regard to exploring the possibility that catastrophizing is a stable, person-based characteristic, it would seem important to evaluate the relation of catastrophizing to other global or relatively stable dispositional variables, such as neuroticism. Although Sullivan et al.1 conclude that catastrophizing is distinct from neuroticism, there is evidence that measures of catastrophizing and measures of neuroticism are highly correlated. For example, one study15 found the CSQ catastrophizing scale to be highly correlated with the NEO16 Neuroticism scale. Another study found that significant associations between CSQ catastrophizing scores and measures of physical and psychosocial disability disappeared when scores on the NEO Neuroticism scale were controlled.17 These results suggest that the CSQ catastrophizing scale provides similar information, as does this measure of neuroticism. Persons high in neuroticism may be prone to catastrophizing, and catastrophizing may mediate the relation between neuroticism and pain intensity ratings.18
Research investigating the relation of pain catastrophizing measures to neuroticism and other personality characteristics (e.g., hypochondriasis, somatization, anxiety, worry, and negative affectivity) would seem to hold considerable potential for increasing our understanding of what dispositional traits or psychologic disorders are associated with catastrophizing responses to pain. For example, one investigation found that patients with psychiatric diagnoses of somatization disorder and hypochondriasis had higher scores on a measure of catastrophizing interpretations of bodily complaints than did behavioral medicine center patients without these diagnoses.19 It might also prove informative to explore the relation between pain catastrophizing scales and measures of worry, such as the Penn State Worry Questionnaire,20 a self-report instrument for the trait assessment of clinically significant, pathologic worry. Studies of people without chronic pain have shown that worry is related to catastrophizing thoughts, and that chronic worriers believe that worry has an adaptive function (e.g., to help avoid failures, mistakes, and catastrophes).5 Worry and catastrophizing have also been found to be associated with a sense of personal inadequacy and lack of confidence in problem-solving skills,5 which suggests other areas to explore in research and treatment involving patients with chronic pain problems.
Another potentially interesting area for study might be the relation of pain catastrophizing to pain-related beliefs, which are generally viewed as fairly stable, although amenable to change (e.g., through education). Similar to catastrophizing, pain beliefs are strongly associated with physical and psychosocial disability. Sullivan et al.1 state that, "in a recent study by Turner et al.21 it was found that pain beliefs mediated the relation between catastrophizing and disability, suggesting that catastrophizing may influence disability indirectly, through other pain appraisals." Although the study did not involve tests of mediation, the results suggested that catastrophizing may explain unique variance in depression, but not in physical disability, beyond that accounted for by age, sex, pain intensity, and measures of beliefs and coping. The findings of the study are consistent with a view of catastrophizing as significantly associated with one's beliefs about one's pain.
Some of the beliefs most importantly associated with physical disability seem to be views of one's pain as disabling, views that pain is a signal of physical harm, and belief that one has little control over pain.21 Factor analytic studies have shown that the belief that one can control or decrease pain loads onto the same factor as catastrophizing, and that control and catastrophizing are inversely related.22 Such beliefs may influence cognitive (e.g., thoughts such as those listed in Table 1) and behavioral (e.g., continuing vs. discontinuing activity when in pain) responses to pain. In turn, these responses may influence subsequent pain experience and physical and psychosocial disability. For example, belief that pain is a signal of physical harm and that one has little control over pain may result in catastrophizing thoughts and verbalizations during pain flare-ups. Catastrophizing thoughts may lead to choices to rest and avoid activity, which can lead to deconditioning and failure to return to work, which may in turn lead to further pain, depression, and additional problems. Catastrophizing verbalizations may elicit sympathetic offers of help and suggestions to rest from significant others, in turn leading to decreased patient participation in customary activities and perhaps reinforcing the patient view that pain is disabling and a signal of harm.
Catastrophizing as a variable, situation-based response to pain
As an alternative to the view that catastrophizing is a disposition that is fairly stable over time, it is useful to consider the possibility that catastrophizing is a response to pain that varies over time and is determined by situational factors. As Sullivan et al.1 note, catastrophizing has been shown to change with targeted interventions. This raises the question of whether there are contextual determinants of catastrophizing and, if so, what they are and how they vary across time and persons with pain. It should be noted that variations in either internal (e.g., sensory or affective states) or external (e.g., environmental cues) conditions may provide the "context" in which catastrophizing thoughts or behaviors are frequently observed. Therefore, for example, it may be the case that even psychologically healthy (e.g., non-neurotic, nondepressed) individuals who do not typically "catastrophize" may have catastrophizing thoughts when experiencing severe pain (an internal sensory state). With regard to external factors, Sullivan et al.1 propose in the "Coping Model" section that catastrophizing may be used instrumentally, via exaggerated verbal reports and pain behaviors, to elicit social support. From this perspective, it would be expected that the extent to which one engages in catastrophizing may change over time as a function of stimulus cues and social responses (e.g., solicitous or punishing) present in the individual's environment.
However, complex interactions may affect the relations between catastrophizing and psychologic distress or physical disability. For example, suppose that communications of catastrophizing are observed only in the presence of certain environmental conditions (e.g., a sympathetic spouse) and when one holds the belief of low control over pain. Such a finding would be consistent with the view that catastrophizing functions as a coping behavior and would lead to a prediction that catastrophizing fluctuates over time, depending on environmental conditions and the internal states of the person (e.g., certain beliefs or mood states). Unfortunately, very little is known about the various conditions in which catastrophizing may increase, decrease, or even disappear in response to pain. Daily process methodologies that have been used to evaluate daily fluctuations of coping responses (e.g., studies by Affleck et al.23,24) and direct observational methodologies (e.g., as used by Romano et al.25-27) may be helpful in addressing these types of questions.
Catastrophizing as both dispositional and situation-influenced
As a final alternative to consider, catastrophizing may be found to represent a fairly stable tendency in some individuals, but with manifestations varying in intensity as a function of certain stimuli/conditions. Analogously, it has been argued that patients with panic disorder have a tendency to catastrophically misinterpret bodily sensations, even when they are not anxious, and that this relatively enduring cognitive trait is amplified when the patient is in an anxious state.6 Similarly, it is possible that certain individuals have a tendency to catastrophize in response to pain but that this tendency is amplified under certain internal conditions, such as when the individual is depressed or anxious or highly stressed, and in certain situations, such as in the presence of a solicitous spouse.
Some methodological considerations
Difference between pain experience and pain report
We comment on more specific aspects of the article by Sullivan et al.1 First, we point out that the authors use the terms "pain" and "pain experience" when it would be more accurate to use terms such as "measures of pain intensity" or "patient ratings of pain intensity." For example, the first sentence in the "Catastrophizing and Pain" section states, "One of the most consistent findings has been that catastrophizing is associated with heightened pain experience." Similar statements are found in other sections of this article. It must be kept in mind that we cannot know the pain experience of others; we can only observe their behaviors (including verbal reports of pain and ratings of pain intensity). A more accurate conclusion would be that self-report measures of the tendency to have catastrophizing thoughts when in pain are correlated with self-report measures of pain intensity.
Correlation does not prove causation
These correlational studies do not prove causal relations. Attention may be called to the first sentence of the fourth paragraph, which states that "...the tendency to 'catastrophize' during painful stimulation contributes to more intense pain and increased emotional distress."1 We must not make the mistake of assuming that catastrophizing "contributes to" more intense pain based on statistical associations between individual ratings of pain intensity and ratings of catastrophizing thoughts. Shared method variance and response/reporting biases may explain these associations to some extent, and causal relations between catastrophizing and pain are unknown. As the authors acknowledge, individuals may be more likely to catastrophize when pain is more intense.
Similarly, further research is needed before concluding that "the pattern of findings appears to support the causal or, at least, antecedent status of catastrophizing" (statement in the Catastrophizing and Pain section).1 Three articles are cited in support of this statement. In one article, students classified as catastrophizers reported more pain than did noncatastrophizers in an experimental pain procedure, and patients classified as catastrophizers reported more pain during an electrodiagnostic procedure than did noncatastrophizers.10 Although these findings are consistent with a model of catastrophizing as causing increased pain (or at least, higher pain ratings), studies in which pain is assessed after experimental manipulations of catastrophizing would provide even stronger support for such a model.
In another of the cited studies, undergraduate students completed a pain catastrophizing measure before a dental procedure involving mild to moderate pain.28 Individuals were assigned randomly to two conditions in which they were asked to write just before the procedure. In the disclosure condition, they were asked to write about the thoughts and feelings they typically experienced during dental treatment, focusing on the aspects of dental treatment they found to be most distressing. In the control condition, they were asked to describe their activities from the previous day. "Catastrophizers" rated their pain during the procedure significantly higher than did "non-catastrophizers" in the control condition, but the two groups did not differ in the disclosure condition. Furthermore, catastrophizers in the control condition reported significantly more pain than did catastrophizers in the disclosure condition. This is a fascinating study; however, because there was not a convention with no writing intervention prior to the dental procedure, we do not know whether catastrophizers would have rated their pain as higher than noncatastrophizers under normal dental procedures. Furthermore, the comparisons between catastrophizers and noncatastrophizers in this study did not account for age, extent of periodontal disease, gender, and other factors that might have influenced pain ratings.
In the third study14 cited as evidence that catastrophizing plays a causal (or antecedent) role in pain, patients with rheumatoid arthritis completed questionnaires assessing catastrophizing, physical disability, depression, and pain intensity at time 1 and 6 months later (time 2). Catastrophizing at time 1 explained a statistically significant proportion of the variance in pain ratings, physical disability, and depression scores at time 2, even after controlling for time 1 scores on these outcome measures. However, the amount of variance explained was very modest and of questionable clinical significance. Further research is needed to shed light on sequential and causal relations between catastrophizing, pain, physical disability, and depression.
In summarizing the current state of knowledge, Sullivan et al.1 have provided a strong foundation to build on in future investigations involving catastrophizing and pain. There are many exciting possibilities for exploration. In this article, we outlined potential directions we believe to hold the most promise to advance our under-standing of the nature and function of catastrophizing. These include the following: (1) questioning how current measures of pain catastrophizing relate to the original construct of catastrophizing and whether current measures should be broadened; (2) examining how pain catastrophizing relates to such stable, dispositional characteristics as neuroticism and worry; (3) exploring sequential relations and causal pathways between personality characteristics, pain beliefs, catastrophizing, pain reports, and physical and psychosocial disability; and (4) identifying conditions under which catastrophizing may function as a variable response to pain. It is our hope that these suggestions will lead to even greater debate and inquiry into the nature of a construct that we do know to be associated importantly with patient reports of pain and with pain-related physical and psychosocial dysfunction.
Acknowledgments: The authors thank Samuel Dworkin, D.D.S., Ph.D., University of Washington, Departments of Oral Medicine and Psychiatry and Behavioral Sciences, Seattle, Washington, U.S.A., for helpful comments on a previous version of this manuscript.
1. Sullivan MJL, Thorn B, Haythornthwaite JA, et al. Theoretical perspectives on the relation between catastrophizing and pain. Clin J Pain
2. Ellis A. Reason and emotion in psychotherapy.
New York. Lyle Stuart, 1962.
3. Beck AT. Cognitive therapy and the emotional disorders.
New York: The New American Library, 1976.
4. Beck AT, Emery G, Greenberg RL. Anxiety disorders and phobias.
New York: Basic Books, 1985.
5. Davey GCL, Levy S. Catastrophic worrying: personal inadequacy and a perseverative iterative style as features of the catastrophizing process. J Abnorm Psychol
6. Clark DM. A cognitive model of panic attacks. In: Rachman SJ, Maser JD, eds. Panic: psychological perspectives.
Hillsdale, NJ: Lawrence Erlbaum; 1988:71-89.
7. Warwick HMC, Salkovskis PM. Hypochondriasis. In: Scott J, Williams JMG, Beck AT, eds. Cognitive therapy in clinical practice: an illustrative casebook.
New York: Routledge; 1989:78-102.
8. Lefebvre MF. Cognitive distortion and cognitive errors in depressed psychiatric and low back pain patients. J Consult Clin Psychol
9. Rosenstiel AK, Keefe FJ. The use of coping strategies in chronic low back pain patients: relationship to patient characteristics and current adjustment. Pain
10. Sullivan MJL, Bishop SR, Pivik J. The pain catastrophizing scale: development and validation. Psychol Assess
11. McCracken LM, Zayfert C, Gross RT. The Pain Anxiety Symptoms Scale: development and validation of a scale to measure fear of pain. Pain
12. McCracken LM, Gross RT. Does anxiety affect coping with chronic pain? Clin J Pain
13. Geisser ME, Robinson ME, Keefe FJ, et al. Catastrophizing, depression and the sensory, affective and evaluative aspects of chronic pain. Pain
14. Keefe FJ, Brown GK, Wallston KA, et al. Coping with rheumatoid arthritis: catastrophizing as a maladaptive strategy. Pain
15. Drossman DA, Leserman J, Li Z, et al. Effects of coping on health outcomes among women with gastrointestinal disorders. Psychosom Med
16. Costa P, McCrae R. NEO Personality Inventory manual.
Odessa, FL: Psychological Assessment Resources, 1985.
17. Martin MY, Bradley LA, Alexander RW, et al. Coping strategies predict disability in patients with primary fibromyalgia. Pain
18. Affleck G, Tennen H, Urrows S, et al. Neuroticism and the pain-mood relation in rheumatoid arthritis: insights from a prospective daily study. J Consult Clin Psychol
19. Rief W, Hiller W, Margraf J. Cognitive aspects of hypochondriasis and the somatization syndrome. J Abnorm Psychol
20. Meyer TJ, Miller ML, Metzger RL, et al. Development and validation of the Penn State Worry Questionnaire. Behav Res Ther
21. Turner JA, Jensen MP, Romano JM. Do beliefs, coping, and catastrophizing independently predict functioning in patients with chronic pain? Pain
22. Keefe FJ, Caldwell DS, Queen KT, et al. Pain coping strategies in osteoarthritis patients. J Consult Clin Psychol
23. Affleck G, Urrows S, Tennen H, et al. Daily coping with pain from rheumatoid arthritis: patterns and correlates. Pain
24. Affleck G, Tennen H, Keefe FJ, et al. Everyday life with osteoarthritis or rheumatoid arthritis: independent effects of disease and gender on daily pain, mood, and coping. Pain
25. Romano JM, Turner JA, Jensen MP, et al. Chronic pain patient-spouse behavioral interactions predict patient disability. Pain
26. Romano JM, Jensen MP, Turner JA, et al. Chronic pain patient-partner interactions: further support for a behavioral model of chronic pain. Behav Ther
27. Romano JM, Turner JA, Friedman LS, et al. Sequential analysis of chronic pain behaviors and spouse responses. J Consult Clin Psychol
28. Sullivan MJL, Neish N. The effects of disclosure on pain during dental hygiene treatment: the moderating role of catastrophizing. Pain