Cognitive-Behavioral Therapy: Nature and Relation to Non-Cognitive Behavioral Therapy

BETH-00620; No of Pages 19; 4C: Available online at www.sciencedirect.com ScienceDirect

Behavior Therapy xx (2016) xxx–xxx

www.elsevier.com/locate/bt

Cognitive-Behavioral Therapy: Nature and Relation to Non-Cognitive Behavioral Therapy

Lorenzo Lorenzo-Luaces John R. Keefe Robert J. DeRubeis University of Pennsylvania

there is any kind of contribution of the “cognitive” in Since the introduction of Beck’s cognitive theory of emotional cognitive-behavioral therapy. disorders, and their treatment with psychotherapy, cognitive- Despite debate regarding their active treatment behavioral approaches have become the most extensively components as well as working mechanisms, CBTs researched psychological treatment for a wide variety of continue to be the most widely studied forms of disorders. Despite this, the relative contribution of cognitive to therapy (Hofmann, Asmundson, & Beck, 2013). A behavioral approaches to treatment are poorly understood uniquely appealing aspect of CBTs is that their theo- and the mechanistic role of cognitive change in therapy is ries of therapeutic change comport well with most widely debated. We critically review this literature, focusing modern conceptualizations of psychopathology. In on the mechanistic role of cognitive change across cognitive this review, we attempt to reconcile skepticism and behavioral therapies for depressive and anxiety disorders. regarding the relative contribution of CT strategies to BT, as well as the mechanisms that account for their efficacy. First, we provide a very brief historical over- Keywords: cognitive-behavioral therapy; cognitive theory; view of the origins of CBT and discuss the support for psychotherapy processes; depression; anxiety the cognitive vulnerability models to depression and anxiety disorders. We discuss methodological chal- THE ORIGIN OF COGNITIVE-BEHAVIORAL THERAPIES lenges in psychotherapy research that have impeded (CBTs) as a family of interventions can be traced to a more thorough understanding of the relative con- the advent of behavioral treatments for psychopa- tributions of cognitive to behavioral techniques. We thology in the 1950s and, later, the so-called “ ” – then focus most of our discussion on research on the cognitive revolution of the 1950 1960s (Dobson, cognitive mechanisms of change in CT, BT, and CBTs 2009). Consequently, CBTs blend techniques that are for depression and anxiety disorders. emphasized in behavioral therapies (BTs) and cogni- We use the terms cognitive therapy (CT) and tive therapies (CTs). However, there remains skepti- cognitive techniques to refer to behaviors therapists cism regarding the relative contributions of CT engage in that are targeted towards changing the strategies to BT strategies in promoting symptom content or process of thoughts, inferences, inter- change within the CBTs (Longmore & Worrell, pretations, cognitive biases, and cognitive schemas.1 2007). Additionally, critics have asserted that changes in thinking are not mechanisms of change in CBTs (e.g., Kazdin, 2007), calling into question whether 1 The terms “cognitive therapy” (CT) and “cognitive-behavioral therapy” (CBT) are often used interchangeably. We believe this is Correspondence regarding this article should be addressed to somewhat unfortunate in that it might be informative to reserve the Lorenzo Lorenzo-Luaces, University of Pennsylvania, Department of Psychology, 3720 Walnut Street D20, Philadelphia PA 19104; term CT to a set of interventions within the broader family of CBTs “ ” e-mail: [email protected]. that are more purely cognitive in nature. However, throughout the article, when we refer to findings in studies of CT or CBT, we are 0005-7894/© 2016 Association for Behavioral and Cognitive Therapies. adhering to the label the study authors use. Additionally, we use CBTs, Published by Elsevier Ltd. All rights reserved. in plural, to refer to the family of cognitive-behavioral therapies.

These interventions can include Socratic question- CT emerged in the context of the so-called cogni- ing, examining the evidence for and against beliefs, tive revolution (Beck, 1991; O’Donohue, Ferguson, cognitive restructuring, and adopting alternative & Naugle, 2003) from the writings of Ellis (1962), core beliefs. We use the terms behavior therapy who described a form of therapy known as rational- (BT) and behavioral techniques to refer to behav- emotive therapy, and Beck (1963). The cognitive iors therapists engage in that are targeted towards models of Ellis and Beck focused on inferential errors a change in observable behavior, including in vivo leading to maladaptive views of one’s self, world, exposure, imaginal exposure, and activity sched- and the future. According to Beck, cognitive biases uling. We use cognitive-behavioral therapies in the and maladaptive cognitive content are the product of plural (CBTs) to refer to the family of interventions the activation of cognitive schemas that typically to which CT and BT belong, and in the singular, develop early in life. Unlike BTs, which were initially CBT, to refer to a treatment package that combines successful in specific phobias and circumscribed cognitive and behavioral techniques. By cognitive anxieties, CTs were focused on depressotypic pre- change, we refer to changes in the content of sentations and more generalized anxiety. Early in thoughts, inferences, interpretations, and cognitive his writing, Beck recognized that his cognitive theory biases. By behavioral change, we refer to changes of psychopathology, which gave a central role to in behavior, such as increasing the frequency of cognition in the etiology of disorder, contrasted selected behaviors (e.g., approaching feared stimuli, with behavioral theories of psychopathology. In his engaging with pleasurable activities) or decreasing highly cited article, “Cognitive Therapy: Nature the frequency of other behaviors (e.g., safety and Relation to Behavioral Therapy,” Beck (1970) behaviors). We include in our paper a discussion of described important differences between the theories issues related to the conceptualization and measure- that underlie BT and CT while recognizing areas of ment of cognitive vs. behavioral interventions as well overlap in the performance of the therapies. Similar- as cognitive vs. behavioral mechanisms of change ities include that both therapies deal with issues in and conclude with a summary and with recommen- the present, are symptom-focused, and require active dations for future research. therapist contribution. Cognitive Therapy: Nature and Relation to Beck (1970) recognized differences between behavioral and cognitive approaches. He applied Behavioral Therapy the principles of his then nascent cognitive theory to Behavioral therapies emerged in the 1950s–1960s account for the mechanisms of action of systematic (O’Donohue & Noll, 1995). The behavioral desensitization, a BT. He concluded that the cog- models emphasized maladaptive learning and self- nitive model “provides a greater range of concepts sustaining behaviors as key to the maintenance of for explaining psychopathology as well as the psychopathology. This made behavioral change mode of action of therapy.” That is, Beck made a the obvious target of treatment, an approach that distinction between the nature of the therapeutic was in stark contrast to the previously dominant interventions (i.e., cognitive vs. behavioral) and psychoanalytic models. Under psychoanalysis, their working mechanisms in providing a cognitive pathological behavior was seen to reflect dysfunc- account of the effects of a behavioral intervention. tion in underlying psychic structures. Behavioral Beck’s paper would become one of the early reflec- change was thus seen as surface-level “symptom tions on the relative contributions of cognitive to reduction” that did not address underlying prob- behavioral strategies and the relevant mechanisms lems. BTs proved very effective, particularly in the of change. Although Beck has provided two up- treatment of phobias and more circumscribed dates to his cognitive model (Beck, 1996; Beck & states of anxiety. Principles of associative learning Haigh, 2014), its basic tenets remain largely intact: were used to account for the efficacy of these that the distinction between different forms of psy- interventions. To the behaviorists, learning had chopathology can be traced to differences in the a specific meaning: an overt change in behavior locus of the cognitive pathology and that cognitive (e.g., approaching a previously avoided stimulus) change, regardless of how this change is achieved, is in the absence of symptoms (e.g., without display- integral to symptom change. ing the fear reaction). This definition avoided “mentalistic” terms. Although early behavioral Cognitive Vulnerability to Depression models featured theoretical accounts focused on and Anxiety associative learning, nonassociative learning, in- Basic research supports the notion that cognitive cluding habituation, was also seen as important. vulnerabilities confer risk to the onset and main- Newer behavioral models also focus on inhibitory tenance of psychopathology (see Mathews & learning (Craske et al., 2008). MacLeod, 2005). Attentional biases to threatening

Please cite this article as: Lorenzo Lorenzo-Luaces, et al., Cognitive-Behavioral Therapy: Nature and Relation to Non-Cognitive Behavioral Therapy, Behavior Therapy (2016), http://dx.doi.org/10.1016/j.beth.2016.02.012 cbt: nature and relation to non-c bt 3 stimuli, along with overestimation of threat, have (2000) reported that inducing bias in the interpre- been implicated in the etiology of anxiety disor- tation of ambiguous information as threatening ders (Bar-Haim, Lamy, Pergamin, Bakermans- leads to increases in state anxiety. In another study, Kranenburg, & Van Ijzendoorn, 2007). Biases MacLeod et al. (2002) manipulated attention to associated with depression include difficulties dis- emotionally negative information. After a stressor engaging from negative material, sustained or sym- task, participants who had had their attention metrical attention to negative, relative to positive, manipulated towards negatively valenced stimuli stimuli (Kircanski & Gotlib, 2015), negative biases showed greater anxiety and depression than par- in the appraisal of life events (Mehu & Scherer, ticipants in the control group. 2015), symmetric memory for negative vs. neutral If cognitive biases increase the risk for depression or positive information (Kircanski & Gotlib, and anxiety states, it follows that strategies that 2015), and negative schemas about the self that address these biases should result in a reduction of foster maladaptive and negative thinking (Beck & risk. This hypothesis has support in basic research Haigh, 2014). on emotion regulation. Webb, Miles, and Sheeran Overall, existing research is supportive of cog- (2012) conducted a meta-analysis of 306 compar- nitive vulnerability models of affective disorders. isons of emotion regulation strategies. Strategies For example, cross-cultural research consistently that focused on cognitive change were estimated to suggests that, on average, healthy individuals have be the most consistently effective ways of regulating a bias towards optimistic thinking that is not emotions (d = 0.36). Strategies aimed at helping found in individuals who are depressed and who, individuals adopt more rational perspectives, as is instead, have a bias towards more negative thinking encouraged in CT, were associated with the largest (Mezulis, Abramson, Hyde, & Hankin, 2004). effect (d = 0.45). Providing even more support for Similarly, in a meta-analytic review of 172 studies cognitive theories, studies that examine the bio- examining biases towards threatening stimuli, logical vulnerabilities to negative emotional states Bar-Haim et al. (2007) found that anxious partic- suggest that, at the phenomenological level, biolog- ipants are biased to attend to threatening stimuli, ical vulnerabilities render individuals more likely to relative to nonanxious participants (d = 0.45). The experience negative emotional states by interfering causal role of these cognitive vulnerabilities, par- with their abilities to engage in cognitive reapprais- ticularly in depression, has been questioned partly al strategies (Firk, Siep, & Markus, 2013; Lemogne because most of the early research on this matter et al., 2011). was correlational in nature (see Ingram et al., More research is needed that characterizes more 2006). Findings from prospective studies, however, precisely the nature of the cognitive biases impli- also support cognitive models. For example, daily cated in depression and anxiety, especially research fluctuations in negative automatic thoughts have that is experimental. The relationship between been found to predict subsequent negative mood, affective disorders and cognition is bidirectional, even controlling for prior levels of automatic thoughts which must also be accounted for in theories of (Wenze, Gunthert, & Forand, 2007; Wenze et al., psychopathology. However, given the amount of 2010). Negative dysfunctional attitudes also pre- evidence and the dearth of competing explanations, dict depressed mood following a stressor (Hankin, it can be safely asserted that the cognitive model is Abramson, Miller, & Haeffel, 2004). In one a valid characterization of the etiology of affective study, participants who were classified as being at disorders. Thus, one would expect considerable highcognitiveriskwerealmost7timesmorelikely support for the hypothesis that change in cognition to report a major depressive episode at 2.5 years mediates symptom change in the context of psy- follow-up, relative to those at low risk (Alloy et al., chotherapy. Instead, the literature contains ques- 2006). tions about whether “we need to challenge thoughts Although prospective studies provide a stronger in cognitive behavior therapy?” (Longmore & level of evidence for causality than correlational Worrell, 2007) and assertions such as “whatever studies, findings from these studies are still subject may be the basis of changes with CT, it does not to third variable confounds, making experimental seem to be the cognitions as originally proposed” designs preferable. Relatively few experiments ( Kazdin, 2007). Why is this so? manipulating cognitions and assessing the effects of the manipulation on mood have been conducted. It’s Complicated The results of these experiments, however, are Previously, we (Lorenzo-Luaces, German, & consistent with models of cognitive vulnerability DeRubeis, 2015) have argued that disagreement (see Mathews & MacLeod, 2005). For example, in among commentators (e.g., Kazdin, 2007; Longmore a series of experiments, Mathews and Mackintosh & Worrell, 2007) regarding the role of cognitive

Please cite this article as: Lorenzo Lorenzo-Luaces, et al., Cognitive-Behavioral Therapy: Nature and Relation to Non-Cognitive Behavioral Therapy, Behavior Therapy (2016), http://dx.doi.org/10.1016/j.beth.2016.02.012 4 lorenzo-luaces et al. change in promoting symptom change in psycho- meta-analytic review of components studies and therapy for depression emerges from different concluded that it is uncommon, in studies that have assumptions regarding the inferences that can be used dismantling designs, for one component of drawn from treatment studies. Below we review a treatment to outperform another. However, in some of these issues in psychotherapies for depres- studies that have used additive designs, adding one sion and anxiety. treatment component to another enhances positive therapeutic outcomes, particularly in the longer experimental designs: additive and term. dismantling studies A meta-analysis by Adams et al. (2015) addressed When two treatment packages are very different the contribution of CT to BT and did not find an (e.g., psychoanalysis vs. exposure and response- added benefit of CT to BT and CBT packages. prevention), it is easy and perhaps even appropriate However, this meta-analysis did not differentiate to interpret findings from studies comparing treat- between additive and dismantling designs. Al- ment packages to reflect the relative efficacy of though, on the face of it the component studies in specific therapy procedures (e.g., analyzing trans- the meta-analysis by Adams et al. seem like they ference vs. engaging in exposure). However, when provide very conclusive answers about the superi- assessing treatments, such as CBT or eye-movement ority or equipotency of CT and BT, component desensitization and reprocessing (EMDR), which studies, as they have been conducted and inter- combine multiple active and overlapping elements, preted, have been extremely problematic. Summar- in this case exposure and cognitive restructuring ily reviewing some of the limitations of component (Tolin, 2014), it becomes more difficult to extrap- studies, Bell et al. (2013) stated: olate conclusions about therapy procedures from … Null results [in component studies] do not directly outcome data. address the issue of specific versus common factors because In lieu of tightly controlled basic research, such there is no group that received only common treatment as the research on emotion regulation strategies components. [...] Component designs may also under- (Webb et al., 2012), researchers have used compo- estimate the contributions of the component. Rehm (2009) suggested that because much improvement typically occurs nent studies as a way of addressing questions about in the early stages of therapy, whichever component is the differential utility of treatment elements. These presented first will appear to be the most effective. Thus, the component studies are often referred to as if they dismantled component (which is never introduced) is likely represent a single class of study design, but there are to appear unnecessary. Component studies are also likely to at least two different types of study designs, additive be statistically underpowered (Kazdin & Whitley, 2003)to detect the relatively small effect sizes that are likely to occur and dismantling designs, that fall under this rubric. with these types of designs. […]. [A] two-group component As described by Bell, Marcus, and Goodlad (2013), study with a presumed effect size of .24 (half the treatment they address different kinds of questions. In additive vs. placebo effect size) would require over 250 patients component studies, in one condition a component in each condition to have a power of .80. Even Kazdin is added to and compared with an already-existing, and Whitley’s (2003) higher estimate of an effect size of .45 for additive design studies would require 78 patients simpler treatment. Butler, Cullington, Munby, in each condition. In contrast, the average sample size for Amies, and Gelder (1984) provide an early example the studies included in the present meta-analysis was 23 of such a study. They examined the value of adding participants in each condition, which would require a large anxiety management to exposure for social anxiety effect size of .84 to have a power of .80. by comparing the combined treatment to exposure only as well as to exposure plus a nonspecific filler. The component methodology evidenced a surge Their findings suggested that adding anxiety man- in popularity following a landmark study by agement to exposure improved treatment outcomes. Jacobson et al. (1996). Jacobson et al. conjectured In dismantling designs, at least one component of a that the full CBT for depression package could be multicomponent treatment package is removed from divided into three components: (1) behavioral the treatment and compared to the full treatment activation (BA); (2) challenging automatic thoughts package or to the other components. For example, (ATs); and (3) modifying core beliefs (CBs). To Foa, Steketee, Grayson, Turner, and Latimer (1984) compare the relative efficacy of these procedures, dismantled exposure and response-prevention (ERP) and perceiving limitations in prior work suggesting and compared its effects with the effects of exposure that CT for depression was superior to BT (Shaw only and response prevention only. Their findings 1977), Jacobson et al. randomized participants to suggested that ERP was superior to either of its single three conditions lasting a maximum of 20 sessions: components and that, for contamination fears, ex- (1) 100% BA; (2) a condition that could use all posure alone may be more effective than response- the elements of BA and could include AT work; and prevention alone. Bell et al. (2013) conducted a (3) a condition that could use all of the elements of

BA, work on ATs, with a required minimum of 8 treatments, BA was more effective than CT or sessions devoted to CB work. In this study, across medications for individuals who were severely various metrics, no statistically significant between- depressed. In the BA condition, 76% of more conditon differences in outcome were reported. severely depressed participants met criteria for re- These findings have generally been misinterepreted sponse or remission, compared to 48% of patients in as indicating that BA is the active component in CT and 49% in antidepressant medications. Among CBT for depression and that the interventions the less severely depressed patients, response rates provided in the cognitive components were inert, on the BDI were 56% in CT, 60% in BA, and 40% thus presenting a major challenge to Beck’s cogni- in ADM. Coffman et al. (2007) identified patients tive theory (Longmore & Worrell, 2007). from the Dimidjian et al. trial who exhibited a Data from assessments of therapists’ adherence pattern of “extreme non-response” (ENR) in CT speak to the construct validity of the experiment (approximately one-fourth of those assigned to CT) by Jacobson et al. (1996). Although, in terms of and noted that none of the patients assigned to BA the absolute frequency with which techniques were evidenced an ENR. At baseline, the CT ENR patients conducted, behavioral work was nearly as frequent were more severely depressed, evidenced more in the CB condition as they were in BA, the relative functional impairment, and reported more problems frequency of BA procedures was greater in BA than with their primary support group. They interpreted in AT and CB. In the follow-up analysis of the trial, these findings to suggest that, relative to CT and Jacobson and colleagues went on further to note: medications, BA may be particularly effective for ... by definition, participants in the BA condition received patients with severe depression that is accompanied more BA than did those in the other treatment conditions. by interpersonal dysfunction. However, it should be Although one might be tempted to infer from this study that noted that the advantage of BA over CT dissipated cognitive interventions are nonessential, our study does not entirely across the trial’s2-yearfollow-up(Dobson directly address the validity of such an interpretation. All we et al., 2008). Moreover, to our knowledge, the can conclude is that adding cognitive interventions to BA is findings of Dimidjian et al. (2006) have not been no more effective than using that time to add more BA. ’ (Gortner, Gollan, Dobson, & Jacobson, 1998, p. 381, replicated. An attempt to replicate Coffman et al. s emphasis added) ENR findings in a separate sample of depressed patients (Koenig, Jarrett, Gallop, Barrett, & Thase, In other words, if the findings from Jacobson 2014) treated with CT found a low (6%) rate of et al. (1996) are taken to mean that CT procedures nonresponse, and severity, functional impairment, are inert, a corollary emerges that was not tested in and interpersonal problems were not good predictors their design: that a BA condition that allowed only of nonresponse. Finally, Webb et al. (2013) found 6 or 7 sessions (one-third of the 20 sessions) should that the therapists in the Dimidjian et al. (2006) trial yield outcomes equivalent to a 20-session course implemented CT with a relatively more behavioral of BA. Thus, per Gortner et al. (1998), the only than cognitive focus, compared to therapists from hypotheses regarding behavioral treatment of de- other CT trials. pression that could have been tested with the study The effect of cognitive change strategies in Beck’s design were that BA is: (a) superior to, or (b) not C(B)T for depression has also been questioned, on inferior to, a cognitive-behavioral treatment. As the basis that much of symptom change occurs early there were no significant differences between the in treatment (Ilardi & Craighead, 1994). However, treatments, the most that can be said is that the BA it has been shown in several studies that CBT condition, in the context of a study with low power, therapists use cognitive change techniques as early was not shown to be inferior to treatments that as session one (Braun, Strunk, Sasso, & Cooper, divided time between cognitive components and 2015; Conklin & Strunk, 2015; Strunk, Brotman, behavioral ones. & DeRubeis, 2010). In fact, at least one therapy Very few dismantling studies have directly manual (i.e., Muñoz & Miranda, 1986) addresses compared “pure” cognitive and behavioral inter- cognitive change exclusively for several sessions ventions. The handful of studies that have com- before addressing behavior. pared purely behavioral (e.g., activity scheduling) In anxiety disorders, among the studies included to purely cognitive (e.g., cognitive restructuring) in the meta-analyses of Adams et al. (2015) and Bell treatments for depression have tended to find little et al. (2013), only eight studies (Barlow, Rapee, if any difference in the acute effects of cognitive Brown, 1992; Borkovec, Newman, Pincus, & Lytle, versus behavioral treatments (Mazzucchelli, Kane, 2002; Emmelkamp & Beens, 1991; Marks, Lovell, & Rees, 2009). In one study comparing BA to CT Noshirvani, Livanou, & Thrasher, 1998; Mattick, (Dimidjian et al., 2006), although there were no Peters, & Clarke, 1989, Szymanski, & O’Donohue, statistically significant differences between the two 1995; White, Keenan, & Brooks, 1992; Williams,

& Falbo, 1996) compared a purely cognitive (Fisher & Wells, 2005; McLean et al., 2001; but see treatment condition to a purely behavioral condi- Öst, Havnen, Hansen, & Kvale, 2015). Another tion. Among these eight studies, we conducted a exception is that, in social anxiety, the CT com- random effects meta-analysis (settings as per ponent appears to add to the efficacy of exposure IntHout, Ioannidis, & Borm, 2014). There was no (Hofmann, 2004; Mayo-Wilson et al., 2014; statistically significant difference in end-state pri- Ougrin, 2011; but see Chambless & Gillis, 1993). mary outcomes between “pure” CT and “pure” It has been suggested that, because with some BT (higher values indicate superiority for CT; g = anxiety disorders cognitive techniques may add little 0.010 [95% CI: -0.203 to 0.222], SE = 0.090, t = or nothing to the efficacy of BTs, exposure and other 0.106, p = 0.919; see Fig. 1). There was minimal BTs are best conducted without the questioning heterogeneity between effect sizes included in the of beliefs or the provision of other CT procedures meta-analysis (Cochrane’s Q = 3.371, df =7,p = (e.g., Arch & Craske, 2008). Indeed, a favored 0.849; 14.90% heterogeneity), consistent with an approach to the treatment of simple phobias has account that trial-level findings were generally been and continues to be one that relies primarily or equivocal, with few meaningful between-trial dif- only on BT techniques (Wolitzky-Taylor, Horowitz, ferences in effects. The results of these studies Powers, & Telch, 2008; but see Choy, Fyer, & suggest that CTs can be as effective as BTs in the Lipsitz, 2007). Some BTs, like the ones focused on treatment of anxiety disorders. Although exposur- relaxation, do not even directly address feared e-based treatments are considered the mainstay stimuli which may be taken to call into question of CBTs for anxiety, other meta-analytic reviews the need to engage in cognitive work. There are at also suggest that ERP, CBT, and CT are about least two things to keep in mind regarding these equally effective across a range of anxiety disorders comparisons. First, there is no evidence that these (Norton & Price, 2007; Ougrin, 2011). This led therapies lead to greater symptom reduction than Arch and Craske (2008) to propose that cognitive CT (e.g., Borkevec et al., 2002; Mayo-Wilson et al., restructuring is a form of exposure whose effects 2014). Second, the fact that these therapies, which are possibly cognitively mediated. One exception to do not directly address thinking, are effective does the pattern of equivalence in CT and BT is that, for not directly inform about their mechanisms. Recall OCD, ERP appears to be more effective than CT that Beck (1970) accounted for the efficacy of

FIGURE 1 Meta-analytic plot of the comparative efficacy of “pure” CT compared to “pure” BT in anxiety disorder RCTs identified by Adams et al. (2015) and Bell et al. (2013). Positive values indicate a superiority of CT over BT. Hedge’s g was calculated as a between-groups effect size of the end-scores of a trial primary outcome measure, selected by RJD and JRK from a results-blinded list of trial outcomes prepared by LLL. No notable changes in effect size or statistical significance resulted from controlling for pre-treatment severity differences by meta-regressing the Hedge’s g of the pre-treatment score differences between treatments (analyses available upon request). Egger’s test did not detect the presence of a significantly asymmetrical funnel plot potentially indicative of publication bias (p = 0.169), and Henmi and Copas’ (2010) test of publication bias proffered a similar between-groups effect estimate (g = 0.017).

Please cite this article as: Lorenzo Lorenzo-Luaces, et al., Cognitive-Behavioral Therapy: Nature and Relation to Non-Cognitive Behavioral Therapy, Behavior Therapy (2016), http://dx.doi.org/10.1016/j.beth.2016.02.012 cbt: nature and relation to non-c bt 7 behavioral therapies, giving the specific example of reverse causality. Inferences about causality rest on desensitization, in cognitive terms. the ascertainment of the correct temporal relation of the criterion and predictor variables. One must techniques vs. mechanisms have confidence that change on the predictor It appears to be a frequent misunderstanding of variable preceded the change in the criterion. This comparative psychotherapy research, including com- has sometimes been interpreted to mean that pre- ponent trials, that outcome studies inform about post changes in an outcome measure, regressed on the mechanisms of treatments. Various authors have an index of early change, establishes temporal cautioned against conflating the efficacy of thera- precedence. However, it frequently is the case that peutic procedures with their mechanisms of action substantial portions of pre to post changes in both (e.g., Hofmann, 2008; Lorenzo-Luaces et al., 2015), proposed mediator and the outcome occur early in and the separation between therapeutic procedures treatment, making early measurement of the medi- and mechanisms has long been recognized (e.g., Beck, ator variable a crucial step in establishing causality. 1970; Foa & Rauch, 2004). Jacobson et al. (1996) Changes in the criterion variable must be assessed explained it this way: subsequent to the assessment of change in the mediator if the aim is to rule out reverse causality. Of course, it is also possible that BA-focused treatments are more effective ways of changing the way people think than Only a small minority of tests of relations between treatments that explicitly attempt to alter thinking. Perhaps symptom changes and proposed mediators of those the exposure to naturally reinforcing contingencies produces changes has conformed to this pattern. changes in thinking more effectively than the explicitly cognitive interventions do. (p. 303) third variable confounds Although it may be tempting to assume that BT Even in studies with the temporal features that allow and CT produce symptom change via different reverse causality to be ruled out, third variable con- theorized mechanisms (see DeRubeis, Brotman, & founds can limit the inferences that can be made. Gibbons, 2005), they may also do so by the same Third variable confounds can never be entirely ruled mechanisms (Hofmann, 2008). It is possible that out because the multitude of variables that may both treatments work because they change think- be confounding the relation between a therapeutic ing. A pattern of findings that would be consistent procedure, a mechanism, and symptom change are with this hypothesis is that both therapies change unknown and potentially unknowable. With obser- cognition and that the cognitive changes lead to vational studies the best one can do is to test for the symptom change irrespective of how the cognitive most plausible confounds, using the best available changes are achieved. This latter point is one of measures of the potentially confounding construct. contention in the research literature because some Experimental designs can protect against many of have erroneously assumed that a cognitive theory the third variable confounds, but unobserved vari- of change implies that change in cognition leads to ables may still act as proxies for the purported symptom change uniquely in CBTs (see Hollon, mediating construct. Yet, as noted by Kazdin (2007), DeRubeis, & Evans, 1987). To the contrary, cogni- most of the experimental designs in psychotherapy tive theories highlight the mechanistic role of cog- are manipulations of the therapy, not of the proposed nition in psychopathology (Lorenzo-Luaces et al., mediator. 2015). In the study of change in psychotherapy, the Given that findings from comparative outcome domains most commonly hypothesized to account studies can, at best, provide food for thought about for symptom change have been: common factors, mechanisms of change, what is needed is more especially the therapeutic alliance; cognitive change; research on the psychological changes that account and behavioral changes. Ideally, a study that was for symptom change in psychotherapy. In the con- attempting to address questions about causality in text of component designs (e.g., Hofmann, 2004), psychotherapy would include measures or manipu- as well as in other kinds of randomized compari- lations of these phenomena. Given that the literature sons, investigations of the mediation of the effects is replete with studies that measure only one or two of psychotherapy promise to advance the under- of these variables, any conclusion that can be drawn standing of the workings of psychotherapy. about common factors, cognitive change, and behavioral change is necessarily tentative. temporality One of the greatest challenge to our understanding statistics and mediation of how BT and CT work is that most studies that Another conceptual hurdle in the understanding of explore these questions have been unable to rule out the relationship between therapeutic procedures,

Please cite this article as: Lorenzo Lorenzo-Luaces, et al., Cognitive-Behavioral Therapy: Nature and Relation to Non-Cognitive Behavioral Therapy, Behavior Therapy (2016), http://dx.doi.org/10.1016/j.beth.2016.02.012 8 lorenzo-luaces et al. cognitive change, and symptom change regards the producing symptom change, there are four guid- role of statistical analyses. Statistical tests cannot ing questions: determine causality. It does not matter how robust the test’s statistic is or whether the test is conducted 1. Are cognitive change procedures more effica- in the context of so-called “causal modeling.” The cious in reducing symptoms than other crucial matter is whether the structure of the data, procedures? which is dependent on the assessment schedule, 2. Do cognitive change procedures generate more meets the assumptions of causal models. In addi- cognitive change than other procedures? tion, plausible third variables confounds must be 3. Does cognitive change lead to symptom ruled out. Thus, a series of carefully planned single- change? case studies, in which mediators are experimentally 4. Is cognitive change a specific predictor of manipulated, and temporal order and third variable symptom change only in the context of cog- issues are addressed, may provide stronger evidence nitive change interventions? concerning the causal status of a mediator than would the use of more seemingly sophisticated Although some writers have assumed that the data analyses developed in a large sample (if there cognitive change model implies that the answer to is overlap in the time periods referenced in the all of these questions should be affirmative (e.g., mediator and outcome assessments). Longmore & Worrell, 2007), we have argued that Furthermore, an assumption in the writing of only the third question is a test of the cognitive several commentators is that mediational tests theory of change (Lorenzo-Luaces et al., 2015). In are the sine qua non for establishing causality (see agreement with Jacobson et al. (1996) it is even Hundt, Mignogna, Underhill, & Cully, 2013; possible that a noncognitive procedure will produce Kazdin, 2007). As it relates to cognitive change more cognitive change than an explicit cognitive procedures and cognitive change, in the mediation focus. The question is whether cognitive change, once framework proposed by Baron and Kenny (1986), it has been produced, leads to symptom change. evidence concerning cognitive change as a mediator of symptom change is given by answers to the Behavioral Change Strategies, Cognitive following questions: Change, and Symptom Change 1. Are cognitive change procedures more effica- Existing evidence largely supports cognitive chang- cious than (at least some) other procedures? es as mechanisms of change in cognitive, behavior- 2. Do cognitive change procedures generate al, and cognitive behavioral therapies. Most of this more cognitive change than those other research has been conducted in the context of social procedures? anxiety, panic disorder (Smits et al., 2012), and 3. Does superiority in cognitive change account depression (Lorenzo-Luaces et al., 2015). Below, for superior symptom change in the context we review evidence for the role of cognitive change of cognitive change procedures vs. those other in symptom change in CT, BT, and CBT. procedures? social anxiety This framework for conducting tests of media- Cognitive models of social anxiety highlight the tion is widely accepted, but can easily be mis- etiological role of cognitive biases in the perceived applied. Two change procedures can result in likelihood that negative social events will occur equivalent magnitudes of changes both on the (Smits, Rosenfield, McDonald, & Telch, 2006), as outcome variable and on a putative mediator well as in the overestimation of costs associated variable. For example, when cognitive and be- with these events (Clark & Wells, 1995). Moreover, havioral techniques produce equivalent outcomes, it has been suggested that individuals with social they may do so because they are equally effective anxiety tend to believe they are less socially desirable at changing cognitions (or behavior). A study may than they actually are (Moscovitch, 2009). find that cognitive and behavioral techniques are The proposed cognitive mediators of outcomes in equally effective in changing symptoms, cogni- social anxiety co-vary with symptom change in tion, and behavior. However, a traditional medi- CBTs (Hofmann, 2008; Wilson & Rapee, 2005). ational analysis will not be informative unless a For example, Boden et al. (2012) reported that conditionisincludedinthedesignandanalysis changes in maladaptive interpersonal beliefs fully that produces less change in the mediator(s) and accounted for changes in social anxiety over the the symptoms because there would be no effect to course of CBT for social anxiety. Hoffart, Borge, mediate. As regards the role of cognitive change in Sexton, and Clark (2009) explored four cognitive-

Please cite this article as: Lorenzo Lorenzo-Luaces, et al., Cognitive-Behavioral Therapy: Nature and Relation to Non-Cognitive Behavioral Therapy, Behavior Therapy (2016), http://dx.doi.org/10.1016/j.beth.2016.02.012 cbt: nature and relation to non-c bt 9 behavioral and four interpersonal processes of ed in the etiology and maintenance of the pathology change in CBT and interpersonal psychotherapy (Clark, 1986). Anxiety sensitivity, the attribution of (IPT) for social phobia. The lone IPT variable that negative somatic, cognitive, and social consequences predicted symptom change, perceived acceptance by to anxiety, has been specifically reported to render others, is arguably a cognitive construct. By con- individuals vulnerable to panic disorder (Reiss, trast, change in each of the four cognitive-behavioral 1991). Additionally, panic self-efficacy, the percep- variables—self-focus, estimated probability of nega- tion of one’s ability to cope with panic attacks, has tive social events, estimated cost of negative social also been implicated in the maintenance of the events, and safety behaviors—predicted changes in disorder (Casey, Oei, & Newcombe, 2004). social anxiety. Changes in positive, but not negative, In one study, fear of bodily sensations and anxiety self-view have been reported to co-vary with the sensitivity, jointly “fear of fear,” were reported reduction of social anxiety symptoms following to mediate the superiority of exposure relative to acute CBT and over a 1-year follow-up (Goldin a wait-list control (Smits, Powers, Cho, & Telch, et al., 2013). Changes in self-focused attention have 2004). Casey, Oei, Newcombe, and Kenardy (2004) also been reported to account for treatment out- reported that changes in catastrophic misinterpreta- comes in individual CT and in group-based CBT tions, as well as changes in self-efficacy, co-varied (Hedman et al., 2013), whereas change in negative with symptom change in CBT. Others, however, and positive self-statements predict outcomes in ACT have reported that changes in self-efficacy, but not in and CBT (Niles et al., 2014). catastrophic beliefs, correlate with symptom change Thus, a variety of cognitive constructs have shown (Fentz et al., 2013; Hoffart, 1995). associations with symptom improvement in CT, Attending to the temporal order of cognitive and CBT, IPT, and ACT. An issue with many of these symptom change, Teachman, Marker, and col- studies, however, is that the measurement of the leagues have reported that cognitive change pre- predictor and criterion variables is contemporane- cedes and predicts symptom change in CBT for ous. In one study that accounted for the temporal panic (Teachman, Marker, & Clerkin, 2010; order of change in a mediator and change in out- Teachman, Marker, & Smith-Janik, 2008). In one come, Goldin et al. (2014) reported that the success study, changes in automatic panic associations (but not the frequency of use) of cognitive reappraisal predicted changes in symptom severity (Teachman strategies predicted subsequent decreases in social et al., 2008). In another study, changes in cata- anxiety symptom. Decreases in social anxiety did strophic misinterpretations predicted subsequent not predict the successful use of cognitive reappraisal change in overall symptom severity, panic attack strategies. This study is encouraging in suggesting frequency, panic apprehension, and avoidance that cognitive process variables predict symptom behavior (Teachman et al., 2010). Gallagher et al. change in social anxiety. However, given the large (2013) also provided evidence for the temporal number of cognitive constructs that have been precedence of changes in anxiety sensitivity and also reported to co-vary with outcomes in social anxiety, found changes in self-efficacy to precede symptom it is likely that at least some of them are products change. These authors observed that overall change rather than predictors of symptom change. Illustrat- in anxiety sensitivity was greater than change in ing this point, Smits et al. (2006) found that self-efficacy, and that changes in self-efficacy oc- changes in probability biases for negative social curred later in treatment than changes in anxiety events predicted changes in social anxiety. However, sensitivity. Taken together, these studies support the after symptoms improved, there was a reduction in mediational roles of anxiety sensitivity, catastrophic patients’ estimates of the costs attributed to negative misinterpretations, and panic self-efficacy as cogni- social events. This pattern of results suggests that tive mediators of treatment effects in CBTs for panic whereas changes in probability biases were causally disorder (Sandin et al., 2015). related to symptom change, changes in cost estimates were the consequence of symptom change. These ptsd findings were replicated successfully and tested Current theories of posttraumatic disorder (PTSD) against the therapeutic alliance as a competing highlight the causal role of associations between predictor of change (Calamaras, Tully, Tone, Price, threatening (unconditioned fear stimuli) and non- & Anderson, 2015). In this study, the alliance was threatening (i.e., conditioned) stimuli in fear struc- not found to predict symptom change. tures of traumatic memories (Cahill & Foa, 2007). Additionally, trauma-related cognitions about the panic self, others, and the world—most commonly mea- In cognitive models of panic disorder, catastrophic sured with the Post-Traumatic Cognitions Inventory misinterpretations of interoceptive cues are implicat- (PCTI; Foa, Ehlers, Clark, Tolin, & Orsillo, 1999)—

Please cite this article as: Lorenzo Lorenzo-Luaces, et al., Cognitive-Behavioral Therapy: Nature and Relation to Non-Cognitive Behavioral Therapy, Behavior Therapy (2016), http://dx.doi.org/10.1016/j.beth.2016.02.012 10 lorenzo-luaces et al. have also been implicated in the etiology and main- data available to them, changes in cognition were tenance of the disorder. actually predicted by changes in symptoms rather In prolonged exposure (PE), changes in negative than vice versa. An almost identical pattern of cognitions about the self have been reported to results was reported by Olatunji et al. (2013): it was co-vary with changes in PTSD symptoms, whether symptom change that preceded and predicted or not the intervention includes cognitive restruc- changes in cognition and in avoidance behavior. turing (Foa & Rauch, 2004). In a sample of patients Interestingly, changes in depression preceded chang- with a severe mental illness comorbid with PTSD, es in OCD symptoms. In a recent trial, Wilhelm, changes in posttraumatic cognitions were found to Berman, Keshaviah, Schwartz, and Steketee (2015) mediate the superiority of CBT relative to treatment reported that, in CT for OCD, changes in perfec- as usual (Mueser et al., 2008). Research attending tionism and certainty beliefs predicted subsequent to the temporality of cognitive change and symp- change. These authors also reported that changes tom change suggests that these findings do not in maladaptive schemas related to dependency and reflect an epiphenomenal status for the cognitive incompetence predicted symptom improvement. changes. In patients who received trauma-focused Taken together, these studies do not provide CT for PTSD, weekly changes in trauma-related strong support for a causal role of any cognitive cognitions predicted subsequent reduction in symp- change variable in OCD treatment. An alternative toms (Kleim et al., 2013). Similarly, using data from mediator to cognitive change—such as behavioral a sample of patients who received PE, Zalta et al. change—has not successfully emerged as a predictor (2014) reported that session-to-session changes in of change. It is possible that a cognitive mediator trauma-related cognitions predicted subsequent accounts for change in OCD but it not well captured changes in PTSD symptoms, but not the other by existing measures. The findings that changes in way around. Thus, there is support for the cognitive depression precede changes in OCD, along with model of symptom change in treatments for PTSD, the findings that changes in thoughts of dependency although more research, with tests of additional and incompetence predict symptom change, may measures of trauma-relevant cognition, would help be taken to suggest that cognitions that have been advance our understanding of how treatments for more typically thought of as depressotypic may func- PTSD work. tion as one of the mechanisms of change in OCD treatments. ocd Cognitive theories of obsessive–compulsive disor- major depression der (OCD) highlight the role of various cognitive Cognitive models of major depression highlight variables. Overly attaching significance to one’s the role of negatively biased cognitive processes thoughts is central to Rachman’s influential cogni- (e.g., overgeneral memory style) and content (e.g., tive theory (Rachman, 1997). Intolerance of uncer- negative schemata, dysfunctional attitudes), as tainty, overestimation of threat, the belief that well as deficiencies in the ability to use cognitive thoughts should be controlled, inflated sense of reappraisal to modulate negative moods. This liter- responsibility, and perfectionism have also been ature has been reviewed elsewhere (see Lorenzo- implicated (Obsessive Compulsive Cognitions Luaces et al., 2015). Contemporaneous associations Working Group, 2003), with disagreement among between cognitive change and symptom change the authors as to which cognitions are key to the during treatments for depression have been observed etiology of OCD (Grayson, 2010; Gwilliam, Wells, in numerous investigations (e.g., Cristea et al., & Cartwright-Hatton, 2004). 2015; Hundt et al., 2013). In a recent meta-analysis, There are few studies that have explored medi- for example, Cristea et al. (2015) reported a very ators of change in psychotherapy for OCD, a strong correlation between symptom change and surprising state of affairs given the number of cognitive change in treatments for depression (r = purported cognitive mediators. Interestingly, many 0.77). Although these findings are consistent with the of the studies that have been conducted have ade- proposal that cognitive change produces symptom quately addressed the temporal relation between change, they are also consistent with the converse. the purported mediators and outcomes, with mixed Support for cognitive mediation models has been results. Woody, Whittal, and McLean (2011) obtained in studies that have attempted to address reported that mediational tests conducted in a reverse causality by modeling subsequent change pre-post fashion suggested that the superiority of in depression using prior cognitive change (Forman CBT relative to stress training were mediated by et al., 2012; Segal et al., 1999; Segal et al., 2006; changes in negative cognitions. However, when Strunk, DeRubeis, Chiu, & Alvarez, 2007; Tang & these authors utilized the full session-by-session DeRubeis, 1999; Tang, DeRubeis, Beberman, &

Pham, 2005). Although these studies provide general cognitive specificity support for cognitive theories of change in symp- There is much excitement regarding the possibility toms, no one variable has emerged as the clear that different treatments could work via different predictor of symptom change in therapy. For hypothesized mechanism. Statistical tests that reveal example, Strunk et al. (2007) found that patients’ the moderation of a predictive effect (e.g., tests of competence in behavioral strategies, automatic moderated mediation) may be taken as evidence for thoughts, and core schematic work predicted a this. Although the interpretation of mediators and lower risk of relapse. predictors within a treatment has been widely Given how heterogeneous depression is, it is likely discussed, moderated mediation and related findings that patients differentially benefit from different have received relatively less attention and are conse- interventions. Adding support for this hypothesis, quently less well-understood (Gelfand & DeRubeis, Sasso et al. (2015) found that more anxious patients 2015). Various findings of this nature, however, have and those with less recurrent depression benefitted been reported (e.g., Cottraux et al., 2001; Hedman more therapists’ use of behavioral strategies than et al., 2013; Niles et al., 2014). Here we focus on the cognitive ones. By contrast, Keefe, Webb, and interpretation of moderated relations between cog- DeRubeis (2016) found that patients with depres- nitive change and symptom change across treatments sion and personality disorders benefitted from for panic disorder. a focus on maladaptive core beliefs but did not Arntz (2002) reported equal amounts of change experience benefit from other techniques such as in catastrophic interpretations following CT and a focus on automatic thoughts and behavioral BT for panic. However, posttreatment beliefs cor- change. related with severity of symptoms at a follow-up in CT but not in BT. In a sample of patients receiving generalized anxiety and specific CBT, imipramine, or combined treatment, analy- phobias ses suggested that changes in panic-related cogni- There have been relatively few investigations of the tions, which were equal across conditions, co- mediation of change in treatments of generalized varied with changes in panic severity in CBT or anxiety disorder (GAD) and specific phobias. combined treatment but not imipramine mono- Current etiological models of GAD highlight the therapy (Hofmann et al., 2007). In a comparisons causal roles of intolerance of uncertainty, positive of capnometry-assisted respiratory training (CART) beliefs about worry, and the avoidance of emotion- versus training in CT strategies, Meuret, Rosenfield, al/internal experiences (including emotional con- Seidel, Bhaskara, and Hofmann (2010) found sup- trasts). Few studies have explored these constructs port for the therapy-specific mediators. In CART, as mediators or explanatory variables in CBTs reduced carbon dioxide in the bloodstream from for GAD, although they each have some support hyperventilation (hypocapnia), but not cognitive in basic research (see Newman, Llera, Erickson, reappraisal, predicted symptom change. In cognitive Przeworski, & Castonguay, 2013). Goldman et al. training, the opposite pattern was obtained; cog- (2007) reported that decreases in intolerance of nitive reappraisal, but not hypocapnia, predicted uncertainty preceded changes in worry over the symptom change. course of treatment. Similarly, Bomyea et al. (2015) Thus, in several studies of treatments for panic reported that changes in intolerance of uncertainty disorder, a “non-cognitive” intervention produced preceded and predicted change in GAD symptoms. similar amounts of cognitive and symptom change Cognitive models of specific phobias highlight relative to an explicitly cognitive intervention, yet the role of exaggerated judgments of dangerous- the relation of cognitive change to symptom change ness, disgust, unpredictability, and uncontrollabil- was only found in the cognitive intervention. It is ity associated with feared stimuli (see Armfield, difficult to reconcile these findings with cognitive 2006). Additionally, these models highlight the role theories of panic, or cognitive theories of psycho- of low expectations for the capacity to cope with pathology more broadly, because these models the consequences of coming into contact with the assume that changes in cognitive variables are feared stimuli, which may include the ability to mechanisms of symptom change across treatment tolerate uncomfortable physiological sensations. modalities (Lorenzo-Luaces et al., 2015). One In one study examining behavioral exposures vs. possibility is that the cognitive change variables behavioral experiments designed to test maladap- examined play a causal role only in interactions tive beliefs, both treatments showed substantial with another variable that is present only in the symptom and cognitive change, and cognitive change treatments in which cognitive change was predic- was associated with symptom levels at posttreatment tive of symptom change (see DeRubeis et al., 1990). and follow-up (Raes et al., 2011). For example, catastrophic misinterpretations of

Please cite this article as: Lorenzo Lorenzo-Luaces, et al., Cognitive-Behavioral Therapy: Nature and Relation to Non-Cognitive Behavioral Therapy, Behavior Therapy (2016), http://dx.doi.org/10.1016/j.beth.2016.02.012 12 lorenzo-luaces et al. panic, in conjunction with increased self-efficacy, difficult to disentangle, conceptually and practical- might mediate changes in both CT and respiratory ly, from symptom measures. training. A further possibility is that the treatments intervene at different stages of the patho- Recommendations for Future Research logical process. For example, one account of panic Although the existing literature allows one to draw disorder could be that hyperventilation leads to conclusions regarding the status of cognitive and hypocapnia, which produces dizziness and dis- behavioral techniques, as well as the mediational comfort, which is misinterpreted, resulting in role of cognitive change, more research is clearly more severe panic symptoms. Breathing retraining needed, particularly research employing experimen- might intervene “early” in this cycle, whereas tal methods. Dismantling studies and randomized cognitive interventions address the appraisal of comparative trials only infrequently are comple- bodily sensations have already been experienced. mented by mediational analyses that could help to In such a case, the cognitive changes observed in clarify what drives change in the interventions. It breathing retraining are only a product of reduced will be important to conduct more experiments symptoms. where interventions are compared and mediational hypotheses are tested. As Kazdin (2007) has noted, Distinguishing Cognition and Behavior however, it is important for research to test the Thus far in this review we have used “cognition,” effects of the manipulations of mechanisms, rather “behavior,” and their adjectival forms to reflect the than continuing a focus on the manipulation of way previous authors have used them in discussions techniques or packages of techniques. of therapeutic techniques or mechanisms. However, Researchers who conduct work on mediators of clear, consensually-agreed-upon definitions of these change in nonexperimental contexts should ensure terms do not exist. In some cases, therapeutic that third variables and competing mediators are procedures or mechanisms are easy to characterize measured. Given that behavioral change, cognitive as cognitive or behavioral. For example, searching change, and common factors are the most frequent- for evidence for or against a belief in the context ly cited mediators of symptom change, it will be of therapy is readily seen as a cognitive procedure, important to measure and model these variables just as engaging in in vivo exposure following a concurrently. Moreover, it might also be important habituation rationale is easily seen as behavioral. to measure multiple aspects of a given construct. However, cognitive techniques may be used to For example, in the behavioral domain, a reduction facilitate exposure (Arch & Craske, 2008), and in safety behaviors along with behavioral exposures cognitive therapists use behavioral techniques to are both purported mechanisms of symptom facilitate tests of patients’ beliefs (McManus, Van change. In the cognitive domain, the multiplicity Doorn, & Yiend, 2012). Thus, it behooves inves- of cognitive variables that have been implicated tigators to specify what counts as cognitive or should also be measured, particularly for disorders behavioral vs. what does not. Similarly, cognitive like depression and OCD, where multiple cognitive and behavioral mechanisms of therapy must be vulnerabilities are hypothesized to be at play. operationalized in a way that allows for falsifica- Additionally, the role of emotions and emotional tion. Arguing that all behavioral changes must, by arousal, whether by themselves or in interactions definition, entail cognitive changes, or that all with other variables, needs to explored. Although cognition is behavior, is unlikely to lead to more non-CBT treatments have traditionally focused more valid models of psychopathology. on emotional experiences than CBTs (Whelton, A related issue to the conceptualization of cog- 2004), there is evidence that clients’ in-session emo- nition and behavior is their measurement. Measur- tional involvement relates to outcomes, at least in ing cognitive and behavioral activity objectively in CT for depression (Castonguay, Goldfried, Wiser, the context of psychotherapy is a difficult enter- Raue, & Hayes, 1996). Hayes and colleagues prise. Although self-reported measures are the most (Grosse Holtforth et al., 2012; Hayes, Beevers, convenient way of measuring cognitive or behav- Feldman, Laurenceau, & Perlman, 2005; Hayes ioral change, they are subject to various biases that et al., 2007) have proposed, and found evidence for, have been well-articulated in the research literature. a model of change in depression treatment where It remains to be ascertained whether self-report behavioral interventions foster emotional arousal measures are more valid ways of ascertaining cog- that, when combined with cognitive processing, leads nitive change than behavioral change or vice-versa. to symptom change. These findings, which can be Moreover, as avoidance behaviors and maladaptive conceptualized using the concept of cognitive- cognitions are symptoms of psychopathology, affective schemata (Beck & Haigh, 2014), warrant assessments of both behaviors and cognitions are further exploration.

Mechanisms of Change: Cognitive and the literature on cognitive biases and cognitive bias Behavioral Treatments modification, Hallion and Ruscio (2011) asserted that: The suggestion that behavioral exposure is superior anxiety is associated with biases in the early, automatic to cognitive restructuring alone for OCD and that stages as well as the later, strategic stages of attention cognitive restructuring adds to behavioral expo- (Bar-Haim et al., 2007), whereas depression is associated sures for social anxiety may be dismissed, given that with [cognitive] biases only in the later stages. (e.g., Gotlib these are not universal observations. However, it is et al., 2004; Joormann, 2004) noteworthy that in a disorder in which cognitive Temporal precedence of cognitive change has procedures appear to be especially potent—social — been easier to establish in anxiety disorders than in anxiety strong evidence has been obtained for depressive disorders. It is possible that this reflects cognitive change as a mechanism of symptom the temporal precedence of cognitive biases in change (see McMillan & Lee, 2010). By contrast, anxiety relative to depression, thought this question in a disorder in which cognitive procedures appear merits further exploration. to be less potent (OCD), there is little if any evidence for a cognitive mechanism of symptom Conclusions change. (These observations, however, should not As has been noted extensively (DeRubeis et al., be taken to imply that efficacy findings allow one to 2005; Haaga, 2007; Kazdin, 2007), characterizing make claims about causal mechanisms.) Much is the process of change in psychotherapy is extremely made and little is done about the potential for challenging. Within a therapeutic orientation, a knowledge of mediator variables to inform treat- given therapeutic package will encourage a set of ment research, but the existing data on the cognitive therapeutic procedures, such as enhancing motiva- mediators in mood and anxiety disorders could be tion for behavioral change, some of which may used to inform treatment. For example, in the overlap with those of therapeutic packages, within context of treatment for social phobia, treatment or outside of that orientation. Factors common to could be targeted especially to the perception that different interventions may also drive symptom negative social events are likely to occur, as op- change, confounding the relation between specific posed to treating symptoms or functioning as the procedures and improvement. Furthermore, a given targets. procedure might not affect the psychological In broad terms, research supports cognitive vul- mechanism(s) it is intended to target, and might nerability models of depression and anxiety disor- instead affect a mechanism that is not its intended ders. Additionally, cognitive changes appear to target. correlate with and predict symptom change in As noted by Beck (1991), a psychotherapeutic therapy for depressive and anxiety disorders. Given approach is not a dissociated collection of tech- this state of the literature, it behooves researchers niques, but rather a set of procedures that follow to characterize the exact nature of the cognitive from “a comprehensive theory of psychopathology vulnerability to psychopathology, as well as the that articulates with the structure of psychothera- nature of the cognitive variables implicated in py” (p. 368). Thus, productive considerations of the symptom change. Often the differences between effectiveness or validity of a procedure or mecha- different cognitive vulnerability theories lie in the nism should be linked to the theory of the disorder stage of information processing that they focus on. or pathological process in question. Critiques of a For example, whereas some investigators have cognitive model of the development, maintenance, emphasized individual differences in the propensity and resolution of psychopathology will advance the to attend to negative stimuli (MacLeod et al., 2002), discussion insofar as they are accompanied by an others have focused on individual differences in alternative theory, along with proposals, and interpretation of life events (Abramson, Metalsky, evidence, regarding alternative noncognitive mech- & Alloy, 1989). The latest characterization of anisms of change. In our view, oft-cited critiques of the cognitive model provided by Beck attempts to the theoretical model that underlie CBTs (Kazdin, represent a more integrative picture of the cognitive 2007; Hayes, 2004; Longmore & Worrell, 2007) vulnerabilities to psychopathology (Beck & Haigh, have employed oversimplifications of the CBT 2014), highlighting the roles of attention, memory, model. The available data across the sciences interpretative biases, emotional processing, and supports cognitive models (Beck & Haigh, 2014; cognitive schemata in the development of psycho- Hofmann et al., 2013) in suggesting that some pathology. However, it may be that vulnerabilities individuals are more likely than others to be ex- in different stages of information processing lead posed to negative environmental stimuli, less likely to different types of psychopathology. Reviewing to be exposed to positive stimuli, or more likely to

Please cite this article as: Lorenzo Lorenzo-Luaces, et al., Cognitive-Behavioral Therapy: Nature and Relation to Non-Cognitive Behavioral Therapy, Behavior Therapy (2016), http://dx.doi.org/10.1016/j.beth.2016.02.012 14 lorenzo-luaces et al. attend preferentially to negative information. In the Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator contexts of stress, these individuals are more likely variable distinction in social psychological research: Con- ceptual, strategic, and statistical considerations. Journal of to experience distress and less likely to engage the Personality and Social Psychology, 51, 1173–1182. http:// kinds of reappraisal and coping strategies that dx.doi.org/10.1037/0022-3514.51.6.1173 would produce naturalistic recovery. Cognitive Beck, A. T. (1963). Thinking and depression: I. Idiosyncratic and behavioral therapies represent an attempt to content and cognitive distortions. Archives of General – activate these resources. The distinctions between Psychiatry, 9,324 333. http://dx.doi.org/10.1001/archpsyc. 1963.01720160014002 behavioral and cognitive change strategies and Beck, A. T. (1970). Cognitive therapy: Nature and relation to procedures are, to some degree, artificial inasmuch behavior therapy. Behavior Therapy, 1, 184–200. http://dx. as CT procedures are often used before and after doi.org/10.1016/S0005-7894(70)80030-2 behavioral exercises, and vice-versa. Moreover, the Beck, A. T. (1991). Cognitive therapy: A 30-year retrospective. – relative efficacy of these procedures, even when American Psychologist, 46, 368 375. http://dx.doi.org/10. 1037/0003-066X.46.4.368 considered in specific contexts, does not provide Beck,A.T.(1996).Beyond belief: A theory of modes, dispositive evidence concerning mechanisms. To personality, and psychopathology. In P. Salkovskis (Ed.), date, the most parsimonious account of change Frontiers of cognitive therapy (p. 1–25). New York: produced in psychotherapy is one that invokes the Guilford Press. importance of changes in cognitive systems or Beck, A. T., & Haigh, E. A. (2014). Advances in cognitive theory and therapy: The generic cognitive model. Annual contents. Although the research literature is limited Review of Clinical Psychology, 10,1–24. http://dx.doi.org/ in its ability to provide strong confirmations of the 10.1146/annurev-clinpsy-032813-153734 causal hypotheses embedded in cognitive theories Bell, E. C., Marcus, D. K., & Goodlad, J. K. (2013). Are of change, a compelling, evidence-based competing the parts as good as the whole? A meta-analysis of com- theory of change has yet to emerge. ponent treatment studies. Journal of Consulting and Clinical Psychology, 81, 722–736. http://dx.doi.org/10.1037/ Conflict of Interest Statement a0033004 Boden, M. T., John, O. P., Goldin, P. R., Werner, K., Heimberg, The authors declare that there are no conflicts of interest. R. G., & Gross, J. J. (2012). The role of maladaptive beliefs in cognitive-behavioral therapy: Evidence from social References anxiety disorder. Behaviour Research and Therapy, 50, 287–291. http://dx.doi.org/10.1016/j.brat.2012.02.007 Abramson, L. Y., Metalsky, G. I., & Alloy, L. B. (1989). Bomyea, J., Ramsawh, H., Ball, T. M., Taylor, C. T., Paulus, Hopelessness depression: A theory-based subtype of depres- M.P., Lang, A. J., & Stein, M. B. (2015). Intolerance of sion. Psychological Review, 96, 358–372. http://dx.doi.org/ uncertainty as a mediator of reductions in worry in a 10.1037/0033-295X.96.2.358 cognitive behavioral treatment program for generalized Adams, T. G., Brady, R. E., Lohr, J. M., & Jacobs, W. J. (2015). anxiety disorder. Journal of Anxiety Disorders, 33,90–94. A meta-analysis of CBT components for anxiety disorders. http://dx.doi.org/10.1016/j.janxdis.2015.05.004 the Behavior Therapist, 38,87–97. Borkovec, T. D., Newman, M. G., Pincus, A. L., & Lytle, R. Alloy, L. B., Abramson, L. Y., Whitehouse, W. G., Hogan, (2002). A component analysis of cognitive-behavioral M. E., Panzarella, C., & Rose, D. T. (2006). Prospective therapy for generalized anxiety disorder and the role of incidence of first onsets and recurrences of depression in interpersonal problems. Journal of Consulting and Clinical individuals at high and low cognitive risk for depression. Psychology, 70, 288–298. http://dx.doi.org/10.1037/0022- Journal of Abnormal Psychology, 115, 145–157. http:// 006X.70.2.288 dx.doi.org/10.1037/0021-843X.115.1.145 Braun, J. D., Strunk, D. R., Sasso, K. E., & Cooper, A. A. Arch, J. J., & Craske, M. G. (2008). Acceptance and (2015). Therapist use of Socratic questioning predicts commitment therapy and cognitive behavioral therapy for session-to-session symptom change in cognitive therapy for anxiety disorders: Different treatments, similar mechanisms? depression. Behaviour Research and Therapy, 70,32–37. Clinical Psychology: Science and Practice, 15, 263–279. http://dx.doi.org/10.1016/j.brat.2015.05.004 http://dx.doi.org/10.1111/j.1468-2850.2008.00137.x Butler, G., Cullington, A., Munby, M., Amies, P., & Gelder, M. Armfield, J. M. (2006). Cognitive vulnerability: A model of (1984). Exposure and anxiety management in the treatment the etiology of fear. Clinical Psychology Review, 26(6), of social phobia. Journal of Consulting and Clinical 746–768. http://dx.doi.org/10.1016/j.cpr.2006.03.007 Psychology, 52, 642–650. http://dx.doi.org/10.1037/0022- Arntz, A. (2002). Cognitive therapy versus interoceptive 006X.52.4.642 exposure as treatment of panic disorder without agorapho- Cahill, S. P., & Foa, E. B. (2007). Psychological theories of bia. Behaviour Research and Therapy, 40, 325–341. http:// PTSD. In M. J. Friedman, T. M. Keane, & P. A. Resick dx.doi.org/10.1016/S0005-7967(01)00014-6 (Eds.), Handbook of PTSD: Science and practice (p. 55–77). Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans- New York: Guilford Press. Kranenburg, M. J., & Van Ijzendoorn, M. H. (2007). Calamaras, M. R., Tully, E. C., Tone, E. B., Price, M., & Threat-related attentional bias in anxious and nonanxious Anderson, P. L. (2015). Evaluating changes in judgmental individuals: A meta-analytic study. Psychological Bulletin, biases as mechanisms of cognitive-behavioral therapy for 133,1–24. http://dx.doi.org/10.1037/0033-2909.133.1.1 social anxiety disorder. Behaviour Research and Therapy, Barlow, D. H., Rapee, R. M., & Brown, T. A. (1992). 71, 139–149. http://dx.doi.org/10.1016/j.brat.2015.06.006 Behavioral treatment of generalized anxiety disorder. Casey, L. M., Oei, T. P., & Newcombe, P. A. (2004). Behavior Therapy, 23,551–570. http://dx.doi.org/10. An integrated cognitive model of panic disorder: The role 1016/S0005-7894(05)80221-7 of positive and negative cognitions. Clinical Psychology

Review, 24, 529–555. http://dx.doi.org/10.1016/j.cpr.2004. py, and antidepressant medication in the acute treatment 01.005 of adults with major depression. Journal of Consulting and Casey, L. M., Oei, T. P., Newcombe, P. A., & Kenardy, J. Clinical Psychology, 74, 658. (2004). The role of catastrophic misinterpretation of bodily Dobson, K. S. (2009). Handbook of cognitive-behavioral therapies. sensations and panic self-efficacy in predicting panic New York: Guilford Press. severity. Journal of Anxiety Disorders, 18,325–340. Dobson, K. S., Hollon, S. D., Dimidjian, S., Schmaling, K. B., http://dx.doi.org/10.1016/S0887-6185(02)00257-8 Kohlenberg, R. J., Gallop, R. J., … Jacobson, N. S. (2008). Castonguay, L. G., Goldfried, M. R., Wiser, S., Raue, P. J., & Randomized trial of behavioral activation, cognitive thera- Hayes, A. M. (1996). Predicting the effect of cognitive py, and antidepressant medication in the prevention of therapy for depression: A study of unique and common relapse and recurrence in major depression. Journal of factors. Journal of Consulting and Clinical Psychology, 64, Consulting and Clinical Psychology, 76, 468–477. http://dx. 497–504. http://dx.doi.org/10.1037/0022-006X.64.3.497 doi.org/10.1037/0022-006X.76.3.468 Chambless, D. L., & Gillis, M. M. (1993). Cognitive therapy Ellis, A. (1962). Reason and emotion in psychotherapy. New of anxiety disorders. Journal of Consulting and Clinical York, NY: Lyle Stuart. Psychology, 61, 248–260. http://dx.doi.org/10.1037/0022- Emmelkamp, P. M., & Beens, H. (1991). Cognitive therapy 006X.61.2.248 with obsessive–compulsive disorder: A comparative evalu- Choy, Y., Fyer, A. J., & Lipsitz, J. D. (2007). Treatment of ation. Behaviour Research and Therapy, 29, 293–300. specific phobia in adults. Clinical Psychology Review, 27, http://dx.doi.org/10.1016/0005-7967(91)90120-R 266–286. http://dx.doi.org/10.1016/j.cpr.2006.10.002 Fentz, H. N., Hoffart, A., Jensen, M. B., Arendt, M., O’Toole, Clark, D. M. (1986). A cognitive approach to panic. Behaviour M. S., Rosenberg, N. K., & Hougaard, E. (2013). Research and Therapy, 24, 461–470. http://dx.doi.org/10. Mechanisms of change in cognitive behaviour therapy for 1016/0005-7967(86)90011-2 panic disorder: The role of panic self-efficacy and cata- Clark, D. M., & Wells, A. (1995). A cognitive model of social strophic misinterpretations. Behaviour Research and Ther- phobia. In R. G. Heimberg, M. R. Liebowitz, D. A. Hope, & apy, 51, 579–587. http://dx.doi.org/10.1016/j.brat.2013. Schneier (Eds.), Social phobia: Diagnosis, assessment, and 06.002 treatment (p. 69–93). New York, NY: Guilford Press. Firk, C., Siep, N., & Markus, C. R. (2013). Serotonin Coffman, S. J., Martell, C. R., Dimidjian, S., Gallop, R., & transporter genotype modulates cognitive reappraisal of Hollon, S. D. (2007). Extreme nonresponse in cognitive negative emotions: a functional magnetic resonance imaging therapy: Can behavioral activation succeed where cognitive study. Social Cognitive and Affective Neuroscience, 8, therapy fails? Journal of Consulting and Clinical Psychol- 247–258. ogy, 75, 531–541. http://dx.doi.org/10.1037/0022-006X. Fisher, P. L., & Wells, A. (2005). How effective are cognitive 75.4.531 and behavioral treatments for obsessive–compulsive disor- Conklin, L. R., & Strunk, D. R. (2015). A session-to-session der? A clinical significance analysis. Behaviour Research and examination of homework engagement in cognitive therapy Therapy, 43, 1543–1558. http://dx.doi.org/10.1093/scan/ for depression: Do patients experience immediate benefits? nsr091 Behaviour Research and Therapy, 72,56–62. http://dx.doi. Foa, E. B., Ehlers, A., Clark, D. M., Tolin, D. F., & Orsillo, org/10.1016/j.brat.2015.06.011 S.M. (1999). The posttraumatic cognitions inventory Cottraux, J., Yao, S. N., Lafont, S., Mollard, E., Bouvard, M., (PTCI): Development and validation. Psychological Assess- Sauteraud, A., … Dartigues, J. -F. (2001). A randomized ment, 11,303–314. http://dx.doi.org/10.1037/1040-3590. controlled trial of cognitive therapy versus intensive 11.3.303 behavior therapy in obsessive compulsive disorder. Psycho- Foa, E. B., & Rauch, S. A. (2004). Cognitive changes during therapy and Psychosomatics, 70, 288–297. http://dx.doi. prolonged exposure versus prolonged exposure plus cogni- org/10.1159/000056269 tive restructuring in female assault survivors with posttrau- Craske, M. G., Kircanski, K., Zelikowsky, M., Mystkowski, J., matic stress disorder. Journal of Consulting and Clinical Chowdhury, N., & Baker, A. (2008). Optimizing inhibitory Psychology, 72, 879–884. http://dx.doi.org/10.1037/0022- learning during exposure therapy. Behaviour Research and 006X.72.5.879 Therapy, 46(1), 5–27. http://dx.doi.org/10.1016/j.brat. Foa, E. B., Steketee, G., Grayson, J. B., Turner, R. M., & 2007.10.003 Latimer, P. R. (1984). Deliberate exposure and blocking Cristea, I. A., Huibers, M. J. H., David, D., Hollon, S. D., of obsessive–compulsive rituals: Immediate and long-term Andersson, G., & Cuijpers, P. (2015). The effects of effects. Behavior Therapy, 15(5), 450–472. http://dx.doi. cognitive behavior therapy for adult depression on dys- org/10.1016/S0005-7894(84)80049-0 functional thinking: A meta-analysis. Clinical Psychology Forman, E. M., Chapman, J. E., Herbert, J. D., Goetter, E. M., Review, 42,62–71. http://dx.doi.org/10.1016/j.cpr.2015. Yuen, E. K., & Moitra, E. (2012). Using session-by-session 08.003 measurement to compare mechanisms of action for accep- DeRubeis, R. J., Brotman, M. A., & Gibbons, C. J. (2005). tance and commitment therapy and cognitive therapy. A conceptual and methodological analysis of the nonspe- Behavior Therapy, 43,341–354. http://dx.doi.org/10.1016/j. cifics argument. Clinical Psychology: Science and Practice, beth.2011.07.004 12, 174–183. http://dx.doi.org/10.1093/clipsy.bpi022 Gallagher, M. W., Payne, L. A., White, K. S., Shear, K. M., DeRubeis, R. J., Evans, M. D., Hollon, S. D., Garvey, M. J., Woods, S. W., Gorman, J. M., & Barlow, D. H. (2013). Grove, W. M., & Tuason, V. B. (1990). How does cognitive Mechanisms of change in cognitive behavioral therapy for therapy work? Cognitive change and symptom change in panic disorder: the unique effects of self-efficacy and anxiety cognitive therapy and pharmacotherapy for depression. sensitivity. Behaviour Research and Therapy, 51, 767–777. Journal of Consulting and Clinical Psychology, 58, http://dx.doi.org/10.1016/j.brat.2013.09.001 862–869. http://dx.doi.org/10.1037/0022-006X.58.6.862 Gelfand, L. A., & DeRubeis, R. J. (2015). Models of specific Dimidjian, S., Hollon, S. D., Dobson, K. S., Schmaling, K. B., and nonspecific change mechanisms in mental health Kohlenberg, R. J., Addis, M. E., … Atkins, D. C. (2006). treatments. The Encyclopedia of Clinical Psychology,1–8. Randomized trial of behavioral activation, cognitive thera- http://dx.doi.org/10.1002/9781118625392.wbecp501

Goldin, P. R., Jazaieri, H., Ziv, M., Kraemer, H., Heimberg, therapy for depression. Journal of Consulting and Clinical R.G., & Gross, J. J. (2013). Changes in positive self-views Psychology, 75, 409–421. http://dx.doi.org/10.1037/0022- mediate the effect of cognitive-behavioral therapy for 006X.75.3.409 social anxiety disorder. Clinical Psychological Science, 1, Hedman, E., Mörtberg, E., Hesser, H., Clark, D. M., Lekander, 301–310. http://dx.doi.org/10.1177/2167702613476867 M., Andersson, E., & Ljótsson, B. (2013). Mediators in Goldin, P. R., Lee, I., Ziv, M., Jazaieri, H., Heimberg, R. G., & psychological treatment of social anxiety disorder: Individ- Gross, J. J. (2014). Trajectories of change in emotion ual cognitive therapy compared to cognitive behavioral regulation and social anxiety during cognitive-behavioral group therapy. Behaviour Research and Therapy, 51, therapy for social anxiety disorder. Behaviour Research and 696–705. http://dx.doi.org/10.1016/j.brat.2013.07.006 Therapy, 56,7–15. http://dx.doi.org/10.1016/j.brat.2014. Henmi, M., & Copas, J. B. (2010). Confidence intervals for 02.005 random effects meta‐analysis and robustness to publication Goldman, N., Dugas, M. J., Sexton, K. A., & Gervais, N. J. bias. Statistics in Medicine, 29, 2969–2983. http://dx.doi. (2007). The impact of written exposure on worry: A org/10.1002/sim.4029 preliminary investigation. Behavior Modification, 31, Hoffart, A. (1995). Cognitive mediators of situational fear in 512–538. http://dx.doi.org/10.1177/0145445506298651 agoraphobia. Journal of Behavior Therapy and Experimen- Gortner, E. T., Gollan, J. K., Dobson, K. S., & Jacobson, N. S. tal Psychiatry, 26,313–320. http://dx.doi.org/10.1016/ (1998). Cognitive–behavioral treatment for depression: 0005-7916(95)00043-7 Relapse prevention. Journal of Consulting and Clinical Hoffart, A., Borge, F. -M., Sexton, H., & Clark, D. M. (2009). Psychology, 66, 377–384. http://dx.doi.org/10.1037/0022- Change processes in residential cognitive and interpersonal 006X.66.2.377 psychotherapy for social phobia: A process-outcome study. Gotlib, I. H., Kasch, K. L., Traill, S., Joormann, J., Arnow, B.A., Behavior Therapy, 40,10–22. http://dx.doi.org/10.1016/j. & Johnson, S. L. (2004). Coherence and specificity of beth.2007.12.003 information processing biases in depression and social Hofmann, S. G. (2004). Cognitive mediation of treatment phobia. Journal of Abnormal Psychology, 113, 386–398. change in social phobia. Journal of Consulting and Clinical http://dx.doi.org/10.1037/0021-843X.113.3.386 Psychology, 72, 392–399. http://dx.doi.org/10.1037/0022- Grayson, J. B. (2010). OCD and intolerance of uncertainty: 006X.72.3.392 Treatment issues. Journal of Cognitive Psychotherapy, 24, Hofmann, S. G. (2008). Common misconceptions about 3–15. http://dx.doi.org/10.1891/0889-8391.24.1.3 cognitive mediation of treatment change: A commentary Grosse Holtforth, M., Hayes, A. M., Sutter, M., Wilm, K., to. Clinical Psychology Review, 28,67–70. http://dx.doi. Schmied, E., Laurenceau, J. P., & Caspar, F. (2012). org/10.1016/j.cpr.2007.03.003 Fostering cognitive-emotional processing in the treatment Hofmann, S. G., Asmundson, G. J., & Beck, A. T. (2013). The of depression: a preliminary investigation in exposure-based science of cognitive therapy. Behavior Therapy, 44, cognitive therapy. Psychotherapy and Psychosomatics, 81, 199–212. http://dx.doi.org/10.1016/j.beth.2009.01.007 259–260. http://dx.doi.org/10.1159/000336813 Hofmann, S. G., Meuret, A. E., Rosenfield, D., Suvak, M. K., Gwilliam, P., Wells, A., & Cartwright-Hatton, S. (2004). Dose Barlow, D. H., Gorman, J. M., … Woods, S. W. (2007). meta-cognition or responsibility predict obsessive–compul- Preliminary evidence for cognitive mediation during sive symptoms: A test of the metacognitive model. Clinical cognitive-behavioral therapy of panic disorder. Journal of Psychology & Psychotherapy, 11, 137–144. http://dx.doi. Consulting and Clinical Psychology, 75, 374–379. http://dx. org/10.1002/cpp.402 doi.org/10.1037/0022-006X.75.3.374 Haaga, D. A. (2007). Could we speed this up? Accelerating Hollon, S. D., DeRubeis, R. J., & Evans, M. D. (1987). Causal progress in research on mechanisms of change in cognitive mediation of change in treatment for depression: Discrim- therapy of depression. Clinical Psychology: Science and inating between nonspecificity and noncausality. Psycho- Practice, 14,240 –243. http://dx.doi.org/10.1111/j.1468- logical Bulletin, 102, 139–149. http://dx.doi.org/10.1037/ 2850.2007.00082.x 0033-2909.102.1.139 Hallion, L. S., & Ruscio, A. M. (2011). A meta-analysis of Hundt, N. E., Mignogna, J., Underhill, C., & Cully, J. A. the effect of cognitive bias modification on anxiety and (2013). The relationship between use of CBT skills and depression. Psychological Bulletin, 137, 940–958. http://dx. depression treatment outcome: A theoretical and methodo- doi.org/10.1037/a0024355 logical review of the literature. Behavior Therapy, 44, Hankin, B. L., Abramson, L. Y., Miller, N., & Haeffel, G. J. 12–26. http://dx.doi.org/10.1016/j.beth.2012.10.001 (2004). Cognitive vulnerability-stress theories of depression: Ilardi, S. S., & Craighead, W. E. (1994). The role of nonspecific Examining affective specificity in the prediction of depres- factors in cognitive-behavior therapy for depression. Clin- sion versus anxiety in three prospective studies. Cognitive ical Psychology: Science and Practice, 1, 138–155. http://dx. Therapy and Research, 28, 309–345. http://dx.doi.org/10. doi.org/10.1111/j.1468-2850.1994.tb00016.x 1007/s10802-008-9228-6 Ingram, R. E., Miranda, J., & Segal, Z. (2006). Cognitive Hayes, S. C. (2004). Acceptance and commitment therapy, vulnerability to depression. In L. B. Alloy & J.H. Riskind relational frame theory, and the third wave of behavioral (Eds.), Cognitive vulnerability to emotional disorders and cognitive therapies. Behavior Therapy, 35(4), 639–665. Mahwah, NJ: Lawrence Erlbaum. http://dx.doi.org/10.1016/S0005-7894(04)80013-3 IntHout, J., Ioannidis, J. P., & Borm, G. F. (2014). The Hayes, A. M., Beevers, C. G., Feldman, G. C., Laurenceau, J. P., Hartung-Knapp-Sidik-Jonkman method for random effects & Perlman, C. (2005). Avoidance and processing as meta-analysis is straightforward and considerably outper- predictors of symptom change and positive growth in an forms the standard DerSimonian-Laird method. BMC integrative therapy for depression. International Journal of Medical Research Methodology, 14, 25. http://dx.doi.org/ Behavioral Medicine, 12, 111–122. http://dx.doi.org/10. 10.1186/1471-2288-14-25 1037/0022-006X.75.3.409 Jacobson, N. S., Dobson, K. S., Truax, P. A., Addis, M. E., Hayes, A. M., Feldman, G. C., Beevers, C. G., Laurenceau, J. P., Koerner, K., Gollan, J. K., … Prince, S. E. (1996). A Cardaciotto, L., & Lewis-Smith, J. (2007). Discontinuities component analysis of cognitive-behavioral treatment for and cognitive changes in an exposure-based cognitive depression. Journal of Consulting and Clinical

Psychology, 64, 295–304. http://dx.doi.org/10.1037/ 1,167–195. http://dx.doi.org/10.1146/annurev.clinpsy.1. 0022-006X.64.2.295 102803.143916 Joormann, J. (2004). Attentional bias in dysphoria: The role of Mattick, R. P., Peters, L., & Clarke, J. C. (1989). Exposure inhibitory processes. Cognition & Emotion, 18, 125–147. and cognitive restructuring for social phobia: A controlled http://dx.doi.org/10.1080/02699930244000480 study. Behavior Therapy, 20,3–23. http://dx.doi.org/10. Kazdin, A. E. (2007). Mediators and mechanisms of change in 1016/S0005-7894(89)80115-7 psychotherapy research. Annual Review of Clinical Psychol- Mayo-Wilson, E., Dias, S., Mavranezouli, I., Kew, K., Clark, ogy, 3,1–27. http://dx.doi.org/10.1146/annurev.clinpsy.3. D.M., Ades, A. E., & Pilling, S. (2014). Psychological and 022806.091432 pharmacological interventions for social anxiety disorder Kazdin, A. E., & Whitley, M. K. (2003). Treatment of parental in adults: a systematic review and network meta-analysis. stress to enhance therapeutic change among children The Lancet Psychiatry, 1, 368–376. http://dx.doi.org/10. referred for aggressive and antisocial behavior. Journal of 1016/S2215-0366(14)70329-3 Consulting and Clinical Psychology, 71, 504–515. http://dx. Mazzucchelli, T., Kane, R., & Rees, C. (2009). Behavioral doi.org/10.1037/0022-006X.71.3.504 activation treatments for depression in adults: A meta- Keefe, J. R., Webb, C. A., & DeRubeis, R. J. (2016). In analysis and review. Clinical Psychology: Science and Practice, cognitive therapy for depression, early focus on maladaptive 16(4), 383–411. http://dx.doi.org/10.1111/j.1468-2850. beliefs may be especially efficacious for patients with 2009.01178.x personality disorders. Journal of Consulting and Clinical McLean, P. D., Whittal, M. L., Thordarson, D. S., Taylor, S., Psychology. Advance online publication. http://dx.doi.org/ Söchting, I., Koch, W. J., … Anderson, K. W. (2001). 10.1037/ccp0000071 Cognitive versus behavior therapy in the group treatment of Kircanski, K., & Gotlib, I. H. (2015). Processing of emotional obsessive–compulsive disorder. Journal of Consulting and information in major depressive disorder: Toward a Clinical Psychology, 69,205–214. http://dx.doi.org/10. dimensional understanding. Emotion Review, 7, 256–264. 1037/0022-006X.69.2.205 http://dx.doi.org/10.1177/1754073915575402 McManus, F., Van Doorn, K., & Yiend, J. (2012). Examining Kleim, B., Grey, N., Wild, J., Nussbeck, F. W., Stott, R., the effects of thought records and behavioral experiments in Hackmann, A., … Ehlers, A. (2013). Cognitive change instigating belief change. Journal of Behavior Therapy and predicts symptom reduction with cognitive therapy for Experimental Psychiatry, 43, 540–547. http://dx.doi.org/ posttraumatic stress disorder. Journal of Consulting and 10.1016/j.jbtep.2011.07.003 Clinical Psychology, 81,383–393. http://dx.doi.org/10. McMillan, D., & Lee, R. (2010). A systematic review of 1037/a0031290 behavioral experiments vs. exposure alone in the treat- Koenig, A. M., Jarrett, R. B., Gallop, R., Barrett, M. S., & ment of anxiety disorders: A case of exposure while Thase, M. E. (2014). Extreme nonresponse to acute phase wearing the emperor’snewclothes?Clinical Psychology cognitive therapy for depression: An attempt to replicate and Review, 30,467–478. http://dx.doi.org/10.1016/j.cpr. extend. Behavior Therapy, 45, 300–313. http://dx.doi.org/ 2010.01.003 10.1016/j.beth.2013.12.005 Mehu, M., & Scherer, K. R. (2015). The appraisal bias model of Lemogne, C., Gorwood, P., Boni, C., Pessiglione, M., Lehéricy, cognitive vulnerability to depression. Emotion Review, 7, S., & Fossati, P. (2011). Cognitive appraisal and life stress 272–279. http://dx.doi.org/10.1177/1754073915575406 moderate the effects of the 5-HTTLPR polymorphism on Meuret, A. E., Rosenfield, D., Seidel, A., Bhaskara, L., & amygdala reactivity. Human Brain Mapping, 32,1856–1867. Hofmann, S. G. (2010). Respiratory and cognitive media- http://dx.doi.org/10.1002/hbm.21150 tors of treatment change in panic disorder: Evidence for Longmore, R. J., & Worrell, M. (2007). Do we need to intervention specificity. Journal of Consulting and Clinical challenge thoughts in cognitive behavior therapy? Clinical Psychology, 78,691–704. http://dx.doi.org/10.1037/ Psychology Review, 27, 173–187. http://dx.doi.org/10. a0019552 1016/j.cpr.2006.08.001 Mezulis, A. H., Abramson, L. Y., Hyde, J. S., & Hankin, B. L. Lorenzo-Luaces, L., German, R. E., & DeRubeis, R. J. (2015). (2004). Is there a universal positivity bias in attributions? It’s complicated: The relation between cognitive change A meta-analytic review of individual, developmental, and procedures, cognitive change, and symptom change in cultural differences in the self-serving attributional bias. cognitive therapy for depression. Clinical Psychology Psychological Bulletin, 130, 711–747. http://dx.doi.org/10. Review, 41,3–15. http://dx.doi.org/10.1016/j.cpr.2014. 1037/0033-2909.130.5.711 12.003 Moscovitch, D. A. (2009). What is the core fear in social MacLeod, C., Rutherford, E., Campbell, L., Ebsworthy, G., & phobia? A new model to facilitate individualized case Holker, L. (2002). Selective attention and emotional conceptualization and treatment. Cognitive and Behavioral vulnerability: assessing the causal basis of their association Practice, 16, 123–134. http://dx.doi.org/10.1016/j.cbpra. through the experimental manipulation of attentional bias. 2008.04.002 Journal of Abnormal Psychology, 111(1), 107–123. http:// Mueser, K. T., Rosenberg, S. D., Xie, H., Jankowski, M. K., dx.doi.org/10.1037//0021-843X.111.1.107 Bolton, E. E., Lu, W., … Wolfe, R. (2008). A randomized Marks, I., Lovell, K., Noshirvani, H., Livanou, M., & Thrasher, controlled trial of cognitive-behavioral treatment for post- S. (1998). Treatment of posttraumatic stress disorder by traumatic stress disorder in severe mental illness. Journal of exposure and/or cognitive restructuring: A controlled study. Consulting and Clinical Psychology, 76, 259–271. http://dx. Archives of General Psychiatry, 55, 317–325. http://dx.doi. doi.org/10.1037/0022-006X.76.2.259 org/10.1001/archpsyc.55.4.317 Muñoz, R. F., & Miranda, J. (1986). Group therapy manual for Mathews, A., & Mackintosh, B. (2000). Induced emotional cognitive-behavioral treatment of depression. Santa Monica, interpretation bias and anxiety. Journal of Abnormal CA: Rand Corporation. Psychology, 109,602–615. http://dx.doi.org/10.1037/0021- Newman, M. G., Llera, S. J., Erickson, T. M., Przeworski, A., & 843X.109.4.602 Castonguay, L. G. (2013). Worry and generalized anxiety Mathews, A., & MacLeod, C. (2005). Cognitive vulnerability to disorder: A review and theoretical synthesis of evidence emotional disorders. Annual Review of Clinical Psychology, on nature, etiology, mechanisms, and treatment. Annual

Review of Clinical Psychology, 9, 275–297. http://dx.doi. Segal, Z. V., Gemar, M., & Williams, S. (1999). Differential org/10.1146/annurev-clinpsy-050212-185544 cognitive response to a mood challenge following successful Niles, A. N., Burklund, L. J., Arch, J. J., Lieberman, M. D., cognitive therapy or pharmacotherapy for unipolar depres- Saxbe, D., & Craske, M. G. (2014). Cognitive mediators of sion. Journal of Abnormal Psychology, 108,3–10. http://dx. treatment for social anxiety disorder: Comparing acceptance doi.org/10.1037/0021-843X.108.1.3 and commitment therapy and cognitive-behavioral therapy. Segal, Z. V., Kennedy, S., Gemar, M., Hood, K., Pedersen, R., Behavior Therapy, 45, 664–677. http://dx.doi.org/10.1016/ & Buis, T. (2006). Cognitive reactivity to sad mood pro- j.beth.2014.04.006 vocation and the prediction of depressive relapse. Archives Norton, P. J., & Price, E. C. (2007). A meta-analytic review of General Psychiatry, 63, 749–755. http://dx.doi.org/10. of adult cognitive-behavioral treatment outcome across the 1001/archpsyc.63.7.749 anxiety disorders. The Journal of Nervous and Mental Shaw, B. F. (1977). Comparison of cognitive therapy and Disease, 195, 521–531. http://dx.doi.org/10.1097/01.nmd. behavior therapy in the treatment of depression. Journal of 0000253843.70149.9a Consulting and Clinical Psychology, 45, 543–551. http://dx. Obsessive Compulsive Cognitions Working Group. (2003). doi.org/10.1037/0022-006X.45.4.543 Psychometric validation of the obsessive beliefs question- Smits, J. A., Julian, K., Rosenfield, D., & Powers, M. B. (2012). naire and the interpretation of intrusions inventory: Part I. Threat reappraisal as a mediator of symptom change in Behaviour Research and Therapy, 41, 863–878. http://dx. cognitive-behavioral treatment of anxiety disorders: A system- doi.org/10.1016/S0005-7967(02)00099-2 atic review. Journal of Consulting and Clinical Psychology, 80, O’Donohue, W., Ferguson, K. E., & Naugle, A. E. (2003). 624–635. http://dx.doi.org/10.1037/a0028957 The structure of the cognitive revolution: An examination Smits, J. A., Powers, M. B., Cho, Y., & Telch, M. J. (2004). from the philosophy of science. The Behavior Analyst, Mechanism of change in cognitive-behavioral treatment of 26(1), 85–110. panic disorder: evidence for the fear of fear mediational O’Donohue, W., & Noll, J. P. (1995). Is behaviorism false hypothesis. Journal of Consulting and Clinical Psychology, because there is no such thing as conditioning? Popper and 72, 646–652. http://dx.doi.org/10.1037/0022-006X.75.4. Skinner on learning. New Ideas in Psychology, 13,29–41. 523 http://dx.doi.org/10.1016/0732-118X(94)E0003-K Smits, J. A., Rosenfield, D., McDonald, R., & Telch, M. J. Olatunji, B. O., Rosenfield, D., Tart, C. D., Cottraux, J., (2006). Cognitive mechanisms of social anxiety reduction: Powers, M. B., & Smits, J. A. (2013). Behavioral versus an examination of specificity and temporality. Journal of cognitive treatment of obsessive–compulsive disorder: An Consulting and Clinical Psychology, 74, 1203–1212. http:// examination of outcome and mediators of change. Journal dx.doi.org/10.1037/0022-006X.74.6.1203 of Consulting and Clinical Psychology, 81, 415–428. http:// Strunk, D. R., Brotman, M. A., & DeRubeis, R. J. (2010). The dx.doi.org/10.1037/a0031865 process of change in cognitive therapy for depression: Öst, L. G., Havnen, A., Hansen, B., & Kvale, G. (2015). Predictors of early inter-session symptom gains. Behaviour Cognitive behavioral treatments of obsessive–compulsive Research and Therapy, 48, 599–606. http://dx.doi.org/10. disorder. A systematic review and meta-analysis of studies 1016/j.brat.2010.03.011 published 1993–2014. Clinical Psychology Review, 40, Strunk, D. R., DeRubeis, R. J., Chiu, A. W., & Alvarez, J. 156–169. http://dx.doi.org/10.1016/j.cpr.2015.06.003 (2007). Patients’ competence in and performance of Ougrin, D. (2011). Efficacy of exposure versus cognitive cognitive therapy skills: relation to the reduction of relapse therapy in anxiety disorders: systematic review and meta- risk following treatment for depression. Journal of Consult- analysis. BMC Psychiatry, 11, 200–212. http://dx.doi.org/ ing and Clinical Psychology, 75, 523–530. http://dx.doi.org/ 10.1186/1471-244X-11-200 10.1037/0022-006X.75.4.523 Rachman, S. (1997). A cognitive theory of obsessions. Szymanski, J., & O’Donohue, W. T. (1995). The potential Behaviour Research and Therapy, 35, 793–802. of state-dependent learning in cognitive therapy with Raes, A. K., Koster, E. H., Loeys, T., & De Raedt, R. (2011). spider phobics. Journal of Rational-Emotive and Cognitive- Pathways to change in one-session exposure with and Behavioral Therapy, 13,131–150. http://dx.doi.org/10.1007/ without cognitive intervention: An exploratory study in BF02354458 spider phobia. Journal of Anxiety Disorders, 25, 964–971. Tang, T. Z., & DeRubeis, R. J. (1999). Sudden gains and http://dx.doi.org/10.1016/j.janxdis.2011.06.003 critical sessions in cognitive-behavioral therapy for depres- Rehm, L. P. (2009). Looking at the components of treatment for sion. Journal of Consulting and Clinical Psychology, 67, depression. Clinical Psychology: Science and Practice, 16, 894–904. http://dx.doi.org/10.1037/0022-006X.67.6.894 412–415. http://dx.doi.org/10.1111/j.1468-2850.2009. Tang, T. Z., DeRubeis, R. J., Beberman, R., & Pham, T. (2005). 01179.x Cognitive changes, critical sessions, and sudden gains in Reiss, S. (1991). Expectancy model of fear, anxiety, and panic. cognitive-behavioral therapy for depression. Journal of Clinical Psychology Review, 11, 141–153. http://dx.doi.org/ Consulting and Clinical Psychology, 73, 168–172. http:// 10.1016/0272-7358(91)90092-9 dx.doi.org/10.1037/0022-006X.73.1.168 Sandin, B., Sánchez-Arribas, C., Chorot, P., & Valiente, R. M. Teachman, B. A., Marker, C. D., & Clerkin, E. M. (2010). (2015). Anxiety sensitivity, catastrophic misinterpretations Catastrophic misinterpretations as a predictor of symptom and panic self-efficacy in the prediction of panic disorder change during treatment for panic disorder. Journal of severity: Towards a tripartite cognitive model of panic Consulting and Clinical Psychology, 78, 964–973. http://dx. disorder. Behaviour Research and Therapy, 67,30–40. doi.org/10.1037/a0021067 http://dx.doi.org/10.1016/j.brat.2015.01.005 Teachman, B. A., Marker, C. D., & Smith-Janik, S. B. (2008). Sasso, K. E., Strunk, D. R., Braun, J. D., DeRubeis, R. J., & Automatic associations and panic disorder: trajectories of Brotman, M. A. (2015). Identifying moderators of the change over the course of treatment. Journal of Consulting adherence-outcome relation in cognitive therapy for and Clinical Psychology, 76, 988–1002. http://dx.doi.org/ depression. Journal of Consulting and Clinical 10.1037/a0013113 Psychology, 83, 976–984. http://dx.doi.org/10.1037/ Tolin, D. F. (2014). Beating a dead dodo bird: Looking at ccp0000045 signal vs. noise in cognitive-behavioral therapy for anxiety

disorders. Clinical Psychology: Science and Practice, 21, therapy for obsessive compulsive disorder: Role of maladap- 351–362. http://dx.doi.org/10.1111/cpsp.12080 tive beliefs and schemas. Behaviour Research and Therapy, Webb, C. A., DeRubeis, R. J., Hollon, S. D., Dimidjian, S., 65,5–10. http://dx.doi.org/10.1016/j.brat.2014.12.006 Amsterdam, J. D., & Shelton, R. C. (2013). Convergence Williams, S. L., & Falbo, J. (1996). Cognitive and performance- and divergence in the delivery of cognitive therapy in based treatments for panic attacks in people with varying two randomized clinical trials. Behaviour Research and degrees of agoraphobic disability. Behaviour Research and Therapy, 51,493–498. http://dx.doi.org/10.1016/j.brat. Therapy, 34, 253–264. http://dx.doi.org/10.1016/0005- 2013.05.003 7967(95)00063-1 Webb, T. L., Miles, E., & Sheeran, P. (2012). Dealing with Wilson, J. K., & Rapee, R. M. (2005). The interpretation of feeling: A meta-analysis of the effectiveness of strategies negative social events in social phobia: Changes during derived from the process model of emotion regulation. treatment and relationship to outcome. Behaviour Research Psychological Bulletin, 138, 775–808. http://dx.doi.org/10. and Therapy, 43,373–389. http://dx.doi.org/10.1016/j. 1037/a0027600 brat.2004.02.006 Wenze, S. J., Gunthert, K. C., & Forand, N. R. (2007). Wolitzky-Taylor, K. B., Horowitz, J. D., Powers, M. B., & Influence of dysphoria on positive and negative cognitive Telch, M. J. (2008). Psychological approaches in the reactivity to daily mood fluctuations. Behaviour Research treatment of specific phobias: A meta-analysis. Clinical and Therapy, 45,915–927. http://dx.doi.org/10.1016/j. Psychology Review, 28, 1021–1037. http://dx.doi.org/10. brat.2006.08.010 1016/j.cpr.2008.02.007 Wenze, S. J., Gunthert, K. C., & Forand, N. R. (2010). Woody, S. R., Whittal, M. L., & McLean, P. D. (2011). Cognitive reactivity in everyday life as a prospective Mechanisms of symptom reduction in treatment for predictor of depressive symptoms. Cognitive Therapy and obsessions. Journal of Consulting and Clinical Psychology, Research, 34, 554–562. http://dx.doi.org/10.1007/s10608- 79, 653–664. http://dx.doi.org/10.1037/a0024827 010-9299-x Zalta, A. K., Gillihan, S. J., Fisher, A. J., Mintz, J., McLean, C. P., Whelton, W. J. (2004). Emotional processes in psychotherapy: Yehuda, R., & Foa, E. B. (2014). Change in negative cognitions Evidence across therapeutic modalities. Clinical Psychology associated with PTSD predicts symptom reduction in prolonged & Psychotherapy, 11(1), 58–71. http://dx.doi.org/10.1002/ exposure. Journal of Consulting and Clinical Psychology, 82, cpp.392 171–175. http://dx.doi.org/10.1037/a0034735 White, J., Keenan, M., & Brooks, N. (1992). Stress control: A controlled comparative investigation of large group therapy for generalized anxiety disorder. Behavioural Psychotherapy, RECEIVED: August 9, 2015 20,97–113. http://dx.doi.org/10.1017/S014134730001689X ACCEPTED: February 25, 2016 Wilhelm, S., Berman, N. C., Keshaviah, A., Schwartz, R. A., & AVAILABLE ONLINE: xxxx Steketee, G. (2015). Mechanisms of change in cognitive