A cognitive model of pathological worry

“We use the term attentional control in this paper to refer to processes underlying the ability to intentionally ignore distracting information (e.g. external noises when trying to work), or shift attention from one topic to another (Miyake et al., 2000).”

“Furthermore, it may be that some reported beliefs about worry are post-hoc attempts at explaining or understanding the experience of worry, rather than being causes of worry (see Nisbett & Wilson, 1977; for a discussion of how post-hoc rationalizations can arise).”

“The latter effect has been revealed in studies in which tasks requiring attentional control capacity for their performance are particularly susceptible to interference from worry (Eysenck et al., 2007; Hayes et al., 2008; Leigh & Hirsch, 2011).”

“Pathological worry has much in common with what has been described as rumination, more often studied in the context of depression (Nolen-Hoeksema, 1991).”

“Faster responding to targets replacing threatening cues, indicating selective attention toward threat, occurs in individuals with pathological worry (Bradley, Mogg, White, Groom, & DeBono, 1999; MacLeod, Mathews, & Tata, 1986; Mogg, Mathews, & Eysenck, 1992).”

“High levels of anxiety and worry (e.g. in GAD) are characterized by selective attention to threatening cues matching emotional concerns (Mathews & MacLeod, 2005), or internally generated representations, such as bodily sensations, mental images or worrisome thoughts (Hayes, Hirsch, & Mathews, 2010;…”

“For example, highly anxious individuals reporting good attentional control disengage more rapidly from threat cues than equally anxious individuals with poor control (Derryberry & Reed, 2002; Lonigan & Vasey, 2009; Peers & Lawrence, 2009).”

“Involuntary “ bottom-up ” in fl uences Earlier models of emotional processing biases have assumed that attention to threat is promoted directly by anxiety or by hypothetical “ threat evaluation systems ” (e.g. Mathews & Mackintosh, 1998; Mogg & Bradley, 1998).”

“Crown Copyright (cid:1) 2012 Published by Elsevier Ltd. Worry is a primary cognitive characteristic of anxiety, and has been described as ‘ a chain of thoughts and images, negatively affect-laden and relatively uncontrollable ’ (Borkovec, Robinson, Pruzinsky, & DePree,1983, p.10).”

“Hence the more threatening the thought content, or the more catastrophic the interpretation, the more likely it is to capture attention, promoting escalation of negative thoughts into a full-blown and protracted worry episode.”

“The present model borrows from several other theoretical views and incorporates the contribution of inappropriate beliefs about worry (Wells, 2006), intolerance of uncertainty (Dugas et al.,1998), and the role of verbal processing in worry (Borkovec & Inz, 1990).”

“Furthermore, the outcome of this competitive process is ‘ biased ’ by top-down (e.g. current goals), as well as bottom-up in fl uences (e.g. salience, arousal; for a review see Mather & Sutherland, 2011).”

“Typically, pathological worry (e.g. in GAD) is associated with longer latencies to name the colours of threatening compared to healthy controls (e.g., Becker, Rinck, Margraf, & Roth, 2001; Mathews & MacLeod, 1985; Mogg, Mathews, & Weinman, 1989).”

“It might seem that there is a con fl ict between these fi ndings and the evidence that images typically provoke more emotion than does verbal processing of the same events (e.g. Holmes & Mathews, 2010).”

“Worry in predominantly verbal form then develops, in fl uenced by conscious processes such as attempts to resolve the perceived threat and the redirection of attentional control resources to worry content, as well as the continuing in fl uence of habitual processing biases.”

“In contrast, a recent meta-analysis (Hallion & Ruscio, 2011) led to the conclusion that the overall effect of cognitive bias modi fi cation across different populations is relatively weak, although it might be useful as an adjunct treatment.”

“In a neuro-imaging study (Bishop, 2009), participants had to respond to a target letter while ignoring other distracting letters under conditions when targets were either easy or dif fi cult to fi nd. High trait-anxious individuals were slowed more by the presence of competing distracters (although…”

“Several research groups (e.g., Davey, 2006; Ruscio & Borkovec, 2004; Wells, 1995) have noted that excessive worriers sometimes endorse certain advantages for worry (e.g. that worry helps them to solve problems, or allows dangers to be avoided), despite also endorsing negative beliefs (e.g., that…”

“…characterized by selective attention to threatening cues matching emotional concerns (Mathews & MacLeod, 2005), or internally generated representations, such as bodily sensations, mental images or worrisome thoughts (Hayes, Hirsch, & Mathews, 2010; Krebs, Hirsch, & Mathews, 2010; Mathews, 1990).”

“Involuntary “ bottom-up ” in fl uences Earlier models of emotional processing biases have assumed that attention to threat is promoted directly by anxiety or by hypothetical “ threat evaluation systems ” (e.g. Mathews & Mackintosh, 1998; Mogg & Bradley, 1998).”

“2).”

“…instructions are later reversed, responses are slowed to the previous “ no-go ” word cues in comparison to baseline, revealing that memories of previously practiced responses (or of withholding responses) were unintentionally retrieved and interfered with current goals (Verbruggen & Logan, 2008).”

“…modify biases of attention (Dandeneau, Baldwin, Baccus, Sakellaropoulo, & Pruessner, 2007; MacLeod, Rutherford, Campbell, Ebsworthy, & Holker, 2002; Mathews & MacLeod, 2002) and interpretation (Grey & Mathews, 2000, 2011; Mathews & Mackintosh, 2000; Murphy, Hirsch, Mathews, Smith, & Clark, 2007;…”

“A fi nal therapeutic direction is suggested by the dominance of verbal linguistic processing and lack of imagery in worry (Borkovec & Inz, 1990; Hirsch, Hayes, Mathews, Perman, & Borkovec, 2012; Hoyer et al., 2001).”

“In a modi fi ed version of the avoidance model, Newman and Llera (2011) present evidence that, rather than avoiding imagery, worry is maintained by the perception that it serves to avoid further increases in distress should a feared event actually occur.”

“However, the few studies that targeted pathological worry with multiple sessions of CBM (e.g. Amir et al., 2009) had at least moderate effect sizes, and it remains to be seen whether improved forms of CBM have even greater effects.”

“Ladouceur, Gosselin, and Dugas (2000) attempted a more direct test of the intolerance of uncertainty hypothesis by manipulating uncertainty in a gambling task and showed that this was associated with increased worry.”

“2).”

“Research using the Posner task (e.g. Fox, Russo, Bowles, & Dutton, 2001; Yiend & Mathews, 2001), in which a single threatening or benign cue is followed by a target either in the same location as the preceding cue, or to a different location has suggested that anxiety is characterized mainly by…”

“However others have reported evidence for selective engagement (Koster, Crombez, Verschuere, van Damme, & Wiersema, 2006), or have pointed out that what appears to be slowed disengagement may be confounded with general slowing due to emotional interference effects (Mogg, Holmes, Garner, & Bradley,…”

“In experimental studies, unresolved problems are particularly likely to be remembered (the Zeigarnik effect; Zeigarnik, 1938).”

“Such capture effects may be compounded by reduced motivation to exert such control due to maladaptive beliefs about worry (Ruscio & Borkovec, 2004; Wells, 1995; Wells & Carter, 2001).”

“However, the content typically labelled as worry concerns thoughts of possible future threats impinging on the individual, whereas the term rumination is more often applied to thoughts of past negative events or negative personal attributes (Watkins, Moulds, & Mackintosh, 2005).”

“…target that was always in the prior location of the benign word (cf. MacLeod et al., 2002); and the second involved listening to benign descriptions presented to one ear whilst ignoring worry-related content presented at the same time to the other (dichotic listening, cf. Mathews & MacLeod, 1986).”

“In keeping with the combined cognitive biases hypothesis (Hirsch, Clark, & Mathews, 2006), we suppose that emotional processing biases and impaired attentional control combine to maintain worry.”

“In the last decade a number of methods have been developed to experimentally modify biases of attention (Dandeneau, Baldwin, Baccus, Sakellaropoulo, & Pruessner, 2007; MacLeod, Rutherford, Campbell, Ebsworthy, & Holker, 2002; Mathews & MacLeod, 2002) and interpretation (Grey & Mathews, 2000, 2011;…”

“2).”

“Typically, pathological worry (e.g. in GAD) is associated with longer latencies to name the colours of threatening compared to healthy controls (e.g., Becker, Rinck, Margraf, & Roth, 2001; Mathews & MacLeod, 1985; Mogg, Mathews, & Weinman, 1989).”

“However, attention can be consciously directed as well (and emotional interpretations can be consciously endorsed or rejected; Ouimet, Gawronski, & Dozois, 2009): in the present paper we therefore consider both automated and controlled in fl uences on worry.”

“For example, highly anxious individuals reporting good attentional control disengage more rapidly from threat cues than equally anxious individuals with poor control (Derryberry & Reed, 2002; Lonigan & Vasey, 2009; Peers & Lawrence, 2009).”

“Mathews, Richards, and Eysenck (1989) showed that individuals with GAD wrote down more threat spellings of ambiguously threatening homophones (words that sound the same but are spelled differently e for example die/dye).”

“Similarly, in the context of an ambiguous description of a job interview, socially anxious individuals were relatively faster to identify words matching threatening rather than benign inferences about the outcome (e.g. ‘ fail ’ rather than ‘ succeed ’ ; Hirsch & Mathews, 1997; 2000).”

“…others have reported evidence for selective engagement (Koster, Crombez, Verschuere, van Damme, & Wiersema, 2006), or have pointed out that what appears to be slowed disengagement may be confounded with general slowing due to emotional interference effects (Mogg, Holmes, Garner, & Bradley, 2008).”

“Similarly, Watkins, Baeyens, and Read (2009) found that training dysphoric individuals to ruminate in an abstract manner (that is, in similar form to verbal worry) resulted in a more negative mood than did equivalent practice in concrete thinking (which is more likely to be accompanied by imagery).”

“Reported proneness to worry varies continuously across the normal population, without any sudden discontinuity (Ruscio, Borkovec, & Ruscio, 2001).”

“2).”

“…slowing was associated with less activation in dorso-lateral pre-frontal cortex, suggesting that anxious individuals recruit less effortful attentional control resources to inhibit competing distracters than non-anxious people, particularly when task demands are low (see also Fales et al., 2008).”

“Several research groups (e.g., Davey, 2006; Ruscio & Borkovec, 2004; Wells, 1995) have noted that excessive worriers sometimes endorse certain advantages for worry (e.g. that worry helps them to solve problems, or allows dangers to be avoided), despite also endorsing negative beliefs (e.g., that…”

“Typically, pathological worry (e.g. in GAD) is associated with longer latencies to name the colours of threatening compared to healthy controls (e.g., Becker, Rinck, Margraf, & Roth, 2001; Mathews & MacLeod, 1985; Mogg, Mathews, & Weinman, 1989).”

“Both faster engagement and slowed attentional disengagement have been found in experiments using fearful gaze as a directional cue (Fox, Mathews, Calder, & Yiend, 2007; Mathews, Fox, Yiend, & Calder, 2003).”

“In a direct study of the latter possibility (Hayes et al., 2008) groups reporting high or low levels of worry were asked to worry, or think about a positive topic, while performing a random key-press task.”

“…Baccus, Sakellaropoulo, & Pruessner, 2007; MacLeod, Rutherford, Campbell, Ebsworthy, & Holker, 2002; Mathews & MacLeod, 2002) and interpretation (Grey & Mathews, 2000, 2011; Mathews & Mackintosh, 2000; Murphy, Hirsch, Mathews, Smith, & Clark, 2007; Wilson, MacLeod, Mathews, & Rutherford, 2006).”

“…to increase general control resources, current evidence suggests that it may be more effective to train the speci fi c control operations required such as ignoring threatening cues, in order to achieve the goal of avoiding capture of attention by worry (Schweizer, Hampshire, & Dalgliesh, 2011).”

“…word), anxious individuals were relatively faster than non-anxious controls to endorse ‘ cancer ’ (versus ‘ plant ’ ) as a word when it followed the priming word ‘ growth ’ , indicating that the ambiguous prime had been interpreted in terms of its more threatening meaning (Richards & French,1992).”

“Genetic variations, such as the low expression allele of the serotonin transporter gene, are associated with emotional processing biases, which in turn mediate increased emotional vulnerability to negative life events (Beevers, Wells, Ellis, & McGeary, 2009).”

“Verbal and image representations differ in several respects: emotional images are usually concrete representations of speci fi c outcomes, whereas verbalworryismoreabstract,withthefearedoutcome(s)lessexactly speci fi ed (Stöber & Borkovec, 2002).”

“As reviewed in the previous section, modifying negative interpretation bias reduces negative intrusive thoughts (Hirsch et al., 2009).”

“Similarly, in the context of an ambiguous description of a job interview, socially anxious individuals were relatively faster to identify words matching threatening rather than benign inferences about the outcome (e.g. ‘ fail ’ rather than ‘ succeed ’ ; Hirsch & Mathews, 1997; 2000).”

“When only one word (or non-word letter string) was presented, no comparable differences were found between groups (Macleod & Mathews, 1991; Mogg, Mathews, Eysenck, & May, 1991).”

“Support for this hypothesis comes from evidence that high worriers require more evidence in uncertain categorization tasks before arriving at a decision (Tallis, Eysenck, & Mathews, 1991).”

“The content of worry typically concerns future events whose outcomes are uncertain, but contain the possibility of one or more negative outcomes (Sibrava & Borkovec, 2006).”

“In non-anxious groups, experiments on viewing aversive fi lms have shown that instructions to worry afterwards lead to a greater frequency of subsequent intrusive thoughts related to the fi lm than imagery instructions (Butler, Wells, & Dewick, 1995).”

“Use of these methods has shown that cognitive bias modi fi cation (CBM) has signi fi cant effects on later emotional reactions to a stressful experience, even 24 h later (Mackintosh, Mathews, Yiend, Ridgeway, & Cook, 2006).”

“…the benign training condition, participants fi rst resolved homographs (single ambiguous words with both threatening and non-threatening meanings such as “ growth ” ; cf. Grey & Mathews, 2000) as cues for solving word fragments corresponding to the homographs ’ benign meaning (e.g. growth e pl_nt).”

“However, in several experimental studies, anxious individuals have been found to perform less well on tasks requiring executive control (e.g. Bishop, 2009; MacLeod & Donnellan, 1993), even when any externally imposed threat is absent.”

“In an experimental study, Fox, Zougkou, Ridgewell, and Garner (2011) have further shown that attentional biases are acquired more readily in those with the low expression allele of the serotonin transporter gene.”

“Another cognitive characteristic associated with worry is an impairment of the central executive function of working memory (Baddeley & Hitch, 1974), or ‘ attentional control ’ (Darke, 1988; MacLeod & Donnellan, 1993; Rapee, 1993).”

“Another cognitive characteristic associated with worry is an impairment of the central executive function of working memory (Baddeley & Hitch, 1974), or ‘ attentional control ’ (Darke, 1988; MacLeod & Donnellan, 1993; Rapee, 1993).”

“Similarly, with more sensitive methods based on a model that draws on both accuracy and speed data, some evidence has suggested that single threat words can be better detected by anxious participants (White, Ratcliff, Vasey, & McKoon, 2010).”

“…are usually regarded as partially automatic in that they typically operate without awareness or deliberate intent: indeed, pre-conscious attention to emotional stimuli may be a better predictor of later physiological stress reactions than self report anxiety measures (Fox, Cahill, & Zougkou, 2010).”

“Thus ambiguous threats that have been repeatedly interpreted previously in particular ways become more likely to be interpreted in that same way when encountered later, via bottom-up in fl uences (e.g. Tran, Hertel, & Joormann, 2011).”

“As noted earlier, Leigh and Hirsch (2011) demonstrated that verbal worry had more marked deleterious effects on attentional control than imagery of worry content, but that use of imagery eliminated any differences between high and low worriers.”

“As with attentional bias, differential interpretation biases emerge when the threatening meaning of an ambiguous event is not completely dominant, but is less apparent to everyone (Calvo & Castillo, 2001).”

“A fi nal therapeutic direction is suggested by the dominance of verbal linguistic processing and lack of imagery in worry (Borkovec & Inz, 1990; Hirsch, Hayes, Mathews, Perman, & Borkovec, 2012; Hoyer et al., 2001).”

“…when high worriers were directly instructed to think about a worry topic in the form of images, rather than in their usual verbal manner, they re-ported fewer subsequent negative intrusions when trying to focus attention elsewhere (Hayes, Perman, Mathews, & Hirsch, 2011; Stokes & Hirsch, 2010).”

“The training used by Hayes, Hirsch, Krebs, et al. (2010) and Hayes, Hirsch, and Mathews (2010) consisted of two consecutive tasks: the fi rst involved pairs of one worry-related and one benign word followed by a to-be-detected neutral target that was always in the prior location of the benign word…”

“The role of attention in worry has been investigated by allocating high worriers to either experimental training designed to reduce attention to worry-related information, or a control condition not intended to modify attention (Hayes, Hirsch, & Mathews, 2010; see also Krebs et al., 2010).”

“For example, highly anxious individuals reporting good attentional control disengage more rapidly from threat cues than equally anxious individuals with poor control (Derryberry & Reed, 2002; Lonigan & Vasey, 2009; Peers & Lawrence, 2009).”

“2).”

“The present model borrows from several other theoretical views and incorporates the contribution of inappropriate beliefs about worry (Wells, 2006), intolerance of uncertainty (Dugas et al., 1998), and the role of verbal processing in worry (Borkovec & Inz, 1990).”

“Instead of this process being restrained by intentional top-down control, as in a non-vulnerable individual, attention to negative thought content takes up the very cognitive resources that would be needed to re-focus attention elsewhere.”

“The ease with which processing bias can be modi fi ed has been found to predict success in the treatment of anxiety (Clarke, Chen, & Guastella, 2012) suggesting that some treatment failures may be attributable to lack of changes in these underlying processing biases.”

“Traumatic experiences in adults or children can also result in negative processing biases (e.g. Lindstrom et al., 2011).”

“Conversely, engaging in worry can have the effect of inducing attention to threat: interference effects due to threat were found to be greater after a worry period that after a control period of mental arithmetic (Oathes, Squillante, Ray, & Nitschke, 2010).”

“Several research groups (e.g., Davey, 2006; Ruscio & Borkovec, 2004; Wells, 1995) have noted that excessive worriers sometimes endorse certain advantages for worry (e.g. that worry helps them to solve problems, or allows cense. dangers to be avoided), despite also endorsing negative beliefs (e.g.,…”

“…evidence of differential bias was absent when emotional categorizations were predictable (i.e. when only decisions of one type occurred in a block), but was present when the decision required was unpredictable (i.e. when decision types are mixed within presentation blocks; Pury & Mineka, 2001).”

“Two well-known theories in this respect are the ‘Attentional Control Theory’ (Eysenck and Derakshan, 2011) and the ‘Cognitive Model’ of pathological worry (Hirsch and Mathews, 2012).”