نشخوار فکری و توجه در افسردگی اساسی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|31350||2007||15 صفحه PDF||سفارش دهید||9605 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Behaviour Research and Therapy, Volume 45, Issue 11, November 2007, Pages 2664–2678
Up until recently, it had been assumed that attentional biases for negative information do not exist in depression. However studies using post-conscious exposure durations have produced contradictory results. The limitations of common attentional tasks, suitability of stimulus materials and differences in stimulus duration times may have contributed to these inconsistencies. We aimed to address many of these issues and examine attentional responses in major depression at two post-conscious exposure times. We also investigated possible roles for rumination and distraction in increasing and lessening attentional biases for negative information. We used a fully controlled experimental design to test the effects of both induced and trait rumination and distraction on attention in patients with major depression and healthy controls. Attention was assessed using the dot-probe task. The findings revealed an attentional bias for negative information in depressed patients only at the longer post-conscious exposure duration. Furthermore although this bias was not influenced by either induced or trait distraction, it was related to trait rumination. Overall, the results showed that depression is associated with a strategic attentional bias towards negative information and that this bias is stronger in individuals who habitually ruminate.
Recently there has been interest in the similarities that may exist in the cognitive and behavioural processes that underlie different psychological disorders (Harvey, Watkins, Mansell, & Shafran, 2004). It could be argued that this transdiagnostic approach is especially relevant to understanding anxiety and depressive disorders given that they are often highly comorbid (e.g. Angold, Costello, & Erkanli, 1999), share considerable joint genetic vulnerability (Kendler, 1996) and can both respond to the same drugs (Kuzma & Black, 2004). However, for many years numerous empirical studies have emphasized the differences between these disorders in terms of information-processing biases (e.g. Bradley & Mathews, 1983; Clark & Teasdale, 1982; Mathews, 1990). In turn, this research led to several theories which sought to explain the dichotomy between anxiety and depression. An example of an influential theory that upheld this dichotomy is that of Williams, Watts, MacLeod, & Mathews (1988) and Williams, Watts, MacLeod, & Mathews (1997), which proposed that attention biases operate in anxiety but not depression. This assertion would seem to be largely based on the considerable body of literature that has demonstrated an anxiety-specific attentional bias for pre-consciously presented negative information (e.g. Bradley, Mogg, Millar, & White, 1995; Mathews, Ridgeway, & Williamson, 1996; Mogg & Bradley, 1998; Mogg, Bradley, & Hallowell, 1994; Mogg, Bradley, & Williams, 1995; Mogg, Bradley, Williams, & Mathews, 1993). By contrast, in depression there is yet to be a single clear instance of attentional biases under pre-conscious conditions (Mathews et al., 1996; Mogg et al., 1995). So far these findings would seem to support Williams et al.'s theory; however, data derived from experiments with post-conscious exposure durations have provided a less clear view of attentional processing in depression. That is, in the majority of experiments using exposure durations of 500–750 ms there has been a failure to find attention biases on the emotional Stroop and dot-probe tasks with clinically depressed individuals (e.g. Bradley et al., 1995; MacLeod, Mathews & Tata, 1986;Mogg et al., 1993). However, in studies using durations of at least 1000 ms (Gotlib & Cane, 1987; Mogg et al., 1995) biases have been found, although overall these investigations have produced mixed results with some inconsistent findings even being generated within the same study (e.g. Bradley, Mogg, & Lee, 1997: Study 1 and 2). Contradictions in the findings may be related to limitations of the experimental tasks, differences in design and/or variability in stimulus duration times. This explanation has been invoked by several authors in this field. For example, Nunn, Mathews, and Trower, (1997) have said that the demonstration of attentional effects in depression may depend ‘on variations in experimental method, or on the material employed’ (p. 490). Similarly, Mogg and Bradley (2005) consider the possibility that ‘variation in methodology (e.g. stimulus presentation conditions) may explain variation in findings across studies’ (p. 31). In terms of experimental tasks, the two main paradigms used to measure attention are the dot-probe and Stroop tasks. In the dot-probe task, two word stimuli are presented in different locations on a computer screen. After the display is terminated, a neutral probe appears in the former location of one of the words. Participants’ responses to the probe are timed and used to infer the allocation of attentional resources as responses will be faster to probes that appear in an attended rather than unattended area (Posner, Snyder, & Davidson, 1980). In the Stroop task, there is only one stimulus with two dimensions (word meaning and ink colour) in the same location. Attentional biases are inferred from slower colour-naming times for negative words compared in neutral words, the assumption being that word meaning has caused interference and disrupted the participants’ ability to colour-name. The use of these different paradigms is likely to account for some of the discrepancies across findings as the dot-probe and Stroop tasks probably do not measure the same processes and it is debatable whether the Stroop task measures selective attention exclusively or even at all. Indeed, MacLeod (2005) has commented ‘The fact that colour information receives inadequate attention does not require the conclusion that attention is diverted instead to the processing of word content’ (p. 52). Furthermore, Mogg, Millar, and Bradley (2000) have concluded that ‘the Stroop task does not provide an unambiguous measure of selective attention’ and go on to say ‘it is unclear whether colour-naming interference reflects competition at the input (attentional) stage or at the output or response selection stage’ (both p. 696). This view would seem to be based on the fact that the Stroop task contains the stimulus and response factors within the same presentation. By contrast, in the dot-probe task the probe occurs after the word display and therefore the manual response is made when the word is no longer on the screen. Several other research groups have expressed similar reservations about the Stroop task and accordingly have expressed a preference for the dot-probe when measuring selective attention (e.g. Musa, Lepine, Clark, Mansell, & Ehlers, 2003). With regard to design factors that could be influencing results, while studies that have examined attentional biases in depression often have some design strengths (e.g. matching word pairs in the dot-probe task according to word frequency), several are flawed in important ways. For example, some older studies did not use depression-relevant negative words and instead used social and physical threat words (e.g. MacLeod et al., 1986; Mathews et al., 1996). Furthermore, some dot-probe experiments from this time used a design that probed all critical but not all non-critical trials, thereby creating a contingency between target words and probes (e.g. MacLeod et al., 1986; Mathews et al., 1996). The use of non-clinical populations in the majority of studies (e.g. Hill & Dutton, 1989) has also created problems as the findings may not generalize to patients with major depression. A last criticism is the inclusion of uncategorized neutral stimuli in many studies (e.g. Bradley et al., 1997: Study 1 and 2; Mogg et al., 1995). This issue is of note as earlier research by Mogg et al. (1993) has shown that word categorization can affect attentional performance especially during post-conscious trials when controlled strategic processes are involved. Indeed, not having a set of categorized neutral words violates a basic standard for any attention experiment which is to ensure that no confound exists between the types of stimuli used. This problem can be rectified by including a group of neutral words that have an organizing theme (i.e. come from the same category) as this helps reduce novelty effects and ensures comparability across word types as negative/symptom words and positive words are also themed. Finally, variability in stimulus exposure times is also a likely contributor to the inconsistent results. An examination of the post-conscious data reveals a pattern: mainly null findings with shorter post-conscious exposure durations (500–750 ms) and more mixed findings with longer post-conscious exposure durations (⩾1000 ms). One interpretation is that attentional processes in depression are highly dependent on the stimulus durations used and biases would have been found more consistently at longer durations if standardized methodologies had been used. This theory is persuasive as excessive elaborative processing of negative material is thought to be characteristic of depression (Mathews & MacLeod, 1994; Nunn et al., 1997) and only long exposures allow stimuli to be analysed in enough detail to enable such elaboration to take place. By contrast, at shorter post-conscious durations, it may not be possible for stimuli to be as extensively evaluated, any biases based on elaboration will be more fragile and the assessment of early and/or middle stages of selective attention will be favoured. Based on these considerations, we attempted to examine the time course of attentional biases in depression across two exposure durations (500 and 1000 ms) in the dot-probe task when several methodological strategies were instituted to avoid confounding and ensure stimulus relevance and comparability (e.g. all trials are probed, depression-related words are used and a group of categorized neutral words are included). We predicted that biases would be found at the longer post-conscious exposure duration but not at the shorter one, despite the numerous methodological safeguards. The latter prediction may seem unremarkable since the combined results from Stroop and dot-probe tasks have provided little evidence of biases in depression at 500 ms. However if the dot-probe data are considered alone, there are only two studies with clinical participants in the literature and they produced conflicting findings: Mathews et al. (1996) found a bias while MacLeod et al. (1986) did not. So when results generated from experiments that have used the Stroop – which provides questionable measurements of attention – are excluded, the evidence is equivocal and still needs to be clarified. In addition to serving this purpose, the central function of the 500 ms duration was to provide a comparison condition against which to test our theory that attention biases in depression are critically dependent on the stimulus durations used and only occur at the later stages of attentional processing and at longer exposure times. This notion of earlier and later stages implies a non-unitary view of attentional processing and has support from brain lesion studies with humans and animals, which indicate the involvement of three separate components: initial attentional shifting, engagement and disengagement (Peterson, Robinson, & Morris, 1987; Posner & Cohen, 1984; Posner, Walker, & Friedrich, 1984). This knowledge has caused researchers to try to hone their theories and identify just one of these stages as the site of biased attention in depression. The most commonly suggested location is the disengagement stage (e.g. Joorman, 2004; Mogg & Bradley, 1998; Nunn et al., 1997; Siegle, Steinhauer, & Thase, 2004). Moreover, it has also been proposed that biases at this stage, and the likely problems they may cause for moving attention away from negative information, may underlie the tendency to ruminate, which is common in depression (e.g. Bradley et al., 1997; Siegle et al., 2004). Independently, researchers investigating rumination have arrived at a similar conclusion. For example, Lyubomirsky and Nolen-Hoeksema (1993) have stated that ‘most people are able to disengage from rumination after a short time by participating in pleasant distracting activities. … a subset of people continue to engage in self-focused rumination…. and these people tend to experience longer periods of depressed mood’ (p. 340). And Ingram, Lumry, Cruet, and Sieber (1987) suggest that ‘depressed individuals who ruminate may not be able to disengage attention when called for’ (p. 358). However, while reference is often made to attention in rumination studies and vice versa, the experimental paradigms used to study attention in cognitive psychology are yet to be applied to the topic of rumination. The present study also had this aim. Thus, in addition to using the dot-probe task to check if attentional biases exist in depression, we also wanted to investigate the impact of rumination on attention. Furthermore, research has shown that rumination is maintaining of and maintained by the tendency to process information in an overgeneral way and that distracting manipulations can reduce this overgenerality (Watkins, Teasdale, & Williams, 2000). If this logic is continued, then distraction—via its effects on overgeneral recall—should also decrease rumination and in turn lessen any difficulty in disengaging attention from negative information. Therefore the impact of distraction was also of interest. In order to explore all of these themes, we examined the relationships between attention and both trait and induced rumination and distraction in patients with major depression and non-depressed control participants. Attention was measured using the dot-probe task and there were two testing times: baseline and following the rumination/distraction inductions. Attention and mood were assessed at both time points; trait rumination and distraction are stable characteristics (Just & Alloy, 1997; Nolen-Hoeksema, Parker, & Larsen, 1994) and so were measured only once (at baseline). We reasoned that the pattern of attentional responses of the two groups at baseline would reveal whether or not biases existed in depression; any differences in attentional responses across the two time points would reflect the impact of the rumination and distraction inductions and the associations between baseline attention and trait measures would elucidate the role of naturally occurring rumination and distraction. Compared with patients with depression, control participants have substantially fewer negative cognitions (Blaney, 1986), and so we did not expect that increasing ruminative focus or distracting from thoughts would impact greatly on attention and have made predictions reflecting this for the control group. Thus, there were three main hypotheses: First, we predicted that at baseline, depressed patients would show an attentional bias for negative words presented for 1000 ms but not for 500 ms; that these biases would not occur for positive or neutral words and that control participants would not show any biases at all regardless of exposure duration and word valence. Second, we predicted that the rumination induction would produce an increase in the attentional bias for negative words in depressed patients at 1000 ms, the distraction induction would produce a decrease and there would be no change in attentional responses for control participants. Third, we predicted that the strength of the baseline attentional bias shown by depressed patients at 1000 ms would be proportionally related to levels of trait rumination and inversely related to levels of trait distraction.