This meta-analysis investigated whether attentional bias, i.e. the preferential allocation of attention to information that is related to pain, is a ubiquitous phenomenon. We also investigated whether attentional bias effects are related to the methodological quality of the study, to procedural differences in their measurement, or to individual differences in pain severity, pain-related fear, anxiety and depression. Results indicated that individuals who experience chronic pain (n=1023) display an attentional bias towards pain-related words or pictures, but this bias was of a small effect size (d=0.134), and did not differ from that in control groups (d=0.082; n=1398). No evidence was found for an attentional bias towards pain-related words and pictures for acute pain (d=0.049), procedural pain (d=0.142), and experimental pain (d=0.069). However, research in which attentional bias towards signals of impending experimental pain in healthy volunteers was investigated, revealed an attentional bias of medium effect size (d=0.676). Moderator analyses in the chronic pain group identified important procedural variables that affected the presence and magnitude of an attentional bias towards pain-related words and pictures, i.e. type
INTRODUCTIONAttending toward, dwelling upon, and switching away from pain, have emerged as core components of many cognitive-affective models that seek to explain pain, distress and disability [45,77,154]. Particularly influential is the idea that patients selectively attend to pain at the cost of other information in the environment. This idea has been variably discussed as somatosensory amplification [12,98], hypervigilance [23,35], and more recently as attentional bias [81,120,129,135].The concept of attentional bias was originally introduced by information processing accounts of psychopathology [13,47,92,103]. Attentional bias, or preferentially attending to information that is related to the content of the emotional concerns of patients, has proven to be a robust phenomenon in many forms of psychopathology [21,25,50,117,165]. For example, patients with phobic and anxiety disorders display an attentional bias to threat-related words or pictures [11]. In many of these models biased information processing is not considered epiphenomenal, but instead is invoked as a predisposing, initiating, exacerbating, or maintaining feature of the disorder [11,58,93]. Although much research is correlational, some longitudinal and interventional studies support these accounts [58].Attentional bias to pain-related information is also the subject of significant research activity in pain [4,9,81,120,154,157]. Pain researchers have typically adopted hypotheses and paradigms from psychopathology research. In the first study on this topic, Pearce and Morley (1989) adapted the modified Stroop task, and presented pain patients with cards containing colored words [116]. They instructed participants to name the color of the words while ignoring word meaning. In comparison with control subjects, patients ...