People with atopic dermatitis show no attentional bias to individually adjusted threat words

Atopic dermatitis (AD) is a common chronic inflammatory skin condition affecting around 20% of children and 3% of adults worldwide.1 Beyond its visible symptoms, AD carries a substantial psychosocial burden. Many patients experience embarrassment, avoidance of social contact and reduced quality of life.2 These difficulties raise an important question for clinicians: are patients' feelings of distress maintained by heightened, automatic attention to disease-related threat cues, a so-called attentional bias? Attentional bias is thought to reflect the preferential capture of attention by information that is personally relevant or threatening. In dermatology, this mechanism has been discussed as a possible contributor to stigma and emotional distress,3 but direct evidence remains scarce. We therefore tested the hypervigilance hypothesis in AD: the idea that disease- or stigma-related cues (e.g. ‘itchy’, ‘ashamed’) automatically draw or hold attention. Across four pre-registered experiments, we examined whether individuals with AD show greater attentional bias towards personally relevant threat words than matched healthy controls (see Table 1). To ensure ecological validity and to overcome a limitation of a previous study,4 each patient generated their own set of AD-related words, which were paired with neutral controls of similar length and frequency. Attentional bias was assessed using a well-established spatial cueing paradigm,3, 5 in which words appeared briefly to the left or right of fixation before a target detection task. Two stimulus-onset asynchronies (SOAs) were used to capture distinct phases of attention: 300 ms (reflecting earlier, more automatic engagement) and 1300 ms (reflecting later, more strategic attention).6, 7 Across all experiments, participants demonstrated robust validity effects, with slower responses when targets appeared in unexpected locations, indicating good task engagement. However, there was no evidence that patients with AD oriented attention preferentially towards disease-related words. Bayesian analyses supported the absence of the critical interaction between Group, Word Type and Validity in three of the four experiments, with one experiment yielding an inconclusive result (see Table 2). These findings converge with our previous study using standardized stimuli,4 suggesting that, within this experimental framework, attentional bias is not a reliable feature of AD. Although null findings can appear less intuitive, they have important clinical implications. The absence of a measurable attentional bias suggests that there is currently no evidence that the emotional burden of AD stems from automatic attentional capture of threat cues. Instead, psychosocial distress may arise from broader processes—such as illness perceptions, self-stigma and maladaptive coping—that are amenable to psychological interventions.8, 9 Clinically, these results underscore the value of routinely screening for emotional and social impact alongside symptom severity, using brief tools such as the Dermatology Life Quality Index.10 Where stigma or avoidance is evident, cognitive–behavioural approaches that target negative beliefs about visibility, rejection and controllability may offer greater benefit than attention-retraining techniques.9 In summary, our pre-registered study provides robust evidence that is incompatible with the assumption of a stable attentional bias to disease-related cues in AD. Rather than focusing on automatic cognitive mechanisms, future research and clinical practice might prioritize addressing shame, stigma and emotional resilience as key determinants of patient well-being. A full description of the study, including all materials and data, can be found at https://doi.org/10.31234/osf.io/n97jr_v1. This research was supported by a grant from the National Eczema Association (grant no. NEA23-ECRG190). HH has received funding from the National Eczema Association. The funder had no influence on the design or interpretation of the research. HH has no other conflicts of interest to declare. CN, AvL and SW all have no conflicts of interest to declare. All procedures were conducted in accordance with the Helsinki Declaration (1975). Ethical approval for the study was obtained from the Faculty of Health Sciences committee at the University of Hull. Images of patients or any other information that could identify an individual patient are not shared in this publication. This study adheres to the STROBE guidelines for reporting case–control studies. We report our sample size determination, all data exclusions, manipulations and measures used. All four experiments were pre-registered prior to data collection (see https://aspredicted.org/pzbc-6trf.pdf, https://aspredicted.org/9bdd-25vp.pdf, https://aspredicted.org/dd58-h3xc.pdf, https://aspredicted.org/5wj6-s252.pdf). All data, analysis code and research materials are available at https://osf.io/6vfsw/?view_only=5665b8abdd2543c29480df97d5a33ca3.