From attentional blink to emotional dysregulation: a study on the cognitive and emotional interactions in depression Authors: Yingying Wang, Qi Wang, Lijie Chen, Xiaomei Dong, Tianchao Xu Published: 2026 ABSTRACT ObjectivesThis study aimed to investigate how negative emotional faces affect target identification in Major Depressive Disorder (MDD) patients using an emotion-induced blindness (EIB) paradigm, and to compare their performance with that of non-clinical controls.MethodsAn EIB task was administered to 32 patients with MDD and 32 non-clinical controls recruited between March and December 2023. Target identification accuracy was assessed using facial expressions of disgust, fear, sadness, and neutrality as distractors presented at three different lags before the target: 100 ms, 300 ms, and 700 ms.ResultsPatients with MDD showed lower overall target identification accuracy than non-clinical controls. Accuracy was lowest at lag 1, indicating stronger early-stage interference. Among the distractors, disgust faces were associated with the lowest target identification accuracy and produced stronger emotional interference than fear or sad faces. In addition, target identification accuracy following disgust and fear distractors was associated with MDD status.ConclusionsThese findings suggest that patients with MDD show reduced rapid target identification performance in an EIB task. Disgust faces may produce particularly strong interference during emotional attentional processing, and performance under disgust and fear distractor conditions may serve as behavioral indicators associated with MDD status. Introduction Major depressive disorder (MDD) is a common psychiatric disorder defined by sustained low mood and loss of interest or pleasure. In addition to emotional symptoms, patients with MDD often experience attentional deficits, slowed information processing, and impaired cognitive control [1, 2]. A key feature of depression is a reliable bias toward negative emotional information. This bias is reflected in patients’ greater tendency to attend to negative emotional stimuli and their difficulty disengaging attention from these stimuli. According to cognitive theories of depression, this attentional bias toward negative information may enhance the processing of negative experiences, thereby contributing to the development and maintenance of depressive symptoms [3–6]. Behavioral, neuroimaging, and experimental studies have also shown that attentional bias related to emotional stimuli plays an important role in the pathological mechanisms of depression [7–9]. Therefore, it is necessary to further examine how negative emotional stimuli affect attentional processing in patients with MDD. Early investigations of emotional attentional bias in depression have relied on several common paradigms, including the dot-probe task, emotional Stroop task, eye-tracking, and visual search [10–13]. Compared to the control group, research using the dot probe has shown that patients with MDD tend to fixate on sad faces [10]. Emotional Stroop research indicates that depressed individuals respond more slowly when processing negative emotional words or negative emotional stimuli, suggesting that their attentional resources may be disrupted by emotional information. Eye-tracking studies further indicate that individuals with major depressive disorder (MDD) may exhibit stronger attention maintenance or greater difficulty in attention disengagement when processing negative facial expressions, particularly under conditions involving sad or disgusted faces [14]. However, the findings remain inconsistent. Some studies have reported a pronounced attentional bias toward emotional stimuli during the early stage of processing (0–100 ms), whereas this bias becomes less stable during later stages (100–500 ms) [15]. In contrast, recent visual search studies have found no significant difference between patients with MDD and non-clinical controls in the allocation of attention to negative emotional faces [13]. Therefore, while MDD appears linked to negative attentional bias, results vary considerably across experimental tasks. The Attentional Blink (AB) refers to a brief and intermittent lapse in attention that occurs when stimuli are rapidly presented, resulting in the subsequent stimulus being missed or ignored [16]. In the traditional AB paradigm, processing the first target (T1) consumes limited attentional resources, creating a temporal bottleneck that impairs the identification of the second target (T2) [17]. The degree to which T1 captures attention thus directly influences the magnitude of the attentional blink [18]. The emotional attentional blink (EAB) paradigm builds on this effect by exploring how emotional signals modulate temporal attention [19]. EAB is often used as a broader term encompassing various forms of emotion-related impairment in rapid target perception. A more specific variant, emotion-induced blindness (EIB), occurs when a task-irrelevant emotional distractor briefly appears and disrupts the detection of a subsequent target [20, 21]. In this paradigm, emotional distractors are thought to capture attentional resources in a stimulus-driven manner, thereby reducing the resources available for subsequent target processing [20]. Therefore, unlike the traditional AB paradigm, the emotional faces in the present study were not treated as T1 targets requiring identification. Instead, they served as task-irrelevant critical emotional distractors presented before a rotated target image. This EIB design enabled us to test how negative facial expressions capture attention over rapid time courses and interfere with target identification in MDD. Existing EAB and EIB studies show that emotional faces reliably interfere with subsequent target identification, although effects differ across emotional categories. Threat-related faces, especially angry and fearful expressions, have been reported to capture attention rapidly and impair target detection when they appear shortly before the target [22, 23]. For instance, Maratos found that angry faces produced a stronger blink response than neutral and happy faces, and similar effects were also observed for fearful and sad expressions [22]. Gutiérrez-Cobo and colleagues also found reduced target recognition following angry and reward-associated faces compared with neutral faces [23]. Other emotional categories may also modulate performance. Fearful faces have been shown to impair early target detection, whereas surprise faces may show a relative processing advantage over angry faces [24, 25]. Disgust-related stimuli have also been found to capture attention and induce an attentional blink effect [26]. Despite these findings in healthy populations, research using clinically diagnosed MDD samples remains scarce. Little is known about how negative emotional distractors influence rapid attentional processing in depression. It remains unclear whether disgust, fear, and sadness produce distinct patterns of attentional capture and release across different time lags in MDD. Further clinical evidence is therefore needed to clarify these effects. The present study used an EIB paradigm to investigate how negative emotional facial distractors affect target identification in patients with MDD and non-clinical controls. Specifically, we focused on three types of negative emotional faces, namely disgust, fear, and sadness, and compared them with neutral faces to explore whether different emotional distractors produce distinct patterns of attentional interference. We also assessed accuracy across multiple temporal lags to capture how attentional capture and release change over time. Based on prior work, we predicted that patients with MDD would show lower overall target accuracy than controls, with greater impairment at shorter lags. Given the mixed evidence regarding emotion-specific attentional bias in MDD, the effects of specific emotional face categories were examined in an exploratory manner. Previous studies have commonly used temporal intervals around 200ms and 800ms in AB paradigms, and similar temporal windows have also been applied in later EAB or EIB studies [27]. However, when faces are used as distractors, attentional capture may occur rapidly, with early effects appearing around 100 ms after stimulus presentation [15]. In addition, target recognition accuracy has been reported to gradually stabilize at longer intervals, particularly around 700ms [28, 29]. Therefore, the present study used three temporal lags: lag 1, lag 3, and lag 7 (SOA 100 ms, 300 ms, 700 ms, respectively). This temporal arrangement allowed us to compare how emotional distractors influence attention during both early and later processing stages. Methods Participants The sample size was determined by an a priori power analysis using G*Power 3.1. For the mixed repeated-measures design, with group as the between-subjects factor and emotional distractor category and lag as within-subject factors, A conventional medium effect size was adopted, resulting in a required total sample size of 60 participants, with approximately 30 participants in each group, with an α level of 0.05, a power of 0.80, and an effect size of f = 0.25 [30]. This study’s sample of patients was gathered from the outpatient clinic of the General Hospital of Northern Theater Command from March to December 2023. Thirty-five patients with a preliminary diagnosis of depression according to DSM-5 criteria [31] were included. All enrolled patients were first-episode or newly diagnosed cases, screened to exclude comorbid psychiatric disorders such as bipolar disorder, and had not received antidepressants or other psychotropic medication prior to enrollment. Meanwhile, we recruited 32 volunteers from the surrounding community who had no history of mental illness as a non-clinical group. Recruitment was conducted through multiple channels, including posters placed on community bulletin boards and student notice boards, as well as online advertisements distributed via social media applications (e.g., WeChat groups). Because three participants failed to complete the task, the final sample selected for the study consisted of 64 subjects, which included 32 patients with MDD and 32 non-clinical controls (NC). All participants in the study were native speakers, right-handed individuals, and reported normal or corrected-to-normal vision. Materials Distracter images consisted of 16 images depicting four categories of emotional faces (e.g., disgust, fear, sad, and neutral), which were chosen from the Chinese Facial Affective Picture System(CFAPS) [32]. Target images comprised 40 building pictures in two rotations: 90° to the left and 90° to the right. Additionally, “filler” images (not targets or distractors) included 132 upright building photos. These 16 facial expressions (4 images of each type) were matched for arousal and intensity [33]. To balance out gender effects, each emotional type included two male and two female faces. Adobe Photoshop 2021 software has been used to process all images, with brightness standardized and resized to 600 ⅹ 600 pixels. These images were presented on a 15.6-in. OLED monitor (1920 ⅹ 1080 pixels resolution) with a refresh rate of 60 Hz and centered on a white background to ensure consistency and standardization. Stimuli were presented through E-Prime Version 3.0 software (Psychology Software Tools, Inc). Research process All participants completed the informed consent, per the guidelines outlined in the Declaration of Helsinki (1964), before the experiment began. Participants completed the general situation survey, diagnostic interview, the Hamilton Depression Scale (HAMD), and EIB tasks. The experiment received ethical approval from the Departmental Ethics Committee of the General Hospital of Northern Theater Command, and the reference number is [Y (2023) 035]. Hamilton depression scale The enrolled patients with MDD went through a HAMD scale interview by a trained interviewer, which is a widely utilized psychological instrument for standardized assessment of depression severity. All depressed patients had to score 19 or above on the scale (M = 28.72, SD = 6.7) [34, 35]. The scores of the healthy control group were all below 8 points (M = 4.62, SD = 1.29). Cronbach’s alpha of 0.78. Emotion-induced blindness task In this task, participants must accurately identify single target stimuli within a rapid serial visual presentation (RSVP) of image stimuli (Fig. 1). The target differs from other upright images in orientation, which is a rotated image (90° left or right). All participants were required to sit in front of the computer and carefully read the instructions before the experiment’s formal start. Following this, the experimental practice continued until participants indicated a complete understanding of the task. Experimentation consists of 110 trials (8 looping practice trials and 102 experimental trials divided into three blocks). In each trial, participants viewed a sequence of 14 color photographs, including 12 filler images, a critical distractor (e.g., emotional faces), and a directional rotation of the target (e.g., 90° left or right). Each image was presented on the screen for 100ms in rapid succession. A “distractor” photo was placed at varying intervals before the target. In the RSVP, distractors (e.g., emotional faces) would randomly appear between the third and sixth positions, with the target appearing downstream at varying “lag” [100 ms (lag 1), 300 ms (lag 3), or 700 ms (lag 7)]. Each distractor category was equally distributed by lag and distractors of emotional faces (e.g., disgust, fear, sad, and neutral). Before the task, participants performed eight looping practice trials (2 trials without a rotated target image, three trials with the target image rotated to the right, and three trials with the target image rotated to the left). These practice trials used neutral faces as distractors. Fig. 1Experimental procedure for the emotion-induced blindness paradigm. The distractors consisted of four kinds of emotional faces that appeared with three lags. Lag 1, the target appeared immediately after the distractor. Lag 3, the target appeared in the third position after the distractor. Lag 7, the target appeared in the seventh position after the distractor Experimental procedure for the emotion-induced blindness paradigm. The distractors consisted of four kinds of emotional faces that appeared with three lags. Lag 1, the target appeared immediately after the distractor. Lag 3, the target appeared in the third position after the distractor. Lag 7, the target appeared in the seventh position after the distractor At the beginning of each trial, a black fixation cross (+) was presented in the center of the screen for a continuous 800ms. Subsequently, after the disappearance of the fixation cross, the image stimuli would be initiated. After all the images were shown, participants were required to complete two tasks. The task assigned to participants involved detecting whether a rotated image was present (Task 1) and identifying the direction of the rotation (Task 2). In Task 1, participants responded by pressing the left response key (F) for “No” or the right response key (J) for “Yes”. In Task 2, to prevent mental rotation during responses, the experiment design matched target stimuli rotated 90° to the left with the left response key (F) and target stimuli rotated 90° to the right with the right response key (J). In cases where no target stimulus (8%) was detected, participants were instructed to press the “B” key to proceed to the subsequent trial. The entire experiment lasted approximately 20⁃25 min. Data analysis The basic design was a 2(Group; MDD, NC)× 4(emotional face; disgust, fear, sad, neutral)× 3 (lag; 1, 3, 7) mixed experimental design. In this design, the group is the between-subjects factor, while emotional faces and delay function as within-subjects factors. The chi-square test was used to compare patients’ demographics with MDD and NCs, including age, gender, educational attainment, and marital status. Participants’ target identification accuracy and emotional interference scores were collected and recorded, and repeated measures Analysis of Variance (ANOVA) was employed. Post hoc comparisons were performed using Bonferroni correction of p-values. Results Participant characteristics Previous studies have shown that individual factors, such as age and gender, can influence the strength of the attentional blink effect [37–40]. Therefore, participants in the study were matched based on these factors to control for their potential impact, while ensuring that both groups had comparable language abilities and intellectual levels [41, 42]. The descriptive statistics of demographic variables for two groups of participants (MDD, NC) are presented in Table 1. The results show that MDD patients and NCs were well-matched in all demographic characteristics with no significant group differences in age, gender, education level, or marital status (p > 0.05) . Fig. 2Main effects of emotional distractor and lag on target identification accuracy. a Target identification accuracy across emotional distractors. b Target identification accuracy across lags. Lag 1, the target appeared immediately after the distractor. Lag 3, the target appeared in the third position after the distractor. Lag 7, the target appeared in the seventh position after the distractor. *p < 0.05; **p < 0.01; ***p < 0.001 Main effects of emotional distractor and lag on target identification accuracy. a Target identification accuracy across emotional distractors. b Target identification accuracy across lags. Lag 1, the target appeared immediately after the distractor. Lag 3, the target appeared in the third position after the distractor. Lag 7, the target appeared in the seventh position after the distractor. *p < 0.05; **p < 0.01; ***p < 0.001 Table 1Descriptive characteristics and differences of the study participants; M = means; SD = standard deviationsGroupMDDNC \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$t/\chi^2$$\end{document} p Age26.88(9.489)28.97(8.11)0.949/0.110.446/0.74 18-352726 35-5556Gender (%)3.2230.20 Male5028.13 Female5071.87Education level (%)6.1170.191 Middle school 40.6318.75 High school 28.1240.63 University31.2540.62Married (%)46.88253.4670.10Total3232MDD Major depressive disorder, NC Non-clinical controlα =0.05 Table 1Descriptive characteristics and differences of the study participants; M = means; SD = standard deviationsGroupMDDNC \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$t/\chi^2$$\end{document} p Age26.88(9.489)28.97(8.11)0.949/0.110.446/0.74 18-352726 35-5556Gender (%)3.2230.20 Male5028.13 Female5071.87Education level (%)6.1170.191 Middle school 40.6318.75 High school 28.1240.63 University31.2540.62Married (%)46.88253.4670.10Total3232MDD Major depressive disorder, NC Non-clinical controlα =0.05 Descriptive characteristics and differences of the study participants; M = means; SD = standard deviations MDD Major depressive disorder, NC Non-clinical control α =0.05 Mixed-design ANOVA of target identification accuracy The descriptive statistics for participant response accuracy are presented in Table 2. As for accuracy, Mauchly’s Test of Sphericity was not significant (p > 0.05), indicating that the assumption of sphericity was met. The results of the mixed-design ANOVA are presented in Table 3. The analysis revealed significant main effects of group, emotional faces, and lag (Fig. 2). Overall, the NC group showed higher target identification accuracy than the MDD group. However, all two-way and three-way interaction effects were non-significant. This pattern suggests that the effects of group, emotional faces, and lag on target identification accuracy were statistically additive rather than interactive. Specifically, the effects of emotional distractor type and lag did not significantly differ between the MDD and NC groups. Therefore, subsequent condition-wise group comparisons were conducted as exploratory analyses. Table 2Means and standard deviations of percent accuracy by group, emotional face, and lag on the rapid serial visual presentation taskGroupEmotional faceLag 1 M(SD)Lag 3 M(SD)Lag 7 M(SD)MDDDisgust0.58(0.23)0.61(0.25)0.60(0.31)Fear0.57(0.29)0.67(0.28)0.72(0.25)Sad0.62(0.32)0.72(0.28)0.70(0.30)Neutral0.61(0.29)0.65(0.28)0.69(0.29)NCDisgust0.75(0.15)0.85(0.16)0.83(0.13)Fear0.82(0.16)0.88(0.19)0.88(0.12)Sad0.78(0.20)0.85(0.18)0.87(0.14)Neutral0.80(0.19)0.89(0.14)0.87(0.11)MDD Major depressive disorder, NC Non-clinical control. Lag 1, the target appeared immediately after the distractor. Lag 3, the target appeared in the third position after the distractor. Lag 7, the target appeared in the seventh position after the distractor Table 2Means and standard deviations of percent accuracy by group, emotional face, and lag on the rapid serial visual presentation taskGroupEmotional faceLag 1 M(SD)Lag 3 M(SD)Lag 7 M(SD)MDDDisgust0.58(0.23)0.61(0.25)0.60(0.31)Fear0.57(0.29)0.67(0.28)0.72(0.25)Sad0.62(0.32)0.72(0.28)0.70(0.30)Neutral0.61(0.29)0.65(0.28)0.69(0.29)NCDisgust0.75(0.15)0.85(0.16)0.83(0.13)Fear0.82(0.16)0.88(0.19)0.88(0.12)Sad0.78(0.20)0.85(0.18)0.87(0.14)Neutral0.80(0.19)0.89(0.14)0.87(0.11)MDD Major depressive disorder, NC Non-clinical control. Lag 1, the target appeared immediately after the distractor. Lag 3, the target appeared in the third position after the distractor. Lag 7, the target appeared in the seventh position after the distractor Means and standard deviations of percent accuracy by group, emotional face, and lag on the rapid serial visual presentation task MDD Major depressive disorder, NC Non-clinical control. Lag 1, the target appeared immediately after the distractor. Lag 3, the target appeared in the third position after the distractor. Lag 7, the target appeared in the seventh position after the distractor Table 3Three-way mixed repeated-measures ANOVA resultsdfF p η2pLag2,12426.769< 0.001***0.302Group1,6217.433< 0.001***0.219Emotional faces3,1865.8730.001**0.087Group × Lag2,1240.2730.7600.004Group × Emotional faces3,1861.6610.1770.026Lag × Emotional faces6,3720.5280.7870.008Group ×Emotional faces ×Lag6,3721.4140.2080.022*p < 0.05, **p < 0.01, ***p < 0.001 Table 3Three-way mixed repeated-measures ANOVA resultsdfF p η2pLag2,12426.769< 0.001***0.302Group1,6217.433< 0.001***0.219Emotional faces3,1865.8730.001**0.087Group × Lag2,1240.2730.7600.004Group × Emotional faces3,1861.6610.1770.026Lag × Emotional faces6,3720.5280.7870.008Group ×Emotional faces ×Lag6,3721.4140.2080.022*p < 0.05, **p < 0.01, ***p < 0.001 Three-way mixed repeated-measures ANOVA results *p < 0.05, **p < 0.01, ***p < 0.001 Post hoc comparisons Bonferroni-corrected post hoc comparisons were conducted for the significant main effects of lag and emotional distractor category. The significant main effect of group showed that the NC group had higher target identification accuracy than the MDD group (NC: mean = 0.840, SE = 0.033; MDD: mean = 0.644, SE = 0.033; p < 0.001) (Fig. 3). Fig. 3Overall target identification accuracy in the two groups. Each point represents one participant. Boxes indicate the interquartile range, horizontal lines indicate medians. MDD, major depressive disorder; NC, non-clinical control; *p < 0.05; **p < 0.01; ***p < 0.001 Overall target identification accuracy in the two groups. Each point represents one participant. Boxes indicate the interquartile range, horizontal lines indicate medians. MDD, major depressive disorder; NC, non-clinical control; *p < 0.05; **p < 0.01; ***p < 0.001 For the lag main effect, target identification accuracy at lag 1 was significantly lower than that at lag 3 and lag 7 (lag 1: mean = 0.692, SE = 0.025; lag 3: mean = 0.764, SE = 0.025, p < 0.001; lag 7: mean = 0.770, SE = 0.024, p < 0.001). No significant difference was found between lag 3 and lag 7 (p = 1.000) (Fig. 2b). For the emotional faces main effect, target identification accuracy following disgust distractors was significantly lower than that following fear, sad, and neutral distractors (disgust: mean = 0.705, SE = 0.023; fear: mean = 0.755, SE = 0.024, p = 0.006; sad: mean = 0.756, SE = 0.027, p = 0.009; neutral: mean = 0.753, SE = 0.026, p = 0.026). No significant differences were observed among the fear, sad, and neutral distractor conditions (all ps > 0.05) (Fig. 2a). EIB interference score analysis To further examine the additional interference produced by negative emotional faces relative to neutral faces, we calculated an emotional interference score. This score was defined as target identification accuracy in the neutral-face condition minus target identification accuracy in the negative emotional-face condition at the same lag. A higher score indicated stronger additional interference caused by the negative emotional face relative to the neutral face. A negative score indicated that target identification accuracy was higher in the negative emotional-face condition than in the neutral-face condition. Descriptive statistics for the emotional interference scores are presented in Table 4. A mixed-design ANOVA on the emotional interference scores revealed a significant main effect of emotion type (F2,124=8.281, p < 0.001, η2p = 0.118), indicating that the relative interference effects differed among different types of negative emotional faces. Bonferroni-corrected post hoc comparisons showed that disgust-related interference scores were significantly higher than fear-related and sadness-related interference scores(ps < 0.01), whereas no significant difference was found between fear-related and sadness-related interference scores (Fig. 4). This suggests that disgust distractors produced stronger additional interference relative to neutral distractors than fear or sad distractors. However, the main effects of group and lag were not significant, and none of the interaction effects reached statistical significance. These findings suggest that the emotional interference effect was mainly reflected in differences among negative emotion types. Table 4Means of EIB interference scores by group, emotional face, and lagGroupEmotional faceLag 1 M(SD)Lag 3 M(SD)Lag 7 M(SD)MDDDisgust0.033(0.236)0.046 (0.192)0.083(0.182)Fear0.046(0.219)-0.014 (0.120)-0.028(0.229)Sad-0.006(0.229)-0.059(0.193)-0.014(0.203)NCDisgust0.051(0.188)0.040(0.189)0.036(0.169)Fear-0.021(0.181)0.018(0.177)-0.013(0.133)Sad0.021(0.163)0.044(0.219)-0.006(0.171)MDD Major depressive disorder, NC Non-clinical control. Lag 1, the target appeared immediately after the distractor. Lag 3, the target appeared in the third position after the distractor. Lag 7, the target appeared in the seventh position after the distractor Table 4Means of EIB interference scores by group, emotional face, and lagGroupEmotional faceLag 1 M(SD)Lag 3 M(SD)Lag 7 M(SD)MDDDisgust0.033(0.236)0.046 (0.192)0.083(0.182)Fear0.046(0.219)-0.014 (0.120)-0.028(0.229)Sad-0.006(0.229)-0.059(0.193)-0.014(0.203)NCDisgust0.051(0.188)0.040(0.189)0.036(0.169)Fear-0.021(0.181)0.018(0.177)-0.013(0.133)Sad0.021(0.163)0.044(0.219)-0.006(0.171)MDD Major depressive disorder, NC Non-clinical control. Lag 1, the target appeared immediately after the distractor. Lag 3, the target appeared in the third position after the distractor. Lag 7, the target appeared in the seventh position after the distractor Means of EIB interference scores by group, emotional face, and lag MDD Major depressive disorder, NC Non-clinical control. Lag 1, the target appeared immediately after the distractor. Lag 3, the target appeared in the third position after the distractor. Lag 7, the target appeared in the seventh position after the distractor Fig. 4Emotional interference scores for negative emotional face distractors. EIB interference scores were defined as accuracy in the neutral-face condition minus accuracy in the negative emotional-face condition at the same lag. Higher scores indicate stronger interference relative to neutral faces. a Emotional interference scores across the negative emotional distractor. b Emotional interference scores in the NC and MDD groups. MDD, major depressive disorder; NC, non-clinical control; **p < 0.01 Emotional interference scores for negative emotional face distractors. EIB interference scores were defined as accuracy in the neutral-face condition minus accuracy in the negative emotional-face condition at the same lag. Higher scores indicate stronger interference relative to neutral faces. a Emotional interference scores across the negative emotional distractor. b Emotional interference scores in the NC and MDD groups. MDD, major depressive disorder; NC, non-clinical control; **p < 0.01 Association between emotional-face-related target accuracy and MDD status Correlation analyses were conducted to examine the associations between target identification accuracy following different emotional faces and MDD status. Our results showed significant linear correlations (p < 0.05) among all variables, as depicted in the correlation heatmap (Fig. 5). Multiple collinearity diagnostics revealed that each predictor’s variance inflation factor (VIF) was below 5. The condition index was less than 30, indicating no serious multicollinearity issues. A backward stepwise binary logistic regression analysis was then conducted with diagnostic status as the dependent variable. Target identification accuracy following disgust and fear distractors was retained in the final model. The Hosmer-Lemeshow test (χ2 = 11.564, p = 0.172) indicated a good model fit. Regression analysis results (Table 5) revealed that the odds ratios (OR) for disgust and fear faces were both less than 1, with regression coefficients (B) < 0. These results indicate that higher target identification accuracy following disgust and fear distractors was associated with lower odds of belonging to the MDD group. Fig. 5Correlation heat map. MDD, whether it is a major depressive disorder. Red indicates a positive correlation, while blue indicates a negative correlation. The darker the color, the stronger the correlation Correlation heat map. MDD, whether it is a major depressive disorder. Red indicates a positive correlation, while blue indicates a negative correlation. The darker the color, the stronger the correlation Table 5Results of binary logistic regression analysisBSEWald p OR (95%CI)Disgust-3.4021.00511.4510.001**0.033 (0.005 ~ 0.239)Fear-2.5040.956.9460.008**0.082 (0.013 ~ 0.526)Cox & Snell R-squared: 0.234Nagelkerke R-squared: 0.312**p<0.01 Table 5Results of binary logistic regression analysisBSEWald p OR (95%CI)Disgust-3.4021.00511.4510.001**0.033 (0.005 ~ 0.239)Fear-2.5040.956.9460.008**0.082 (0.013 ~ 0.526)Cox & Snell R-squared: 0.234Nagelkerke R-squared: 0.312**p<0.01 Results of binary logistic regression analysis Cox & Snell R-squared: 0.234 Nagelkerke R-squared: 0.312 **p<0.01 Discussion This study used an emotion-induced blindness (EIB) paradigm to examine how negative emotional face distractors influenced target identification in patients with MDD and non-clinical controls (NC) [36]. Three results are central to the interpretation. First, the MDD group showed lower target identification accuracy than the NC group. Second, accuracy was lowest at lag 1, showing that target identification was most disrupted when the emotional distractor and the target appeared in close temporal proximity. Third, disgust distractors led to lower target identification accuracy than fear, sadness, and neutral distractors. Because the interaction effects were not significant, the emotion-related effects should be interpreted as effects observed across groups rather than as evidence for a specific emotional response pattern in MDD. According to the spatiotemporal competition account of the attentional blink, emotionally salient distractors can automatically capture attention and initiate bottom-up processing when they appear in close temporal proximity to a target [37]. Under this condition, emotional distractors and target stimuli compete for limited attentional resources, which reduces the resources available for target processing [38]. The present data fit this temporal account: target identification accuracy was significantly lower at lag 1 across groups. Accuracy improved as the distractor–target interval increased [24, 25, 39]. This pattern suggests that emotional distractors exert their strongest influence during the early temporal window of target processing [40]. Accuracy did not differ significantly between lag 3 and lag 7, which means that attentional allocation had largely recovered by 300 ms in the present task. Compared with studies in which recovery has often been assessed at approximately 800 ms, the present result places the recovery point earlier under the current stimulus and task conditions [20, 41, 42]. The group effect was expressed as lower overall target identification accuracy in patients with MDD than in NC participants. This result points to a general reduction in rapid target identification performance in the MDD group, rather than a selective deficit for one emotional category or one lag condition [7]. Previous studies have linked MDD with emotional dysregulation, impaired cognitive control, and reduced attentional control [43, 44]. In the present task, emotional faces served as task-irrelevant distractors, and the lower accuracy in the MDD group therefore reflects weaker performance under rapid attentional competition. This interpretation remains at the level supported by the current data: MDD was related to reduced overall accuracy, but the non-significant interaction effects do not support a group-specific emotional interference pattern. Our study revealed that disgust distractors produced a stronger interference effect than fear, sad, and neutral distractors. The emotional interference score analysis led to the same conclusion, with disgust-related interference stronger than fear- or sadness-related interference. This disgust effect was observed across groups. It suggests that disgust faces may be particularly effective in capturing attentional resources and disrupting subsequent target identification. This interpretation is consistent with previous evidence that disgust-related or emotionally salient facial expressions can have a pronounced impact on attentional processing [44, 45]. Moreover, under conditions of spatiotemporal competition, disgusted faces may be more likely to dominate attention and reduce the allocation of attentional resources [26]. However, sad faces did not produce the lowest target identification accuracy in the raw accuracy results. In the emotional interference score analysis, sadness-related interference scores in the MDD group were descriptively negative across lags. This means that, in this dataset, sad face distractors did not create additional interference relative to neutral distractors and were not linked to poorer target identification in the MDD group. This result differs from studies reporting attentional bias toward sad information in depression [10, 46]. One reason for this difference is the task structure. The present EIB task measured rapid interference from task-irrelevant face distractors, whereas other studies have often examined sustained attention to emotional material or explicit processing of emotional stimuli. For example, Milders and his colleagues used an AB task in which emotional faces appeared as targets rather than task-irrelevant distractors, and found no significant group difference in the identification of sad faces [47]. Recent work has also related sad mood to changes in cognitive effort, but this mechanism was not directly tested here [48]. This study also examined whether target identification accuracy following different emotional distractors was related to MDD status. We found a negative correlation between correct identification rates and depression when disgust, fear, sadness, and neutral faces were used as distractors. In the logistic regression analysis, target identification accuracy following disgust and fear distractors was associated with MDD status. Higher accuracy under these two distractor conditions corresponded to lower odds of belonging to the MDD group. This result does not mean that disgust or fear produced a selective interference effect only in MDD. Rather, accuracy under disgust and fear distractor conditions carried group-relevant information in the regression model. Fear-related stimuli may rapidly capture attentional resources and interfere with subsequent target processing [40]. Reduced attentional control and cognitive inflexibility in MDD may also contribute to poorer performance in rapid attentional competition tasks [49, 50]. Prior evidence indicates that neural hyperactivation in response to fearful faces and atypical attentional disengagement from disgusted faces are closely linked to depression-related emotional and attentional processing [51, 52]. In the present study, accuracy following disgust and fear distractors functioned as behavioral indicators associated with MDD status. Overall, the present results show that patients with MDD had lower target identification accuracy in an EIB task. Target identification was most vulnerable at the shortest lag, and disgust faces produced stronger overall interference than fear or sad faces. This stronger interference may be related to the ability of disgust faces to activate stimulus-driven attentional processing, thereby increasing competition for limited attentional resources [19]. In patients with MDD, the lower overall task performance may reflect reduced attentional control or less efficient cognitive resource allocation [50]. Thus, the EIB task offers a quantifiable behavioral approach for exploring the interaction between cognition and emotion. Previous studies have shown that attentional bias toward negative stimuli, particularly sadness, can affect the outcomes of attentional bias modification training (ABMT) [53]. ABMT has been shown to alleviate difficulties in attention allocation over time, thereby reducing the attentional blink effect [54, 55]. Consequently, we propose that emotion-induced blindness tasks may be useful for assessing changes in emotional attention processing during ABMT and may provide a behavioral measure for evaluating its potential role in alleviating depressive symptoms [56]. It would be useful to test whether disgust- and fear-related distractors can be applied in cognitive training paradigms for patients with MDD [57]. Conclusion This study found that patients with MDD showed lower target identification accuracy on an EIB task than NC participants. Emotional interference was strongest at the shortest lag, indicating that target identification was most affected during the early stage of attentional processing. Among the negative emotional faces, disgust faces produced the strongest interference effect, while sad faces showed no clear additional interference relative to neutral faces in patients with MDD. Target identification accuracy following disgust and fear distractors was also associated with MDD status in the logistic regression analysis. These findings provide evidence that EIB tasks can be used to examine rapid emotion-related attentional processing in MDD and may help clarify the interaction between emotional dysregulation and cognitive control in depression. Limitation Several limitations need to be considered. First, this study included negative and neutral facial expressions but did not include positive emotional stimuli such as happy faces. This design was appropriate for testing negative emotional distractors, but it cannot describe the full pattern of emotional processing in MDD. Studies including both positive and negative emotional faces would provide a broader comparison [13, 47]. Second, although the facial stimuli were selected from a standardized picture system and were matched for arousal and emotional intensity, some low-level visual features, including contrast, spatial frequency, and specific facial characteristics, were not fully controlled. These features could have influenced attentional capture. Third, the sample size was relatively small, which may have limited the statistical power to detect subtle interaction effects, especially given the heterogeneity of MDD. Finally, the cross-sectional design limits causal interpretation. Future longitudinal studies with larger samples, more detailed clinical assessments, and broader emotional stimuli are needed to further clarify emotion-related attentional processing in MDD. In addition, detailed clinical characteristics were not fully examined. Future studies should consider these factors because they may affect cognitive control and emotional processing.