Anxiety, fatigue, and attentional bias toward threat in patients with hematopoietic tumors

Abstract

Cancer patients with hematopoietic tumors exhibit particularly high rates of anxiety disorders and depression, and often develop negative affect. In addition, psychological problems experienced by cancer patients impair their quality of life. When cancer patients feel anxious, they tend to direct their attention toward stimuli associated with threat in the surrounding environment. If attentional bias occurs in patients with hematopoietic tumors, who are at particular risk of developing negative affect, resolution of the bias could be useful in alleviating their anxiety. The current study examined the association between attentional bias and negative affect in patients with hematopoietic tumors and tested the hypothesis that negative affect would be more severe in those who exhibited greater attentional bias. Twenty-seven patients with hematopoietic tumors participated in the study. Reaction time (RT) was measured as the time between the presentation of the threatening and neutral images, and the subject's button press to indicate choice (neutral expressions). Eight combinations of "threatening" expressions with high emotional valence and "neutral" expressions with low emotional valence were presented. The images used to measure attentional bias were taken from the Japanese Female Facial Expression Database and had been rated as expressive of anger, sadness, or neutrality, with predetermined emotional valence. Psychological testing was performed with the Profile of Mood States (POMS). To examine the association between attentional bias and negative affect, we calculated Spearman's rank correlation coefficients for RTs and POMS. Subjects' mean RT was 882.9 (SD = 100.9) ms, and 19 of the 27 subjects exhibited slower RTs relative to healthy individuals. RT was significantly positively correlated with Tension-Anxiety (r = .679, p < .01) and Fatigue (r = .585, p < .01) subscale scores. The results of the study suggested that attentional bias toward threatening expressions could be positively correlated with the mental intensity of anxiety and fatigue in patients with hematopoietic tumors.

Fig 1. Procedure for measuring attentional bias.

Once the facial photographs had been displayed for 500 ms, the subject was instructed to press a button to choose the most neutral face. The images were presented simultaneously, one above the other, at 1,600 (vertical) × 900 pixels (horizontal). The buttons could only be pressed after the images had been displayed, and the time taken to select the neutral image was recorded. Trials were performed 128 times, and RTs of <200 or >2,000 ms were excluded from the statistical analysis as outliers.

Fig 2. Analysis of the association between RTs and POMS scores.

The horizontal axis shows POMS subscale scores and the vertical axis indicates RTs; scores for the 27 subjects are represented by dots in the gray regions. Bold boxes indicate significant correlations. There were significant positive correlations between RTs and scores for the Tension-Anxiety (r = 0.679, p < 0.01) and Fatigue (r = 0.585, p < 0.01) subscales. Correlation coefficients (r) were analyzed via Spearman's rank-order correlation analysis. n = 27, *p < .05.

Fig 3. Comparison of negative POMS scores between subjects with fast and slow RTs.

Mean RTs for the slower (n = 8) and faster (n = 19) groups were 778 (SD = 33) ms and 927 (SD = 86) ms, respectively. Both Tension-Anxiety (t = 4.108, p < .001, r = .64) and Fatigue (t = 2.724, p < .01, r = .48) subscale scores in subjects with slower RTs were significantly higher relative to those observed in subjects with faster RTs. Non-paired t test, *p < .05.