Response inhibition and attentional control in anxiety

Abstract

Traditionally, anxiety has been associated with a selective attentional bias for threat and a decreased capacity in attentional control. In two different experiments, we investigated whether individuals with different levels of self-reported state anxiety (Experiment 1) and induced anxiety (Experiment 2) had impaired response inhibition processes (attentional control deficit) as characterized by a different response style in the presence of negative stimuli under low and high perceptual load conditions. A go/no-go paradigm with emotional distractors (angry, happy, and neutral faces) was used to provide measures of perceptual sensitivity, inhibition, and response style. Our findings showed that perceptual sensitivity, as assessed by the d′ parameter of signal detection theory, was reduced in all participants for angry faces under low perceptual load, where enough perceptual resources were available to be attracted by distractors. Importantly, despite similar perceptual sensitivity, the beta parameter indicated that high state anxiety individuals in both experiments were less flexible at adjusting to task demands in the presence of angry face distractors by adopting a stricter criterion. Implications of findings are discussed within current models of attentional control in anxiety.

Keywords:

• Self-reported anxiety
• Anxiety induction
• Response inhibition
• Go/no-go paradigm
• Perceptual load
• Emotional distractors
• Response style

Acknowledgments

This work was supported by a research position for Antonia Pilar Pacheco-Unguetti funded by the Junta de Andalucía, Secretaría General de Universidades, Investigación y Tecnología (HUM1017), and research grants funded by the Spanish Ministerio de Ciencia y Tecnología and the Junta de Andalucía (P07-SEJ-03299 to Alberto Acosta, CONSOLIDER-INGENIO2010 CSD2008-00048 and PSI2008-03595PSIC to Juan Lupiáñez). The first experiment was carried out by the first author under the supervision of Nazanin Derakshan at the Laboratory of Affective and Cognitive Neuroscience at Birkbeck, University of London.

Notes

¹ This interaction was also significant when the analysis was performed with STAI State as a continuous predictor, Face × Load × Anxiety–SA, F(2, 212) = 3.40, p = .0366. While the Face × Anxiety–SA interaction was significant for low load conditions, F(2, 212) = 3.91, p = .0228, it was far from significance for high load conditions (F < 1).

² The pictures' numbers for negative induction were 3000, 3071, 3080, 3150, 3170, 3350, 3550, 6312, 9040, and 9410 (mean valence = 1.78 and arousal = 7.49), and for positive induction the numbers were 2040, 2091, 2340, 2501, 2540, 4599, 5260, 5830, 8540, and 8600 (mean valence = 7.77 and arousal = 4.41). IAPS values range from 1 to 9. The differences in valence, F(1, 18) = 1,393,128, p < .0000, and arousal, F(1, 18) = 113,029, p < .0000, were significant between the two sets of pictures.

³ The goal of inducing a low anxiety mood by means of positive pictures was to increase the differences in the anxiety levels between groups, more than to produce a positive mood. The negative mood induction gave rise to a greater state anxiety level, some increase in hostility and depression, and a higher reduction in happiness (this group is named high state anxiety throughout the paper). In the group receiving the positive induction (defined as low state anxiety), participants had a reduction in anxiety, hostility, and depression levels and an increase in happiness.

⁴ The EVEA is a scale that includes 16 items (adjectives referring to mood states) with four factors evaluated in a Likert scale (ranging from 0 to 10): Fear–Anxiety, Anger–Hostility, Sadness–Depression, and Joy–Happiness. The alpha coefficients for these factors fluctuate from .88 to .93.