Time course of amygdala activation during aversive conditioning depends on attention

Abstract

The time course of amygdala activation during aversive conditioning is a matter of debate. While some researchers reported rapid habituation, others found stable or no amygdalar responses to conditioned stimuli at all. In the present event-related fMRI study, we investigated whether the activity of the amygdala during aversive conditioning depends on attentional conditions. Subjects underwent aversive delay conditioning by pairing an electrical shock (unconditioned aversive stimulus) with a visual conditioned stimulus (CS+). For each singular presentation of the CS+ or a nonconditioned visual stimulus (CS−), subjects attended in random order to features that either differed between both stimuli (identification task) or that did not differ (distraction task). For the identification task trials, increased responses of the left amygdala to CS+ versus CS− were rapidly established but absent at the end of the conditioning trials. In contrast, under the distraction condition, amygdala activation to CS+ versus CS− was present during the late but not the early phase of conditioning. The results suggest that the time course of amygdala activity during aversive associative learning is strongly modulated by an interaction of attention and time.

Introduction

Lesion studies in both animals and humans suggest a critical role of the amygdala in aversive conditioning (Bechara et al., 1995, LaBar et al., 1995, LeDoux, 1996). In accordance with this hypothesis, a number of functional imaging studies showed increased activation of the amygdala to conditioned aversive stimuli in humans (e.g., Büchel et al., 1998, Büchel et al., 1999, LaBar et al., 1998, Morris et al., 1998, Morris et al., 2001, Critchley et al., 2002, Phelps et al., 2004). However, several studies failed to find this effect (e.g., Knight et al., 1999, Knight et al., 2004a, Fischer et al., 2002). Furthermore, a rapid decrease or habituation of amygdalar responses with increasing number of conditioning trials has been described (Büchel et al., 1998, Büchel et al., 1999, LaBar et al., 1998, Morris et al., 2001, but see Critchley et al., 2002, Tabbert et al., 2005). This might indicate that amygdala activity is contingent with a fast establishment of the association between unconditioned (UCS) and conditioned stimulus (CS) and with the strength of the conditioned responses rather than with the representation of the acquired UCS–CS relation (Büchel et al., 1998, Büchel et al., 1999, Cheng et al., 2003).

However, whether the observed decrease of brain activation during aversive conditioning is a function of time per se or influenced by the subjects’ attentional focus is still unknown. Studies, which reported a decrease, used either a passive viewing design or tasks that forced subjects to focus the attention to the stimuli. Such conditions allow for the evaluation and top-down processing of the stimuli and/or of the behavioral reactions. Outside the context of learning paradigms, numerous studies have investigated the effects of attention on amygdala activation to visual aversive stimuli. Several of these studies showed that amygdala activation to visually presented threat-related stimuli seems to be more pronounced during attentional distraction as compared to non-distraction conditions (e.g., Critchley et al., 2000, Lange et al., 2003, Anderson et al., 2003, Straube et al., 2004, Straube et al., 2005, Straube et al., 2006b, Williams et al., 2005). These outcomes suggest, on the one hand, that amygdala responses are elicited rather automatically (but see Pessoa et al., 2002) and, on the other hand, that amygdala activation is partly inhibited during more detailed evaluation of stimuli (Critchley et al., 2000, Hariri et al., 2000, Taylor et al., 2003). This raises the question as to whether there is also an effect of attention, especially an interaction of attention and time on amygdala activation during associative aversive learning.

Therefore, the aim of the present study was to investigate whether changes in the attentional focus during aversive conditioning affects brain responses to the CS in the amygdala and/or other areas. If the decrease in amygdala activity depends mainly on time one would expect similar time courses of amygdala activation regardless of whether subjects have to attend to the stimuli or not. However, if this decrease is modulated by attention then amygdala responses should show an interaction of attention with time leading to stronger amygdala activation to the CS+ during distracting as compared to non-distracting conditions at the end of the conditioning procedure.