Contributions of Amygdala and Striatal Activity in
Emotion Regulation
Todd A. Hare, Nim Tottenham, Matthew C. Davidson, Gary H. Glover, and B.J. Casey
Background: Emotional information can facilitate or interfere with cognitive processes. In this study, we examined the influence of
emotional information in biasing performance and the biological basis underlying this influence.
Methods: Ten human subjects (five female) were scanned with functional magnetic resonance imaging while performing an
emotional go/nogo task.
Results: Subjects were slower to approach fearful target expressions and had more difficulty avoiding happy nontarget expressions.
The amygdala was recruited most for negative emotional context, and activity in this region was positively correlated with response
time when detecting negative expressions. Increased signal in the right caudate nucleus was observed when avoiding nontargets and
was negatively correlated with the number of false alarms subjects made.
Conclusions: Emotional context can alter behavioral and biological responses when approaching or avoiding a stimulus. We showed
that recruitment of the amygdala, a region implicated in evaluating emotional significance, was associated with longer response
latencies when approaching negative information, whereas recruitment of the caudate nucleus, a structure previously implicated in
reward and impulse control, was most active when avoiding positive information. Our findings have significant implications for
exaggerated and inhibited emotional responses that are characteristic of a number of psychiatric disorders.
Key Words: Amygdala, striatum, emotion regulation, go/nogo, fMRI
E
motional values assigned to objects and events play an
important role in cognitive processes. Interpretations of
objects and events, however, can change as a function of
emotional context (Kim et al 2004; Russell and Fehr 1987). The
psychologic construct of emotion regulation has been defined as
the dynamic influences between emotions and other psychologic
or physiologic processes (Campos et al 1994; Cole
et al 1994; Fox 1994). As such, emotion regulation is not the
simple presence of an emotion like happiness or sadness, but the
manner in which an emotion facilitates or interferes with other
processes (Cole et al 1994; Frijda 1988). Processes like cognitive
or attentional control (Derryberry and Rothbart 1988, 1997;
Rothbart and Ahadi 1994; Rothbart et al 1994) and delay of
gratification (Mischel 1958, 1961, 1966; Mischel and Metzner
1962; Mischel and Underwood 1974; Toner and Smith 1977;
Walls and Smith 1970) can modulate emotional influences and
thus play an important role in emotion regulation. The objective
of this study was to examine the influence of emotional infor-
mation in biasing performance during a go/nogo task and the
biological basis underlying this influence.
Emotional modulation of cognitive control has been reported
as attentional biases to positive and negative information (Der-
ryberry 1988; Mogg et al 2000). Behavioral responses are faster or
slower in different emotional contexts and vary as a function of
subject traits (e.g., negative attention bias in affective disorders;
Vasey et al 1996). Attentional bias toward negative or threatening
information is hypothesized by some to play a major role in the
etiology and maintenance of anxiety and depression (Beck 1967;
Bower 1981; Lonigan et al 2004; Mathews and MacLeod 1994;
Vasey and MacLeod 2001; Williams et al 1997). This bias in the
processing of negative emotional information in affective disor-
ders has been assessed through experimental studies examining
the behavioral performance of subjects on tasks such as the
Emotional Stroop (Williams et al 1996). Studies from our group
have found that positive and negative emotional valence can
affect cognitive control in normal populations as well (Casey and
Tottenham, unpublished data).
Neuroimaging and lesion studies (see Phillips et al 2003a)
have described neural substrates of affective processing,
especially in the amygdala, and demonstrated its influence on
behavior (Adolphs et al 1998; Bechara et al 2003). Patients
with bilateral lesions of the amygdala are impaired in judging
emotional facial expressions and make suboptimal decisions
(see Bechara et al 2003). Negatively valenced stimuli in-
creased blood oxygen level– dependent (BOLD) signal in the
amygdala, and reaction times to negative stimuli were slower
compared with neutral stimuli (Monk et al 2003; Simpson et al
2000). Likewise, positively valenced stimuli can also increase
activity in regions of the amygdala (Breiter et al 1996; Canli
et al 2002; Garavan et al 2001; Killgore and Yurgelun-Todd
2001; Pessoa et al 2002; Somerville et al 2004; Whalen et al
1998b; Yang et al 2002), and happy expressions (positive
stimuli) have been found to speed reaction times (Casey and
Tottenham, unpublished data) and bias decisions (O’Doherty
et al 2003). Thus the amygdala plays an important role in the
processing of emotional information, although amygdala ac-
tivity and emotional influences can be attenuated by other
cognitive processes (Hariri et al 2000). Decreases in amygdala
activity are seen when subjects attempt to reduce emotional
responses to affective stimuli (Beauregard et al 2001; Levesque
et al 2003; Ochsner et al 2002; Schaefer et al 2002). Performing
a cognitive task can also reduce the amygdala response to
emotional stimuli (Hariri et al 2000, 2003; Monk et al 2003).
Understanding the interactions between cognitive and emo-
tional processes is important for understanding behavior in
both normal and clinical populations. Therefore, the details of
these cognitive and emotional interactions have become the
focus of an increasing number of studies but at this point
remain unclear.
From The Sackler Institute (TAH, NT, MCD, BJC), Weill Medical College of
Cornell University, New York; Institute of Child Development (NT), Uni-
versity of Minnesota, Minneapolis, Minnesota; and Lucas Magnetic Res-
onance Imaging Center (GHG), Stanford University, Stanford, California.
Address reprint requests to Dr. Todd A. Hare, The Sackler Institute, Weill
Medical College of Cornell University, 1300 York Avenue, Box 140, New
York, NY 10021; E-mail: toh2003@med.cornell.edu.
Received October 6, 2004; revised December 7, 2004; accepted December
28, 2004.
BIOL PSYCHIATRY 2005;57:624 – 632 0006-3223/05/$30.00
doi:10.1016/j.biopsych.2004.12.038 © 2005 Society of Biological Psychiatry
l