Editorial

C o n fl ic t A d a p ta tio n in G e n e ra liz e d A n x ie ty
D iso rd e r: S m a ll P a ra d ig m Tw ist, L a rg e
S c ie n tifi c L e a p

I n this issue of the Journal, Etkin and colleagues (1) have published a functional MRI
(fMRI) study of implicit affective regulation in health and anxiety that is both provoca-
tive and exciting. Provocative because it demonstrates that the elusive process of im-
plicit emotion regulation may be measurable, and exciting because this approach is
applied to a clinical question.
The study examines the effects of trial repetition on task performance, particularly for
trials with conflicting information. This work relies on more than a decade of research
investigating “the conflict monitoring theory” (2). This theory posits that the resolution
of a conflict—defined as the concomitant presence of stimuli calling for incompatible
responses—requires initial detection of the conflict as well as its subsequent resolution.
The detection site resides in medial prefrontal structures, including the dorsal anterior
cingulate cortex extending dorsally to premotor regions. Once activated, the detection
system alerts a cognitive control system that ini-
tiates conflict resolution by modulating informa-
tion processing, i.e., facilitating the processing of “The ‘conflict
the appropriate response while hampering the adaptation’ deficit
processing of the incorrect, conflictual, response.
The resolution site resides in regions of the dor-
might very well
solateral prefrontal cortex. The modulated in- explain the inability
formation-processing streams lay in posterior of generalized anxiety
brain regions (e.g., the parietal cortex), which are
dedicated to sensory-perceptual integration. The
disorder patients to
study by Etkin and colleagues draws on the criti- dampen feelings of
cal observation that performance in a conflict tri- anxiety.”
al is influenced by the level of conflict present in
the preceding trial. Specifically, performance on
a conflict trial is facilitated when preceded by a similar conflict trial. This effect, which
relies upon small behavioral adjustments, has been called conflict adaptation (3).
The paradigm used by Etkin and colleagues employs pairs of stimuli that combine fear-
ful or happy faces with the words “FEAR” or “HAPPY.” In this study, “conflict” entails the
presence of competing stimuli (e.g., “HAPPY” printed on a fearful face) in incongruent
trials, and “absence of conflict” refers to synergistic stimuli (e.g., “HAPPY” on happy faces)
in congruent trials. Subjects were instructed to identify the facial emotion (target stimu-
lus) as fast as possible, while ignoring the word (distracter stimulus). Consistent with the
conflict monitoring theory described above, the incongruent stimuli engage the conflict
detection node. This node, in turn, activates the conflict resolution system, which biases
information processing and promotes correct responding. It is at this critical point that
the focus of this study lies. The idea is that the conflict resolution node, by remaining ac-
tive, would facilitate response to a subsequent incongruent stimuli trial; that is, facilita-
tion of the response to an incongruent trial when preceded by an incongruent trial. This
facilitation, termed “conflict adaptation” is implicit, because individuals are not aware of
it. Notably, the effect can be captured by a speeded reaction time and is reflected at the
neural level by activation of a prefrontal region that initiates conflict resolution (2).
As described above, Etkin and colleagues use emotional stimuli, in contrast to the clas-
sic studies that employed nonemotional stimuli (such as the color Stroop task in which

Am J Psychiatry 167:5, May 2010 [Link] 489

F IG U R E 1 . C o n fl ic t A d a p ta tio n in Ta sk s W ith E m o tio n a l a n d C o g n itiv e In c o n g ru e n c e a

Congruent
Emotional

Dorsal anterior Parietal cortex

Amygdala

Incongruent

GREEN

Congruent

prACC
activated

GREEN
Amygdala
DLPFC activated downregulated
Incongruent

RED
Cognitive

Incongruent
a
The sequential trials above the hatched line pertain to emotional conflict adaptation. Etkin et al. (1) show that
the naming of the facial emotion in an incongruent trial (the word “HAPPY” on a fearful face) that follows a
congruent trial (the word “HAPPY” on a happy face) activates the dorsal anterior cingulate cortex (highlighted
green area). The same task in a subsequent incongruent trial (the word “FEAR” on a happy face) activates most
significantly the pregenual anterior cingulate cortex (prACC, in red). Findings suggest that this prACC activation
down-regulates the amygdala (blue arrow). The key finding is that individuals with generalized anxiety disorder
do not show this neural adaptation. The sequential trials below the hatched line pertain to cognitive conflict
adaptation. Previous works (2, 4) have shown that when the stimuli and the task do not involve emotional at-
tributes, such as in the Stroop color naming task, the region recruited during conflict adaptation is within the
dorsolateral prefrontal cortex (DLPFC, blue area) and regulates information flow in posterior areas, such as the
parietal cortex (arrow to purple area).

subjects have to identify the actual color of a word when the word itself reads a different
color, i.e., answering “green” when the word “red” is printed in green-colored type [2,
5]). Doing so, Etkin and colleagues extend previous, purely cognitive work into the do-
main of affective neuroscience. In this context of emotional stimuli, conflict adaptation
recruits the pregenual anterior cingulate cortex instead of the dorsolateral prefrontal
cortex (Figure 1). This pregenual anterior cingulate cortex activation presumably down-
regulates amygdala function, as inferred by the negative functional connectivity found
between these two regions in healthy individuals. Notably, the conflict resolution node
differs in the context of emotional stimuli (pregenual anterior cingulate cortex) versus
nonemotional stimuli (dorsolateral prefrontal cortex) (4). Of particular importance is
the novel finding that patients with generalized anxiety disorder exhibit an absence of
conflict adaptation; these individuals did not exhibit speeded reaction time or recruit-

490 [Link] Am J Psychiatry 167:5, May 2010

ment of the pregenual anterior cingulate cortex. The research and clinical implications
of this finding are impressive.
With this study, Etkin and colleagues provide a research tool to measure regulation of
emotional information that does not depend on self-reports and is sensitive to psycho-
pathology. Furthermore, these findings support existing theories of emotion regulation
that implicate the dorsolateral prefrontal cortex and anterior cingulate cortex (6). Most
of this earlier research has used strategies that instruct study participants to modulate
emotional responses (e.g., reappraisal or redirected attention) and that rely on explicit
manipulations. In contrast, Etkin and colleagues examine implicit, automatic regula-
tion. Related to this fundamental difference, previous investigations focused on expe-
rienced emotion, whereas the present study examines cognitive response (selective
attention) to conflict between emotional stimuli. In other words, whereas prior work
examined “emotion regulation” per se, Etkin and colleagues probe “conflict regulation”
and, by using emotional stimuli, challenge neural circuits of emotion appraisal. Wheth-
er and how such automatic emotion-related responses contribute to daily emotional
experiences remain unknown. However, the finding that individuals with generalized
anxiety disorder are impaired on this function suggests that this deficit may contribute
to affective dysregulation. In the same vein, the possibility that patients with general-
ized anxiety disorder could also be impaired on a purely cognitive conflict adaptation
paradigm might mitigate the specific relevance of the present findings to emotion pro-
cessing. Nevertheless, this would not diminish the potential treatment implications.
Indeed, the “conflict adaptation” deficit might very well explain the inability of general-
ized anxiety disorder patients to dampen feelings of anxiety. As a result, strengthening
the underlying neural network of “conflict adaptation” via relevant behavioral training
could remedy this deficit (similar to the novel Attention Bias Modification Treatment
developed for other anxiety disorders [7]).
Finally, the premise of this study does raise one important caveat. The claim that the
task probes “emotion regulation” rests on the emotional nature of the stimuli and the
neural responses involved (e.g., pregenual anterior cingulate cortex is typically recruit-
ed in affective processes, whereas the dorsolateral prefrontal cortex is recruited in stud-
ies of cognitive conflict adaptation [2]). However, no evidence exists that the conflict
adaptation effect exhibited in the present study modulates emotional experience. This
type of inferences from specific patterns of neural activation to behavior is common in
the interpretation of neuroimaging studies. Because they can alternately foster novel
theories or stimulate invalid lines of research, such inferences can have tremendous
consequences.
Despite this caveat, the study by Etkin and colleagues in this issue of the Journal has
the potential to ignite a novel, influential, and highly promising line of research that
ingeniously applies a well-described purely cognitive phenomenon to mechanisms
pertaining to emotion processing. This line of research will be particularly relevant to
automatic mechanisms underlying biased processing of affectively laden stimuli that
contribute to anxiety disorders. The study provides insights that may eventually trans-
late into innovative therapies for anxiety disorders.

R e fe re n c e s
1. Etkin A, Prater KE, Hoeft F, Menon V, Schatzberg AF: Failure of anterior cingulate activation and connectivity
with the amygdala during implicit regulation of emotional processing in generalized anxiety disorder. Am J
Psychiatry 2010; 167:545–554
2. Botvinick MM, Braver TS, Barch D M, Carter CS, Cohen JD : Conflict monitoring and cognitive control. Psychol
Rev 2001; 108:624–652
3. Gratton G , Coles MG , D onchin E: O ptimizing the use of information: strategic control of activation of re-
sponses. J Exp Psychol Gen 1992; 121:480–506
4. Egner T, Etkin A, Gale S, Hirsch J: D issociable neural systems resolve conflict from emotional versus nonemo-
tional distracters. Cereb Cortex 2008; 18:1475–1484
5. Kerns JG , Cohen JD, MacD onald AW III, Cho RY, Stenger VA, Carter CS: Anterior cingulate conflict monitoring
and adjustments in control. Science 2004; 303:1023–1026

Am J Psychiatry 167:5, May 2010 [Link] 491

6. O chsner KN, Gross JJ: The cognitive control of emotion. Trends Cogn Sci 2005; 9:242–249
7. See J, MacLeod C , Bridle RJ: The reduction of anxiety vulnerability through the modification of attentional
bias: a real-world study using a home-based cognitive bias modification procedure. J Abnorm Psychol 2009;
118:65–75

M O N IQ U E E R N S T, M .D., P h.D.

Address correspondence and reprint requests to Dr. Ernst, Head, Neurodevelopment of Reward Systems, Emo-
tional Development and Affective Neuroscience Branch (EDAN), NIMH Mood and Anxiety Disorders Program,
National Institutes of Health, 15K North Dr., Bethesda, MD 20892; ernstm@[Link] (email). Editorial
accepted for publication February 2010 (doi: 10.1176/[Link].2010.10010234).

The author reports no financial relationships with commercial interests.

Supported by the NIMH Intramural Program.

492 [Link] Am J Psychiatry 167:5, May 2010