Yohimbine Enhancement of Exposure Therapy
for Social Anxiety Disorder: A Randomized
Jasper A.J. Smits, David Rosenfield, Michelle L. Davis, Kristin Julian, Pamela R. Handelsman,
Michael W. Otto, Peter Tuerk, Michael Shiekh, Ben Rosenfield, Stefan G. Hofmann, and
Mark B. Powers
Background: Preclinical and clinical trials suggest that yohimbine may augment extinction learning without significant side effects.
However, previous clinical trials have only examined adults with specific phobias. Yohimbine has not yet been investigated in the
augmentation of exposure therapy for other anxiety disorders.
Methods: Adults (n ¼ 40) with a DSM-IV diagnosis of social anxiety disorder were randomized to placebo or yohimbine HCl (10.8 mg) 1
hour before each of four exposure sessions. Outcome measures were collected at baseline, each treatment session, posttreatment, and
Results: Yohimbine was well tolerated. Yohimbine augmentation, relative to placebo augmentation, resulted in faster improvement and
better outcomes on self-report measures of social anxiety disorder severity (Liebowitz Social Anxiety Scale, d ¼ .53) and depressed mood
severity (Beck Depression Inventory, d ¼ .37) but not on the clinician-rated measures (Clinical Global Impressions-Severity Scale, d ¼ .09;
Clinical Global Impressions-Improvement Scale, d ¼ .25). Between-group differences on the Liebowitz Social Anxiety Scale were
moderated by the level of fear reported at the end of an exposure exercise (end fear), such that the advantage of yohimbine over
placebo was only evident among patients who reported low end fear.
Conclusions: The results provide moderate support for yohimbine as a therapeutic augmentation strategy for exposure therapy in social
anxiety disorder, one that may be especially effective when coupled with successful exposure experiences. Beneficial effects for
yohimbine were readily evident for self-report measures but not for clinician-rated outcomes of social anxiety severity and improvement.
Key Words: Cognitive behavioral therapy, cognitive enhancer,
exposure therapy, social anxiety disorder, social phobia, yohimbine
translational research is the investigation of the efficacy of
D-cycloserine (DCS), a partial agonist of the N-methyl-D-aspartate
receptor, for enhancing exposure therapy outcomes (2). Specifi-
cally, after demonstrating that fear extinction is N-methyl-
D-aspartate receptor dependent (3), Walker et al. (4) showed that
DCS enhanced the retention of fear extinction. Importantly, these
preclinical findings on fear extinction have been translated to the
clinical paradigm of exposure therapy, with findings accumulating
to suggest that DCS can improve its outcomes (2,5–9).
Among other compounds, yohimbine has also shown to
facilitate fear extinction in rodents and thus emerges as another
nimal research on fear extinction has proven to be a fruitful
model for the development of augmentation strategies for
exposure therapy (1). A good example of this type of
potential augmentation strategy for exposure therapy (10). An
α2-adrenoceptor antagonist, yohimbine increases central nora-
drenergic activity (11), which has been implicated in the con-
solidation of emotional memories (12–14). The increase of
extracellular levels of norepinephrine in regions critical to fear
extinction (e.g., amygdala, hippocampus, prefrontal cortex) has
therefore been put forth as one plausible mechanism underlying
the effects of yohimbine on fear extinction retention (10,15). In
animals, the systemic administration of yohimbine has been
shown to facilitate fear extinction retention in several (16–18)
but not all studies (19). In accounting for the differences in results
across studies, Holmes and Quirk (10) point to the fact that these
studies employed different strains of rodents, which may imply
that yohimbine enhancement of fear extinction is strain-specific.
Yohimbine enhancement of exposure therapy has been tested
in two published studies that employed samples of adults with
specific phobias. In the first study, Powers et al. (20) enrolled 24
students with claustrophobic fears into a one-session in vivo
exposure protocol and randomly assigned them to either pill
placebo or 10.8 mg of yohimbine hydrochloride (HCl) 1 hour
before the session. Consistent with the hypothesis, participants
who received yohimbine HCl evidenced greater outcomes relative
to those who received placebo. They did not observe between-
group differences in peak fear ratings during exposure, adverse
events, or participants’ guesses of pill content (yohimbine or
placebo), thus challenging the explanation that yohimbine
augmentation is mediated by increased somatic arousal (15).
In the second study, Meyerbroeker et al. (21) enrolled a flying
phobia sample (n ¼ 45) into a multisession protocol. Patients
received four 1-hour sessions of virtual reality exposure therapy
combined with either placebo or yohimbine HCl (10 mg),
From the Department of Psychology and Institute for Mental Health
Research (JAJS, MLD, MBP), The University of Texas at Austin, Austin;
and Department of Psychology (DR, KJ, PRH, MS), Southern Methodist
University, Dallas, TX; Department of Psychology (MWO, SGH), Boston
University, Boston, MA; Department of Psychiatry (PT), Medical University
of South Carolina, Charleston, SC; and Department of Mathematics (BR),
University of Minnesota, Minneapolis, MN.
Address correspondence to Jasper A.J. Smits, Ph.D., The University of
Texas at Austin, Department of Psychology and Institute for Mental
Health Research, 108 E Dean Keeton Stop A8000, Austin, TX 78712;
Received Jun 5, 2013; revised Sep 16, 2013; accepted Oct 3, 2013.
BIOL PSYCHIATRY 2014;75:840–846
& 2014 Society of Biological Psychiatry
administered 1 hour before each session. Analyses of salivary
α-amylase samples confirmed that yohimbine administration
resulted in increased noradrenergic activity. However, the treat-
ment conditions did not differ with respect to baseline to
posttreatment improvements on clinical outcomes. As discussed
by the authors, one possible explanation for the null findings is a
floor effect (21). This hypothesis is in line with the preclinical
findings (16), indicating that yohimbine facilitation of fear
extinction was evident in animals that received partial extinction
training (i.e., 5 extinction trials; 20 minutes of exposure) but not
evident in animals that received full extinction training (i.e., ?10
extinction trials; 40 minutes of exposure).
Growing evidence from investigations of DCS efficacy for
facilitating exposure therapy outcomes indicates the cognitive-
enhancing effects of agents such as yohimbine may also depend
on the success of exposure therapy sessions. Specifically, in
reanalyzing the data from two published studies (9,22), we found
that the advantage of DCS over placebo augmented exposure
therapy was significantly greater when patients reported low fear
at the end of exposure than when they reported moderate to
high end fear, regardless of pre-exposure fear levels. Interestingly,
insufficient exposure training may also account for the failure of
yohimbine to augment extinction for drug cues (23). As reviewed
by Holmes and Quirk (10), yohimbine can impair extinction of
cocaine conditioned place preference, through a mechanism
other than α2-adrenoceptor activity (23). Similarly, DCS has been
shown to fail to enhance extinction to drug cues (24,25), perhaps
indicating that the learning trials used for the extinction of
appetitive versus fear cues are insufficient, so that positive
associations are retained and strengthened by these agents.
Indeed, a recent report indicates that yohimbine, when adminis-
tered to healthy subjects before fear conditioning, can strengthen
fear associations (26). Accordingly, it is possible that the end point
success of exposure sessions moderates the efficacy of yohimbine
augmentation such that the advantage of yohimbine over
placebo is evident when combined with sessions ending with
low fear but not evident or impaired when combined with
sessions ending with higher fear.
The present study sought to build upon the aforementioned
research by 1) examining yohimbine enhancement of a subopti-
mal dose of exposure therapy in adults suffering from social
anxiety disorder; and 2) testing whether yohimbine enhancement
depends on exposure success. The five-session exposure para-
digm employed in the current study matched that of two studies
documenting the efficacy of DCS for facilitating exposure therapy
outcomes in social anxiety disorder (6,27). Participants were
randomized to receive either 10.8 mg of yohimbine or matching
pill placebo 1 hour before sessions 2 to 5. We hypothesized that
1) participants receiving yohimbine would evidence better out-
comes relative to participants receiving placebo; and 2) end fear
would moderate the effects of yohimbine augmentation on
Methods and Materials
Participants (n = 40), 65% male subjects, were recruited from
the Dallas community from 2009 through 2012 (Table 1). Eligi-
bility required a DSM-IV diagnosis of social anxiety disorder and
substantial fear of public speaking [Clinical Global Impressions
Scale (28) for severity of public speaking anxiety $3]. Because
yohimbine administration in larger doses (?20 mg) has been
associated with increases in anxiety in vulnerable populations,
such as those with panic disorder (29,30) and posttraumatic stress
disorder (31,32), we excluded participants with these two con-
ditions and also screened participants for reactivity to the 10.8 mg
dose used in the current study. Other psychiatric exclusion criteria
included 1) organic brain syndrome, mental retardation, or
significant cognitive dysfunction; 2) obsessive-compulsive disor-
der, bipolar disorder, or psychotic or delusional disorders; and
3) eating disorder, any substance or alcohol abuse or dependence
(other than nicotine), or suicidal ideation or suicidal behaviors
within 6 months of study entry. Individuals were excluded if
receiving concurrent psychotherapy directed toward the treat-
ment of social anxiety disorder in the last 3 months or reporting
prior nonresponse to adequately delivered exposure therapy.
Stable (8 weeks before randomization) antidepressant or anxio-
lytic medication (with the exception of monoamine inhibitors or
tricyclic antidepressants) use was accepted. Other exclusion
criteria included significant general medical conditions; untreated
resting blood pressure of $160 systolic and/or 100 diastolic;
history of seizures; and women who were pregnant, lactating, or
planning to become pregnant.
The Southern Methodist University Institutional Review Board
approved the study protocol, and participants provided written
Screening. Interested individuals completed an online pre-
screen survey. Those appearing to meet criteria were invited to
the clinic for a standardized clinical interview [Structured Clinical
Interview for DSM-IV Axis I Disorders, Patient Edition (33)].
Participants appearing eligible after the diagnostic interview
attended a medical evaluation with the study physician involving
a physical examination, pregnancy test, and review of medical
history. At this medical evaluation, all participants were given a
test dose of yohimbine HCl (10.8 mg) to assess for tolerability.
Treatment. One week after baseline, participants began group
sessions of a previously validated 5-week exposure therapy proto-
col (6,27). In the first treatment session (60 minutes), participants
received psychoeducation about the nature of social anxiety
disorder and the treatment rationale, emphasizing the importance
of repeated and prolonged confrontation to the feared stimulus as
an effective method for overcoming anxiety disorders. In sessions 2
through 5 (90 minutes each), participants completed exposure
exercises with the aim to achieve fear extinction. The exposure
exercises involved giving a speech about challenging topics in front
of the therapists, other group members, and confederates. All
participants were encouraged to repeat these exercises at home (i.
e., giving speeches in front of a mirror) and expand their social
interaction and performance activities. Therapists were doctoral
students trained and supervised by the first author.
Medication. Yohimbine HCI and matching placebo capsules
were compounded at Abrams Royal Pharmacy in Dallas. The 10.8
mg dose was selected to be consistent with previous trials.
Randomization. Participants were randomized to receive
either yohimbine HCI or placebo pills 1 hour before sessions 2
through 5, using a computer-generated allocation schedule.
Integrity of the blind was assessed at posttreatment by asking
patients to indicate how confident (0–10 scale) they were they
had been assigned to the yohimbine condition.
Structured Clinical Interview for DSM-IV Axis I Disorders,
Patient Edition. The Structured Clinical Interview for DSM-IV Axis
J.A.J. Smits et al.
BIOL PSYCHIATRY 2014;75:840–846
I Disorders, Patient Edition (33) was used to assess diagnosis of
social anxiety disorder and other exclusionary diagnoses. Inter-
views were conducted by trained doctoral students, and diag-
noses were reviewed in weekly supervision meetings with the
first author before participants were enrolled in the study.
Liebowitz Social Anxiety Scale. Participants completed the
Liebowitz Social Anxiety Scale (LSAS) (34) at baseline, weekly
during treatment, at posttreatment (week 6), and at follow-up
(week 9). We elected to use the self-report version of the LSAS,
which has sound psychometric properties, comparable with the
clinician-rated version (35).
Clinical Global Impression Severity and Improvement Scales.
Study clinicians, blind to condition administered the Clinical Global
Impression-Severity (CGI-S) and Clinical Global Impression-Improve-
ment (CGI-I) scales (36) at baseline (CGI-S only), weekly during
treatment, at posttreatment (week 6), and at follow-up (week 9).
Beck Depression Inventory-II. Participants completed the
Beck Depression Inventory-II (BDI-II), an established measure of
depression severity with sound psychometric properties (37), at
baseline, weekly during treatment, at posttreatment (week 6), and
at follow-up (week 9).
Fear Ratings. Clinicians introduced patients to a Subjective
Units of Distress Scale (SUDS) (0–100 units) (38) during session 1
and provided standard anchor points (0 ¼ no anxiety, 50 ¼
moderate anxiety, 100 ¼ extreme anxiety). Participants provided
SUDS ratings before beginning each exposure and immediately
before the end of the exposure (end fear). They provided their
peak fear rating following the completion of the exercise.
Adverse Events. This study employed procedures for adverse
events and safety assessments identical to those of previous
studies conducted by our group (6,7,20,39,40). Specifically, vital
signs were taken before each dose administration, after which
Table 1. Sample Characteristics and Clinical Severity Measures at Baseline
Yohimbine-Augmented CBT (n ¼ 20)
Placebo-Augmented CBT (n ¼ 20)
Hispanic or Latino Ethnicity
Living with a partner
Separated, widowed, or divorced
High school graduate
Partial high school
Missing or not given
Student or dependent on spouse
BDI-II, Beck Depression Inventory-II; CBT, cognitive behavioral therapy; CGI-S, Clinical Global Impression-Severity;
LSAS-SR, Liebowitz Social Anxiety Scale-Self-Report.
842 BIOL PSYCHIATRY 2014;75:840–846
J.A.J. Smits et al.
trained staff asked participants whether they experienced any
symptoms since the last dose administration. All new complaints
and symptoms were recorded on the adverse events form, as
were preexisting complaints or symptoms that increased in
intensity or frequency since enrollment.
We employed multilevel modeling (MLM) to test hypotheses,
using restricted maximum likelihood estimation, because it
includes all available data from all participants, is an intention-
to-treat analysis, and it produces unbiased estimates of regression
coefficients and variances in samples as small as 30 participants
(41). The repeated assessments of the outcomes (LSAS, CGI-S, CGI-I,
and BDI-II) were nested within subjects. We initially modeled time
as quadratic but dropped the quadratic terms since they were
found to be nonsignificant (ps ? .536). We also included various
demographic variables as covariates in the initial analyses but
found none of them to be significantly related to outcome, so
they too were dropped from the analyses. Initial severity on the
outcome, along with the interaction of severity with time, were
included as covariates in all analyses (all ps ? .01).
We conducted two tests to determine whether treatment
condition (yohimbine vs. placebo) enhanced the effect of exposure
therapy on outcomes (hypothesis 1). The first was the test of the
treatment condition ? time interaction, which examined whether
the yohimbine condition evidenced steeper slopes of improvement
relative to the placebo condition. Second, we investigated treatment
condition differences at the last assessment by centering time at the
final assessment, which tested whether estimated scores at the final
assessment (1-month follow-up) differed by treatment condition.
To test whether end fear at the conclusion of one session
moderated the effect of condition (yohimbine vs. placebo) on
outcome at the next session (hypothesis 2), we examined a second
MLM model, which included the following predictors of outcome at
the next session: end fear, treatment condition, and end fear ?
condition. Because end fear was not randomized, we included
several variables to control for potential third variables (42). Based
on previous work (43), we included sex, education, race, cohabiting
versus not, baseline severity on the outcome, severity rating at the
current session, and the interactions of severity with treatment
condition. This analysis included only the sessions in which
yohimbine (or placebo) was administered (end fear at sessions 2,
3, 4, 5 predicting outcome at sessions 3, 4, 5, 6, respectively).
To correct for the possibility of inflation of the experiment-
wise type I error rate, we first performed each analysis as a
multivariate MLM, including the four dependent variables as a
multivariate outcome. We further examined the significant results
from the multivariate MLM using univariate analyses of the
individual dependent variables. To correct for possible inflation
of the false discovery rate due to the multiple univariate tests, we
used the Benjamini-Hochberg-Yekutieli procedure (44) to correct
the p values of the univariate tests. All p values reported for the
univariate tests are the corrected values.
There were no significant between-group differences on any
of the demographics or clinical severity measures assessed at
baseline (ps ? .320; Table 1). One participant in the yohimbine
condition was taking a stable dose of zolpidem for sleep
throughout the study period.
Adverse Events During Test Dose at Screen
The most commonly reported adverse events (each reported
by three participants; 6% of the sample) were jitteriness, anxiety,
hot flashes, chills, and nausea. These adverse events were all
reported to be mild in intensity, with the exception of jitteriness,
anxiety, and chills, which were each reported as being of
moderate intensity by at least one participant. Tiredness, light-
headedness, sweating, alertness, feeling dazed, heart palpitations,
head pressure, insomnia, hunger, open sinuses, headache, and
pressure around the eyes were reported by one participant each
(2% of the sample). All of these events were reported to be mild
in intensity, with the exception of one report of sweating, one
report of feeling dazed, one report of heart racing and feeling
jittery, and one report of insomnia; these were reported as being
moderate in intensity. None of these reported adverse events
were unexpected, and none were serious adverse events. Of the
53 patients tested, 1 participant (2%) decided against random-
ization because of adverse events.
Adverse Events During Intervention
Throughout the treatment phase of the study, two adverse
events were reported among individuals in the yohimbine
condition. One participant endorsed trouble sleeping that was
mild in intensity, and another participant endorsed crying that
was moderate in intensity. Participants in the placebo condition
reported sweaty palms (one report of mild intensity), headache
(three reports, with one of moderate intensity and two of mild
intensity), and a decrease in sexual arousal (one report of mild
intensity). No unexpected adverse events or serious adverse
events were reported.
Attrition and Treatment Integrity
The attrition rate during the 5-week treatment phase was 10%
in the yohimbine condition and 15% in the placebo condition
(p ¼ .633). The mean number of the four exposure sessions
attended was 3.70 (SD ¼ .57) in the yohimbine condition and 3.75
(SD ¼ .44) in the placebo condition (t38¼ .31, p ? .7).
Session 2 Session 3 Session 4 Session 5 Post Follow-Up
Predicted LSAS Score
Figure 1. The efficacy of yohimbine (YOH) for enhancing the effects of
exposure therapy on social anxiety disorder symptom severity. LSAS,
Liebowitz Social Anxiety Scale; PBO, placebo.
J.A.J. Smits et al.
BIOL PSYCHIATRY 2014;75:840–846
An analysis of variance comparing the two treatment conditions
on participants’ confidence that they had been assigned to the
yohimbine condition showed no significant between-group differ-
ences (Myohimbine¼ 4.50 vs. Mplacebo¼ 3.25; F1,32¼ 2.34, p ¼ .136).
The Effect of Yohimbine on Fear Activation during Treatment
To verify that the arousal created by yohimbine did not impact
fear activation, we performed MLM analyses with treatment
condition, time, and treatment ? time as predictors of peak
SUDS ratings or initial SUDS ratings. We observed no between-
group differences (ps ? .537).
Multivariate analyses showed that the slope of improvement
over time was steeper for participants in the yohimbine condition
than for those in the placebo condition (b ¼ .06, t208¼ 1.98, p ¼
.049). Univariate analyses indicated that this faster improvement
for yohimbine participants was evident for the LSAS (b ¼ 1.93,
t66¼ 2.32, p ¼ .046; Figure 1) and the BDI-II (b ¼ .65, t79¼ 2.39,
p ¼ .046; Figure 2) but not for the CGI-S (b ¼ ?.02, t45¼ ?.26,
p ¼ .793) or the CGI-I (b ¼ .03, t42¼ .56, p ¼ .770). Similarly, the
multivariate MLM also showed that participants in the yohimbine
condition reported lower scores at follow-up than those in the
placebo condition (b ¼ .23, t213¼ 2.29, p ¼ .023). Univariate
analyses indicated that this difference was significant on the LSAS
(M ¼ 31.6 vs. 42.1, b ¼ 10.46, t55¼ 2.50, p ¼ .046, d ¼ .53) and on
the BDI-II (M ¼ 3.65 vs. 6.59, b ¼ 2.94, t60¼ 2.33, p ¼ .046, d ¼
.37) but not on the CGI-S (M ¼ 2.93 vs. 2.81, b ¼ .12, t41¼ .36, p ¼
.721, d ¼ .09) or the CGI-I (M ¼ 1.87 vs. 2.08, b ¼ .21, t35¼ .77,
p ¼ .597, d ¼ .25).
The multivariate analysis revealed a significant end fear ?
treatment condition interaction predicting outcome at the next
session (b ¼ ?.01, t268¼ 1.99, p ¼ .048). Univariate analyses
indicated a significant end fear ? treatment condition interaction
predicting LSAS at the next session (b ¼ .21, t168¼ 2.56, p ¼ .049)
but not BDI-II, CGI-S, or CGI-I scores at the next session (ps ?
.338). To elucidate the form of the significant interaction, we
utilized the approach developed by Aiken and West (45), which
involves calculating model-based predicted differences between
yohimbine and placebo for participants who have different levels
of end fear (Figure 3). For participants with end fear that was 1 SD
below the mean (M ¼ 39.1; SD ¼ 17.7), those receiving yohimbine
were 4.47 points more improved on the LSAS at the next session
than those receiving placebo (b ¼ 4.47, p ¼ .038). Conversely, for
those with end fear 1 SD above the mean (end fear ¼ 56.8),
participants receiving yohimbine were 2.95 points less improved
on the LSAS than those given placebo, although this difference
was not statistically significant (b ¼ 2.95, p ¼ .127). Similarly, this
analysis also showed that end fear was significantly related to
LSAS at the next session for those given yohimbine (b ¼ ?.26,
p ? .001) but not for those given placebo (b ¼ .05, p ¼ .343).
The present study showed that yohimbine augmentation
resulted in faster improvement and better outcome on self-
report measures of social anxiety disorder severity and depressed
mood severity relative to placebo augmentation. However, we did
not find significant differences on clinician-rated assessments of
social anxiety symptoms. Interestingly, the same pattern of
differential outcomes on self-reported and clinician-rated instru-
ments was evident in the pilot study of DCS augmentation of
exposure in social anxiety disorder by Hofmann et al. (6). In that
study, the effect size for the self-report measure at 1-month
follow-up (Social Phobia and Anxiety Inventory; d ¼ 1.43) was
roughly double that for the clinician-rated measures (CGI-S and
clinican-rated version of the LSAS: d ¼ .73 and d ¼ .69, res-
pectively). Similarly, in the current study, the effect size for the
advantage of yohimbine over placebo for the patient-rated LSAS
(d ¼ .53) was roughly double that for the clinician-rated CGI-I
(d ¼ .25), although the effect size for the CGI-S was much smaller
(d ¼ .09). It is unclear why the self-report measures in these
augmentation trials would be more sensitive to differential
Session 2 Session 3 Session 4 Session 5 Post Follow-Up
Predicted BDI Score
Figure 2. The efficacy of yohimbine (YOH) for enhancing the effects of
exposure therapy on depressive symptom severity. BDI, Beck Depression
Inventory; PBO, placebo.
End Fear = M-1SD End Fear = M End Fear = M+1SD
Predicted LSAS Score at the Next Session
Figure 3. Session-end fear moderates the efficacy of yohimbine (YOH) for
enhancing the effects of exposure therapy on social anxiety disorder
symptom severity. LSAS, Liebowitz Social Anxiety Scale; PBO, placebo.
844 BIOL PSYCHIATRY 2014;75:840–846
J.A.J. Smits et al.
change than the clinician measures. Indeed, a meta-analysis of
trials of selective serotonin reuptake inhibitors for social anxiety
disorder showed roughly equivalent effect sizes for the CGI-I and
LSAS (46). However, it should be noted that our augmentation
study included only four exposure sessions. Thus, it is possible
that patients self-report fewer symptoms on paper after this
abbreviated treatment, while being unable to fully describe the
relative improvement from baseline to a clinician. A full course of
augmented exposure therapy would help determine if clinician-
rated scores eventually mirror self-report data.
In published studies to date, yohimbine augmentation has
significantly enhanced outcomes for exposure interventions in
two of three studies. In the current study, we replicated the
positive findings reported by Powers et al. (20) for augmentation
of single-session treatment for claustrophobia. Null results were
reported by Meyerbroeker et al. (21) for a four-session fear of
flying intervention. It may be that yohimbine augmentation
effects parallel those for DCS, with much smaller effects evident
as more complete exposure-based cognitive behavioral therapy
(CBT) is offered. For example, Hofmann et al. (6,7) reported large
effect sizes for DCS augmentation of 5 sessions of CBT for social
anxiety disorder but much smaller effects with 12 sessions of CBT.
Our study also replicates both published studies of yohimbine
augmentation of exposure therapy in finding that low doses of
yohimbine HCl (ten 10.8 mg in these studies) are not associated
with adverse events (20,21). These findings are inconsistent with
studies documenting increased nervousness among healthy
control subjects after administration of intravenous yohimbine
(47) or 20 mg of oral yohimbine (48) but consistent with other
reports on the safety of acute and chronic administration of lower
doses of oral yohimbine in healthy individuals (11).
In terms of mechanism of action, our study was designed to
minimize the influence of enhanced autonomic symptoms and
anxiety on exposure outcome. To minimize the potential influence
of these effects, we utilized a relatively low dose of yohimbine,
selected a cohort of patients that were less likely to experience
anxiety or panic in response to yohimbine administration, com-
pleted a test dose of yohimbine as part of the selection criteria, and
then assessed whether participants who received yohimbine
experienced more fear during exposure than those undergoing
placebo augmentation. With these selection criteria, we found no
evidence of differential fear during exposure; hence, the observed
significant differences in outcome for yohimbine versus placebo
augmentation are unlikely due to differential effects of anxious
arousal, a finding consistent with similar analyses by Powers et al.
(20). Moreover, participants were unable to guess whether they had
taken an active agent. Together, these lines of evidence help
discount the differential experience of autonomic arousal as a
mechanism for yohimbine augmentation effects.
The current study also documented apparent extinction-
retention effects of yohimbine augmentation by showing a link
between a session index of fear reduction and the level of
symptoms 1 week later. Specifically, greater relative benefit of
yohimbine augmentation was evident for those patients completing
their exposure session with low fear, replicating similar findings
reported for DCS (9,22). The advantage of yohimbine was lost for
those patients who completed exposure with higher fear ratings,
again replicating similar findings for DCS (9,22). This relationship is
consistent with the hypothesized role of yohimbine in aiding the
consolidation of extinction learning from one session to the next
(10), helping patients retain end-of-session evaluations of the feared
stimulus. As noted, consolidation effects for yohimbine or DCS
should be less apparent in those protocols offering a greater
number of exposure sessions (7); that is, with greater opportunities
for learning and consolidation, the relative advantage of memory
enhancement should be attenuated. As such, these augmentation
strategies might have their strongest role in increasing the efficiency
of treatment and helping those individuals who otherwise have
difficulty with extinction; although this latter role, with treatment-
resistant anxiety patients, has yet to be systematically tested.
The primary limitation of this study is the relatively small sample
size comprised mostly of male subjects. Future work would also be
strengthened by utilizing a longer follow-up and including domain-
specific clinician-rated measures that are perhaps more sensitive to
detecting augmentation effects than the CGI-I. Finally, the present
research provides only limited insight into the mechanisms by
which yohimbine facilitated exposure therapy outcomes. We did
not assess catecholamine levels and therefore cannot test the
hypothesis that the observed main effects and moderator effects
were mediated by increased noradrenergic activity. Importantly, as
mentioned previously, accumulating findings from the animal
literature indicate that yohimbine may exert its effects on fear
extinction via mechanisms independent of the α2-adrenoceptor
(10) and that the effects of yohimbine on extinction learning may
be particularly potent for extinction resistant subjects (10). The
examination of these alternative mediator and moderator hypoth-
eses awaits testing in future research.
Overall, this study provides moderate support for yohimbine
augmentation of exposure therapy for patients with social anxiety
disorder. This study is also noteworthy in drawing attention to the
conditions by which putative memory-enhancing agents, such as
DCS or yohimbine, offer greater benefit in that it joins two others in
showing that the greater relative benefit over placebo augmentation
is achieved for patients who attain low fear levels at the conclusion
of exposure sessions. As we have hypothesized elsewhere, when
memory-consolidating agents are used, longer sessions or the use of
smaller incremental steps along a hierarchy of exposure may be
useful to help maximize the likelihood of low levels of end-of-
exposure fear and, hence, better exposure-augmentation outcomes
(9,22). The study we report on here indicates that yohimbine offers
potential as one such augmentation strategy.
Drs. Smits and Hofmann receive royalties from various book
publishers for work unrelated to this study. Dr. Otto served as a
consultant for MicroTransponder Inc., ProPhase, and Concert Pharma-
ceuticals; received royalties from ProPhase for use of the SIGH-A, and
receives royalties from various book publishers for work unrelated to
this study. Dr. Hofmann receives royalties from various book publishers
for work unrelated to this study. Dr. Rosenfield, Dr. Powers, Dr. Tuerk,
Dr. Shiekh, Ms. Davis, Ms. Julian, Mr. Rosenfield, and Ms. Handelsman
report no biomedical financial interests or potential conflicts of interest.
ClinicalTrials.gov: Yohimbine to Enhance Cognitive Behavioral
Therapy (CBT) for Social Anxiety; http://clinicaltrials.gov/ct2/show/
Supplementary material cited in this article is available online at
1. Anderson KC, Insel TR (2006): The promise of extinction research for the
prevention and treatment of anxiety disorders. Biol Psychiatry 60:319–321.
2. Davis M, Ressler K, Rothbaum BO, Richardson R (2006): Effects of D-
cycloserine on extinction: Translation from preclinical to clinical work.
Biol Psychiatry 60:369–375.
3. Falls WA, Miserendino J, Davis M (1992): Extinction of fear-potentiated
startle: Blockade by infusion of an NMDA antagonist into the
amygdala. J Neurosci 12:854–863.
J.A.J. Smits et al.
BIOL PSYCHIATRY 2014;75:840–846
4. Walker DL, Ressler KJ, Lu KT, Davis M (2002): Facilitation of condi- Download full-text
tioned fear extinction by systemic administration or intra-amygdala
infusions of D-cycloserine as assessed with fear-potentiated startle in
rats. J Neurosci 22:2343–2351.
5. Ressler KJ, Rothbaum BO, Tannenbaum L, Anderson P, Graap K,
Zimand E, et al. (2004): Cognitive enhancers as adjuncts to psycho-
therapy: Use of D-cycloserine in phobic individuals to facilitate
extinction of fear. Arch Gen Psychiatry 61:1136–1144.
6. Hofmann SG, Meuret AE, Smits JA, Simon NM, Pollack MH, Eisenmenger
K, et al. (2006): Augmentation of exposure therapy with D-cycloserine
for social anxiety disorder. Arch Gen Psychiatry 63:298–304.
7. Hofmann SG, Smits JA, Rosenfield D, Simon N, Otto MW, Meuret AE, et al.
(2013): D-cycloserine as an augmentation strategy with cognitive-
behavioral therapy for social anxiety disorder. Am J Psychiatry 170:751–758.
8. Norberg MM, Krystal JH, Tolin DF (2008): A meta-analysis of D-
cycloserine and the facilitation of fear extinction and exposure
therapy. Biol Psychiatry 63:1118–1126.
9. Smits JA, Rosenfield D, Otto MW, Powers MB, Hofmann SG, Telch MJ,
et al. (2013): D-cycloserine enhancement of fear extinction is specific
to successful exposure sessions: Evidence from the treatment of
height phobia. Biol Psychiatry 73:1054–1058.
10. Holmes A, Quirk GJ (2010): Pharmacological facilitation of fear
extinction and the search for adjunct treatments for anxiety disor-
ders–the case of yohimbine. Trends Pharmacol Sci 31:2–7.
11. Tam SW, Worcel M, Wyllie M (2001): Yohimbine: A clinical review.
Pharmacol Ther 91:215–243.
12. McGaugh JL (2000): Memory–a century of consolidation. Science 287:
13. McGaugh JL (2004): The amygdala modulates the consolidation of
memories of emotionally arousing experiences. Annu Rev Neurosci 27:1–28.
14. Kaplan GB, Moore KA (2011): The use of cognitive enhancers in animal
models of fear extinction. Pharmacol Biochem Behav 99:217–228.
15. Janak PH, Corbit LH (2011): Deepened extinction following compound
stimulus presentation: noradrenergic modulation. Learn Mem 18:1–10.
16. Cain CK, Blouin AM, Barad M (2004): Adrenergic transmission facilitates
extinction of conditional fear in mice. Learn Mem 11:179–187.
17. Hefner K, Whittle N, Juhasz J, Norcross M, Karlsson RM, Saksida LM, et al.
(2008): Impaired fear extinction learning and cortico-amygdala circuit
abnormalities in a common genetic mouse strain. J Neurosci 28:8074–8085.
18. Morris RW, Bouton ME (2007): The effect of yohimbine on the extinction
of conditioned fear: a role for context. Behav Neurosci 121:501–514.
19. Mueller D, Olivera-Figueroa LA, Pine DS, Quirk GJ (2009): The effects of
yohimbine and amphetamine on fear expression and extinction in
rats. Psychopharmacology (Berl) 204:599–606.
20. Powers MB, Smits JA, Otto MW, Sanders C, Emmelkamp PM (2009):
Facilitation of fear extinction in phobic participants with a novel
cognitive enhancer: A randomized placebo controlled trial of yohim-
bine augmentation. J Anxiety Disord 23:350–356.
21. Meyerbroeker K, Powers MB, van Stegeren A, Emmelkamp PM (2012):
Does yohimbine hydrochloride facilitate fear extinction in virtual
reality treatment of fear of flying? A randomized placebo-controlled
trial. Psychother Psychosom 81:29–37.
22. Smits JA, Rosenfield D, Otto MW, Marques L, Davis ML, Meuret AE,
et al. (2013): D-cycloserine enhancement of exposure therapy for
social anxiety disorder depends on the success of exposure sessions. J
Psychiatr Res 47:1455–1461.
23. Davis AR, Shields AD, Brigman JL, Norcross M, McElligott ZA, Holmes
A, Winder DG (2008): Yohimbine impairs extinction of cocaine-
conditioned place preference in an alpha2-adrenergic receptor
independent process. Learn Mem 15:667–676.
24. Price KL, Baker NL, McRae-Clark AL, Saladin ME, Desantis SM, Santa
Ana EJ, Brady KT (2013): A randomized, placebo-controlled laboratory
study of the effects of D-cycloserine on craving in cocaine-dependent
individuals. Psychopharmacology (Berl) 226:739–746.
25. Kennedy AP, Gross RE, Whitfield N, Drexler KP, Kilts CD (2012): A
controlled trial of the adjunct use of D-cycloserine to facilitate
cognitive behavioral therapy outcomes in a cocaine-dependent
population. Addict Behav 37:900–907.
26. Soeter M, Kindt M (2012): Stimulation of the noradrenergic system
during memory formation impairs extinction learning but not the
disruption of reconsolidation. Neuropsychopharmacology 37:1204–1215.
27. Guastella AJ, Richardson R, Lovibond PF, Rapee RM, Gaston JE,
Mitchell P, Dadds MR (2008): A randomized controlled trial of the
effect of D-cycloserine on enhancement of exposure therapy for social
anxiety disorder. Biol Psychiatry 63:544–549.
28. National Institute of Mental Health (1970): CGI: Clinical Global
Impressions. In: Guy W, Bonaro R, editors. Manual for the ECDEU
Assessment Battery. Chevy Chase, MD: National Institute of Mental
29. Charney DS, Woods SW, Krystal JH, Nagy LM, Heninger GR (1992):
Noradrenergic neuronal dysregulation in panic disorder: The effects of
intravenous yohimbine and clonidine in panic disorder patients. Acta
Psychiatr Scand 86:273–282.
30. Gurguis GN, Vitton BJ, Uhde TW (1997): Behavioral, sympathetic and
adrenocortical responses to yohimbine in panic disorder patients and
normal controls. Psychiatry Res 71:27–39.
31. Southwick SM, Krystal JH, Bremner JD, Morgan CA 3rd, Nicolaou AL,
Nagy LM, et al. (1997): Noradrenergic and serotonergic function in
posttraumatic stress disorder. Arch Gen Psychiatry 54:749–758.
32. Bremner JD, Innis RB, Ng CK, Staib LH, Salomon RM, Bronen RA, et al.
bral metabolic correlates of yohimbine administration in combat-
related posttraumatic stress disorder. Arch Gen Psychiatry 54:
33. First MB, Spitzer RL, Gibbon M, Williams JBW (2001): Structured Clinical
Interview for DSM-IV Axis-I Disorders, Clinician Version. Arlington, VA:
American Psychiatric Publishing, Inc.
34. Liebowitz MR (1987): Social phobia. Mod Probl Pharmacopsychiatry 22:
35. Fresco DM, Coles ME, Heimberg RG, Liebowitz MR, Hami S, Stein MB,
Goetz D (2001): The Liebowitz Social Anxiety Scale: A comparison of
the psychometric properties of self-report and clinician-administered
formats. Psychol Med 31:1025–1035.
36. Liebowitz MR, Schneier F, Campeas R, Hollander E, Hatterer J, Fyer A,
et al. (1992): Phenelzine vs atenolol in social phobia. A placebo-
controlled comparison. Arch Gen Psychiatry 49:290–300.
37. Beck AT, Steer RA, Ball R, Ranieri WF (1996): Comparison of Beck
Depression Inventories-IA and -II in psychiatric outpatients. J Pers
38. Wolpe J (1958): Psychotherapy by reciprocal inhibition. Pavlov J Res
39. Otto MW, Tolin DF, Simon NM, Pearlson GD, Basden S, Meunier SA,
et al. (2010): Efficacy of d-cycloserine for enhancing response to
cognitive-behavior therapy for panic disorder. Biol Psychiatry 67:
40. Tart CD, Handelsman PR, Deboer LB, Rosenfield D, Pollack MH,
Hofmann SG, et al. (2013): Augmentation of exposure therapy with
post-session administration of D-cycloserine. J Psychiat Res 47:
41. Maas CJ, Hox JJ (2005): Sufficient sample sizes for multilevel modeling.
Eur J Res Meth Behav Soc Sci 1:86–92.
42. Cook TD, Steiner PM (2010): Case matching and the reduction of
selection bias in quasi-experiments: The relative importance of pretest
measures of outcome, of unreliable measurement, and of mode of
data analysis. Psychol Methods 15:56–68.
43. Smits JA, Hofmann SG, Rosenfield D, Deboer LB, Costa PT, Simon NM,
et al. (2013): D-Cycloserine augmentation of cognitive behavioral
group therapy of social anxiety disorder: Prognostic and prescriptive
variables [published online ahead of print August 12]. J Consult Clin
44. Benjamini Y, Yekutieli D (2001): The control of the false discovery rate
in multiple testing under dependency. Ann Stat 29:1165–1188.
45. Aiken LS, West SG (1991): Multiple Regression: Testing and Interpreting
Interactions. Newbury Park, CA: Sage Publications.
46. Hedges DW, Brown BL, Shwalb DA (2009): A direct comparison of
effect sizes from the Clinical Global Impression-Improvement scale to
effect sizes from other rating scales in controlled trials of adult social
anxiety disorder. Hum Psychopharmacol 24:35–40.
47. Vasa RA, Pine DS, Masten CL, Vythilingam M, Collin C, Charney DS,
et al. (2009): Effects of yohimbine and hydrocortisone on panic
symptoms, autonomic responses, and attention to threat in healthy
adults. Psychopharmacology (Berl) 204:445–455.
48. Charney DS, Woods SW, Heninger GR (1989): Noradrenergic function
in generalized anxiety disorder: Effects of yohimbine in healthy
subjects and patients with generalized anxiety disorder. Psychiatry
846 BIOL PSYCHIATRY 2014;75:840–846
J.A.J. Smits et al.