Effectiveness and Mediating Mechanisms of Acceptance
and Commitment and Cognitive Behavioral Therapies in
the Treatment of Mixed Depression and Anxiety
Evan M
M. Forman
Forman, James D
D. Herbert
Herbert, Peter D
D. Yeomans
Yeomans, Katherine McGrath
McGrath, Ethan Moitra
Moitra, Pamela A
A. Geller
Drexel University
Results
Introduction
Comparative Effectiveness: ANCOVA (continued)
90
Effectiveness of ACT. Acceptance and Commitment
Therapy (ACT), as one of the newer, third-wave variations
of cognitive-behavioral therapy (CBT), is particularly
concerned with discouraging experiential avoidance,
decreasing attempts to control internal experiences, and
moving in the direction of valued goals. To date, only a
handful of randomized controlled trials have been
conducted evaluating the effectiveness of ACT for the
treatment of mood and anxiety disorders, and even fewer
directly comparing ACT to CBT. However, findings to date
have strongly supported the effectiveness of ACT.
Mechanisms of Action. Surprisingly little is known about
the mechanisms by which CBT and ACT exert their
positive effects.
Of note, the existing research on
mechanisms of CBT effect has often failed to find support
for the theoretically proposed mediators of this therapy,
such as reducing dysfunctional attitudes. Some findings
point to increased acceptance of undesirable thoughts
and feelings as a specific mechanism of action for ACT.
Some aspects of mindfulness (awareness, acceptance,
ability to describe experiences), an explicit component of
ACT, overlap with the constructs of defusion and metacognitive awareness which may mediate the effects of
both CBT and ACT.
OQ
General Effectiveness: pre-post Cohen’s d
1.6
85
CBT
ACT
1.4
80
12
1.2
1
d
0.8
0.6
0.4
75
CBT
70
ACT
65
0.2
60
0
BDI
BAI
OQ
Both treatments showed large decreases in
symptoms levels between baseline and the 3
month follow-up. For this analysis, and comparative
effectiveness analyses below, only patients with
moderate symptoms at baseline were included (ns =
38, 20, 33).
Long-term Effectiveness: pre-6 mo post Cohen’s d
1
Baseline
3 mo
ANCOVA’s were utilized to examine the effect of
y p
with baseline symptoms
y p
condition on Time 2 symptoms
as a covariate. In the case of all three outcomes, the
effect of treatment condition was not significant (BDI:
F(1, 38) = 1.74, p=.20, partial η2 = .05); BAI: F(1, 26) =
0.00, p=.96, partial η2 = .00; OQ: F(1, 19) = .12, p=.73
partial η2 = .01). Thus, data indicate that CBT and
ACT were equally effective.
CBT
ACT
0.8
(Proxy) Mediational Analyses
0.6
d
Current Study. The current study randomly assigns a
mixed client population to either ACT or CBT. The aim of
the current study is to investigate the effectiveness of ACT
in terms of general outcome improvement and the relative
effectiveness of ACT versus CBT. Additionally, it seeks to
test whether dysfunctional attitudes, experiential
avoidance and mindfulness variables mediate outcome
differentially for the two treatments.
Percent change in mediating variables
0.4
40
0.2
0
BDI
BAI
OQ
Both treatments showed moderate-large decreases
in symptoms levels between baseline and 6-month
post-termination follow-up, with indications of longterm advantage of ACT for BAI and OQ (n = 15).
Method
Outcomes were assessed at baseline,, at a 3-month
follow-up point, and, for some patients (n = 15), 6 months
post-termination. Outcome variables were depression
(BDI-2), anxiety (BAI) and social/vocational functioning
difficulties (OQ-45). Also assessed were potential
mediators, i.e. dysfunctional attitudes (ATQ), experiential
avoidance (AAQ), and mindfulness (KIMS; subscales:
observing, act with awareness, describing, acceptance).
Allegiance. Study therapists are graduate student
trainees who have received specialized training and
supervision in both ACT and CBT. Allegiance effects
were minimized by including investigators with
backgrounds in both CBT and ACT, and through
attempting to ensure equality of training time and
emphasis across the two treatments. An anonymous
measure of therapist allegiance indicated therapists
had no clear allegiance to either of the therapies. If
forced to pick an allegiance, approximately half chose
each treatment (55% CBT, 45% ACT).
ACT
20
10
0
-10
ATQ f* ATQ-b*
ATQ-f*
ATQ b*
Comparative Effectiveness: ANCOVA
This presentation reports findings from an ongoing
randomized clinical trial in which patients with mixed
anxiety and depression are assigned to either ACT or
CBT. Patients (n = 46) are health professional students
seeking treatment at a university counseling center who
vary by gender (79% female), ethnicity (7% African
American, 12% Asian American, 72% White), age (M =
27.8, SD = 7.1), and primary diagnosis (36% depressive
disorder, 41% anxiety disorder, 7% eating disorder).
CBT
30
25
AAQ*
Treatment
Condition
BDI
Obs
Desc
Mediator
Aware Accept
Outcome
20
Correlations between mediators (residualized) and
outcomes (residualized)
15
CBT
ACT
10
0.8
BDI
0.6
5
BAI
OQ
0.4
0
Baseline
25
3 mo
0
BAI
-0.2
20
ATQ-f
15
10
0.2
ATQ-b
AAQ
Obs*
Desc* Aware* Accept*
Proxy mediational analyses suggest that (frequency
of and belief in) dysfunctional thoughts may mediate
outcome for both ACT and CBT, while certain
dimensions of mindfulness (describing, acting with
awareness and acceptance) mediate outcome only
for ACT.
CBT
ACT
5
0
Baseline
3 mo
*Valence reversed in figures above where marked by asterisk.
Corresponding author: Evan Forman, Department of Psychology, 245 N. 15th Street, MS 515, Philadelphia PA 19102; 215-762-4021; evan.forman@drexel.edu.