Transcript Please Highlight this Area and Add your Main Title

A Randomized Controlled Trial of Cognitive Behavioral Social Rhythm
Group Therapy for Male Veterans with PTSD, Major Depressive
Disorder, and Sleep Problems
Patricia Haynes,1,2 Monica Kelly,2,1,3 Sairam Parthasarathy,1,2 & Richard Bootzin3
1Southern
Arizona VA Healthcare System, 2College of Medicine, University of Arizona, 3Department of Psychology, University of Arizona
Introduction
Sleep problems are highly prevalent in both
Posttraumatic Stress Disorder (PTSD) and Major
Depressive Disorder (MDD)
•Sleep symptoms may precede and exacerbate
both MDD1 and PTSD2
•PTSD treatments often do not address sleep
symptoms completely3,4
Group therapies for PTSD may not be as efficacious
as individual therapies. A large, multicenter RCT
found that Present Centered Group Therapy (PCGT)
had comparable outcomes to Trauma Focused
Exposure Group Therapy, with less attrition.5
•Cognitive Behavioral Therapy for Insomnia (CBTi)
is efficacious in a group format6
•CBTi has lower attrition rates (0-8%) than
exposure therapies (~25%)
•Adjunctive, group therapies may improve PTSD
exposure therapy outcomes.7
We developed Cognitive Behavioral Social Rhythm
Therapy (CBSRT) as a potential adjunctive, group
therapy to CBT for PTSD for veterans with PTSD and
MDD (a population with worse psychopathology and
treatment outcomes8)
•Open-trial results indicated that veterans receiving
CBSRT have significant improvements in sleep,
PTSD, and depression.9
Hypotheses
The purpose of this study was to test whether group
CBSRT is superior to PCGT for sleep and psychiatric
outcomes in veterans with PTSD, Depression, and
sleep problems.
Methods
Therapies
Summary of Findings
Cognitive Behavioral Social Rhythm
Group Therapy (CBSRT)
Duration
Despite randomization, individuals in CBSRT had
significantly worse baseline symptom severity
measures than individuals in PCGT, complicating
interpretation of findings.
Present Centered Group Therapy
(PCGT)
12 week, once per week
2 hours
12 week, once per week
2 hours
Modality
Group, manualized
Group, manualized
Format
Structured, 1st hour homework review, 2nd hour
new material
Relatively unstructured, guided by group process
needs
Model
Social Rhythm Hypothesis14
CBSRT is designed to improve mood and sleep by
stabilizing social rhythms, increasing exposure to
ambient light, changing dysfunctional bed/bedtime
associations, activating the imagery system by
changing nightmare content, and challenging
dysfunctional automatic thoughts that might
contribute to behavioral inactivation and
nonadherence to the therapy protocol. There is no
discussion of past traumatic events.
Yalom-based Process Group Model15
PCGT includes education about the typical
symptoms and features associated with PTSD and
MDD, with a focus on how these symptoms affect
interpersonal relationships. It uses the group
format to decrease isolation, normalize symptoms,
and provide the experience of giving and receiving
support. Some relaxation training is provided early
in therapy. There is no discussion of past traumatic
events.
To collaboratively increase the frequency and
consistency of daily behaviors and empirically
refute thoughts that may be impede goal
attainment.
To develop positive group cohesion and the
atmosphere of safety and trust.
Therapist
Goals
Results
Sleep outcomes:
•During the treatment and follow-up periods, the
rates of change in sleep efficiency in CBSRT v.
PCGT were significantly different (see graph).
•In both CBSRT and PCGT, there were significant
reductions in sleep onset latency and number of
awakenings over the treatment period; this
improvement slowed in the follow-up period.
•From baseline to 6 Month FU, there were few
qualitative differences in TST or WASO.
Psychiatric outcomes:
•Both groups improved over time on PTSD and
depression outcomes. There were no significant
differences between groups.
•Both groups had >10 point change on CAPS,
indicating clinically significant improvement on
PTSD symptoms. As expected, individuals in
both groups had significant levels of residual
symptoms.
Parameter Estimates (and Standard Errors) for Growth Models Examining the Rate of Change x
Condition in Symptom Measures Over the Course of Treatment and Through 6 Month Follow-Up
SE
Intercept
TIB
B
SE
64.45*** 3.24
B
SE
504.03*** 21.46
Condition
12.01* 4.58
Time
0.73** 0.27
Condition x Time
-0.89* 0.40
PostTx Time
-0.75* 0.36
Condition x PostTx Time 1.18* 0.55
-11.80
-6.36***
3.71
9.36***
-6.77ᵗ
30.29
1.70
2.56
2.32
3.53
Conclusions
TST
WASO
SOL
No. Awakes
B
SE
333.37*** 21.54
B
SE
47.44*** 10.78
B
SE
75.70*** 11.64
B
SE
2.88*** 0.30
44.94
-2.03
-0.89
4.16ᵗ
0.10
30.39
1.64
2.47
2.24
3.41
2.12
-0.49
1.05
0.78
-2.07
15.19
0.76
1.14
1.03
1.58
-41.81*
-1.90**
1.24
2.46*
-1.98
16.37
0.69
1.04
0.94
1.44
-0.10
-0.08**
0.05
0.09**
-0.07
0.42
0.02
0.04
0.03
0.05
A compound symmetry variance matrix was used to model the error variance. PCGT is the reference condition.
ᵗp < .10, *p < .05, **p < .01, ***p < .001
Design
• Double-blind, behavioral RCT comparing CBSRT to
PCGT, an active attention control psychotherapy
CBSRT and PCGT appear to produce comparable
psychiatric and sleep outcomes (SOL and number of
awakenings). As compared to PCGT:
•CBSRT is associated with faster improvements on
sleep efficiency
•CBSRT is associated with fewer therapy drop-outs
•Given the lower attrition rate, CBSRT may be a
valuable, adjunctive group therapy option for
individuals with PTSD and MDD. Research is
necessary to test whether CBSRT improves exposure
therapy outcomes.
•Active mental health treatments (such as PCGT)
may have positive effects on sleep that may rival
behavioral sleep treatments, over time.
Recruitment
• Southern Arizona VA Healthcare System
•Results are consistent with previous studies
indicating that Vietnam Veterans that are VA Users
may have limited treatment responsiveness5,16
Measures
•Diagnosis: Structured Clinical Interview DSM-IV10
•Sleep: Daily Sleep Diary11
•PTSD: Clinician Administered PTSD Scale (CAPS)12
•Depression: Hamilton Depression Rating Scale
(HamD)13
•Secondary and objective sleep outcome analyses
are pending, as are analysis of therapist fidelity,
patient adherence, and adequate dosing issues,
which may affect overall findings
Statistical Analyses
• Mixed modeling with intent-to-treat approach
•Results must be qualified by small sample size / low
power to detect < large effects
Participants
•43 male veterans (21 CBSRT, 22 PCGT)
•Randomization stratified by military era (44%
Vietnam, 26% OEF/OIF, 30% Other)
•No differences in groups on age or ethnicity
•M Age = 48.42 Yrs (SD = 13.51 Yrs)
•56% Caucasian, 23% Hispanic, 7% African
Americian, 7% American Indian, 7% Other
Exclusion criteria
• Age < 18 or > 65
• Shift work, moderate to severe primary sleep
disorders (as determined by PSG), alteration in
medications, current substance abuse (< 30 days),
uncontrolled medical illness, severe traumatic brain
injury/neurological disorder
References
1.
2.
3.
4.
5.
6.
7.
8.
HamD
Intercept
Condition
Time
Condition x Time
B
SE
23.30***
-2.37
0.21ᵗ
0.1
1.23
1.72
0.11
0.17
CAPS
Intercept
Condition
Time
Condition x Time
B
SE
76.65***
-5.25
-0.76*
0.39
4.59
6.39
0.32
0.45
CBSRT was associated with fewer therapy dropouts (χ2 = 2.75, p < .10). Only 14% of the CBSRT group (n
= 3) attended less than 75% sessions versus PCGT, where 36% of the sample (n = 8) attended less than
75% of sessions.
Ford & Kamerow DB. (1989). JAMA, 262
Mellman TA, et al. (2002). Am J Psychiatry, 159
Zayfert & DeViva (2004). J Trauma Stress., 17
Galovski et al. (2009). J Trauma Stress, 22
Schnurr et al. (2003). Arch Gen Psychiatry, 60
Bastien et al. (2004). J Consult Clin Psychol, 72
Cloitre et al. (2002), J Consult Clin Psychol, 70
Holtzheimer et al. (2005). Am J Psychiatry, 162
Haynes et al. (2009). Sleep (Abstract), 32
Spitzer et al. (1992). Arch Gen Psychiatry, 49
Buysse DJ et al. (2006).Sleep, 29
Weathers et al. (2001). Depress Anxiety, 13
Hamilton (1960). J Neurol, Neurosurg
Psychiatry, 23
14. Ehlers et al. (1988). Arch Gen Psychiatry, 45
15. Yalom (2005). Theory and Practice of Group
9.
10.
11.
12.
13.
Psychotherapy
16. Frueh et al. (2000). Clin Psychol Rev, 20
Acknowledgements
This project was supported by Department of Defense
(Grant #W81XWH-08-2-0121).