PS65CH11-Craighead ARI 31 October 2013 12:37

ANNUAL
REVIEWS Further Combination Psychotherapy
Click here for quick links to
Annual Reviews content online,
including:
and Antidepressant Medication
• Other articles in this volume
• Top cited articles
Treatment for Depression: For
• Top downloaded articles
Whom, When, and How
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

• Our comprehensive search

Access provided by 179.7.195.19 on 05/08/20. For personal use only.

W. Edward Craighead1,2 and Boadie W. Dunlop1

1
Department of Psychiatry and Behavioral Sciences and 2 Department of Psychology, Emory
University, Atlanta, Georgia 30322; email: ecraigh@emory.edu, bdunlop@emory.edu

Annu. Rev. Psychol. 2014. 65:267–300 Keywords

First published online as a Review in Advance on major depressive disorder, CBT, antidepressant medications, combination
September 13, 2013
treatment, mechanisms of therapeutic change, neurobiology of depression
The Annual Review of Psychology is online at change
http://psych.annualreviews.org

This article’s doi: Abstract

10.1146/annurev.psych.121208.131653
Major depressive disorder (MDD) is among the most frequent and debili-
Copyright  c 2014 by Annual Reviews. tating psychiatric disorders. Efficacious psychotherapy and antidepressant
All rights reserved
medications have been developed, and two-thirds of depressed patients
respond to single-modality treatment; however, only about one-third of
patients remit to single-modality treatments with no meaningful differences
in outcomes between treatment types. This article describes the major
clinical considerations in choosing between single-modality or combination
treatments for MDD. A review of the relevant literature and meta-analyses
provides suggestions for which treatment to use for which patient and when
each treatment or combination should be provided. The review summarizes
the moderators of single-modality and combination-treatment outcomes.
We describe models of mechanisms of treatment efficacy and discuss recent
treatment-specific neurobiological mechanisms of change.

267
 PS65CH11-Craighead ARI 31 October 2013 12:37

Contents
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
CURRENT TREATMENTS FOR MDD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
Antidepressant Medications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
Psychotherapies for MDD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
TREATMENT OUTCOMES FOR MDD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
Outcome Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
Efficacy of Single-Modality Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Concerns with Single-Modality Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
COMBINATION TREATMENTS FOR MDD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
Strategies for Combination Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

Rationale for Combination Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273

Considerations for Interpreting Combination-Treatment Study Outcomes . . . . . . . . 276
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

Limitations of Meta-Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278

EFFICACY OF COMBINATION TREATMENTS: ACUTE EFFECTS . . . . . . . . . . 279
Medication Alone versus Combined Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Psychotherapy Alone versus Combined Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
Summary of Acute-Phase Combination Treatments for MDD . . . . . . . . . . . . . . . . . . . . 280
EFFICACY OF COMBINATION TREATMENTS: PREVENTION OF
RELAPSE/RECURRENCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
Design Issues During Maintenance-Phase Interventions . . . . . . . . . . . . . . . . . . . . . . . . . . 281
Design Improvements for Maintenance-Phase Interventions . . . . . . . . . . . . . . . . . . . . . . 282
Mechanisms of Combination Relapse Prevention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
Summary of Maintenance-Phase Combination Treatments for MDD . . . . . . . . . . . . . 284
IMPROVING APPLICATIONS OF COMBINATION TREATMENT . . . . . . . . . . . 284
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
Moderators of Outcomes with Combination Treatment: Which Patients
Need Combination Treatment? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
FUTURE DIRECTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290

INTRODUCTION
Major depressive disorder (MDD) is among the most prevalent and debilitating of psychiatric
problems affecting about 1 in 6 women and 1 in 10 men over the life span (Kessler et al. 2005). MDD
negatively impacts an individual’s quality of life and produces impairments in role functioning
(Wells et al. 1989). MDD is a leading cause of family dysfunction and death from suicide (Arató
et al. 1988, Rich et al. 1988), and it contributes to increased risk of morbidity and mortality from
diabetes, cardiovascular disease, and stroke (Barth et al. 2005, Egede 2006, Loeb et al. 2012). MDD
reduces productivity and is a major health expense, thereby becoming a societal and economic
problem. It is the fourth leading cause of disability worldwide, and by 2020 it is projected to be
second only to ischemic heart disease as a cause of disability (Murray & Lopez 1997). By 2030,
MDD is expected to be the largest worldwide contributor to disease burden (World Health Organ.
2004).

268 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

Although highly variable across studies, the average age of onset for initial depressive episodes
is late adolescence or emerging adulthood; the majority of individuals who will experience MDD
will have done so by early adulthood (Hankin & Abramson 1999, Kessler et al. 2005). MDD is
also a recurrent disorder, and it frequently becomes a chronic problem. For example, Lewinsohn
et al. (1999; see also Fergusson & Woodward 2002) found that ∼45% of adolescents who suffer
an episode of MDD experience a recurrence by age 24, and Rohde and colleagues (2012) found
that cumulatively among the same group ∼50% had a subsequent depressive episode by age
30. Episodes in which full criteria for MDD are present continuously for 2 years or more are
considered chronic; about 30% of depressed individuals in the community and 50% of those who
are in treatment suffer from chronic depression (Klein & Black 2013, Mueller et al. 1999, Waraich
et al. 2004). The duration of MDD episodes is also highly variable; estimates of the average time
to spontaneous recovery range from 5 months to 13 months (Angst et al. 2003, Lewinsohn et al.
1999). Risk factors for the development of MDD include positive family history for a mood
disorder, early life trauma, prior anxiety disorders, and substance abuse (see Ritschel et al. 2013).
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

MDD as defined by current nosologies [e.g., Diagnostic and Statistical Manual of Mental Disorders,
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

5th Edition (DSM-5; APA 2013)] is a heterogeneous syndrome, with multiple likely etiologies
and pathological forms. As shall become apparent, this heterogeneity contributes strongly to the
variable response to treatment observed in clinical trials of treatments for depression.

CURRENT TREATMENTS FOR MDD

Antidepressant medications (ADM) and psychotherapies are the two primary outpatient treat-
ments for MDD. Various forms of stimulation treatments (e.g., transcranial magnetic stimulation,
deep brain stimulation, vagus nerve stimulation, electroconvulsive therapy) exist for patients who
do not respond to ADM or psychotherapy, but stimulation treatments are not well studied as forms
of combination treatment and are not discussed further in this review (Kennedy & Giacobbe 2007).

Antidepressant Medications
Although the first antidepressants were identified through serendipitous discovery, more recent
developments in pharmacology are driven by current understandings of the biological functions
of neurotransmitter systems within the central nervous system. There are several ADM classes,
which are categorized by their structural or functional relationships. No antidepressant class or
individual medication has consistently proven to be superior to others for the treatment of MDD
(APA 2010). Thus, recommendations for first-line ADM treatments for MDD are based primarily
on tolerability and safety rather than on efficacy. Selective serotonin reuptake inhibitors (SSRIs),
serotonin-norepinephrine reuptake inhibitors (SNRIs), bupropion, or mirtazapine are considered
first-line antidepressants (APA 2010). Owing to their greater health risks, tricyclic antidepressants
and monoamine oxidase inhibitors are generally used only after first-line agents fail. Treatment
with ADM should be sustained for at least 4–8 weeks prior to determining efficacy. The goal
of treatment is essentially to eliminate depressive symptoms; dosage is increased every 2–4 weeks
(unless side effects prevent further increases) until either that goal is achieved or the FDA maximum
recommended dose is reached (APA 2010).
DSM-5 (APA 2013) identifies several clinical subtypes of MDD, including those character-
ized by melancholic, atypical, or psychotic features. Standard of care for MDD with psychotic
features (i.e., episodes characterized by hallucinations, delusions, or catatonia) includes medica-
tion treatment with antipsychotics or electroconvulsive therapy. Beyond depression with psychotic

www.annualreviews.org • Combination Treatment of MDD 269

 PS65CH11-Craighead ARI 31 October 2013 12:37

features, there currently are no reliable clinical or biological measures to match individual patients
to specific treatments to maximize outcomes.

Psychotherapies for MDD

Psychological treatments for MDD are derived from the theoretical constructs about the nature of
the disorder and the psychological factors that permit its persistence. Evidence-based psychother-
apies for the acute treatment of MDD include Beck’s eponymous cognitive-behavioral therapy
(CBT; Beck et al. 1979), interpersonal therapy (IPT; Klerman et al. 1984), and behavioral acti-
vation (BA; Jacobson et al. 2001). Other forms of psychotherapy with some support for efficacy
in MDD include behavioral marital therapy (Beach et al. 1990), problem-solving focused therapy
(e.g., Nezu et al. 2013), and short-term psychodynamic psychotherapy (Driessen et al. 2010). All
these therapies are short term, educational, and fairly directive, and they typically comprise 16–20
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

sessions over 12–16 weeks. Although CBT has been more extensively evaluated and is discussed
more thoroughly in this article, there is minimal empirical evidence to support the choice of one
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

treatment over another; therefore, the decision of which treatment to use is typically based on
therapist competence, therapist availability, marital status, and patient preference.

TREATMENT OUTCOMES FOR MDD

Outcome Definitions
The most widely accepted outcome target in the treatment of MDD is remission, in which a
patient has returned to the pre-episode level of functioning and has no or only a few mild and
infrequently experienced residual depressive symptoms. In clinical trials, remission is defined most
often as a rating scale score below a certain threshold, most commonly a Hamilton Depression
Rating Scale (HDRS; Hamilton 1960) 17-item total score of ≤7. Some studies use alternative
rating scales, such as the Montgomery Åsberg Depression Rating Scale (MADRS; Montgomery
& Åsberg 1979), the Quick Inventory of Depressive Symptoms (QIDS; Rush et al. 2003), or the
Beck Depression Inventory-II (BDI-II, or a self-report measure; Beck et al. 1996). One important
caveat regarding the clinical trial definitions of remission is that these cutoffs were arrived at by
consensus (Frank et al. 1991); likewise, subsequent analyses suggest that these thresholds may
not accurately or adequately reflect functional recovery (Dunlop et al. 2012b, Zimmerman et al.
2012).
Another commonly used outcome measure is response, typically defined as a 50% or more
decrease from the baseline score on a rating scale. Response reflects clear improvement with
treatment, but it does not imply that full remission has been achieved. Response short of remission
is generally an unsatisfactory outcome because the presence of ongoing significant symptoms at
the end of treatment is the strongest predictor of future depressive episodes ( Judd et al. 2000,
Paykel et al. 1995). A return to a full depressive episode within two months of the end of an episode
is termed a relapse, whereas such an episode occurring more than two months after remission or
response is considered a recurrence, denoting a new major depressive episode. Consequently,
maintaining remission for at least two months is referred to as recovery (Frank et al. 1991).
The validity of these distinctions has been questioned (Rush et al. 2006), but they are noted
here because they are important concepts for interpreting treatment and relapse-prevention trials
discussed below.

270 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

Efficacy of Single-Modality Treatments

Only about one-third of individuals with MDD seek and receive treatment of any type, and
members of minority groups are particularly unlikely to seek or receive depression treatment
(Gonzalez et al. 2010). Among those who do receive treatment, evidence-based psychotherapies
and ADM are equally efficacious, on average (APA 2013, Cuijpers et al. 2008, Gelenberg et al.
2010, Khan et al. 2012, Spielmans et al. 2011). The data for head-to-head trials of psychother-
apy alone versus medication alone are not included in this review; they have been extensively
reviewed by others (Cuijpers et al. 2008, Spielmans et al. 2011). The outcome data from in-
dividual clinical trials of treatments for MDD are highly variable. Nevertheless, reviews and
meta-analyses consistently conclude that regardless of the form of treatment, about two-thirds
of all patients with MDD will show a positive response to an active treatment in a clinical trial,
but only 30–40% of treated patients with MDD will achieve remission (e.g., DeRubeis et al.
2005b).
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

The overwhelming majority of clinical trials for MDD involve strict inclusion and exclusion
criteria that may reduce generalizability of the results to clinical practice. Four common exclusions
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

used in studies of adults with MDD are (a) presence of psychotic symptoms; (b) presence of clinically
significant suicidal ideation; (c) mild severity, typically identified as a screening-visit score below a
threshold on a depression rating scale; and (d ) a medical illness that may interfere with treatment or
impair response to treatment. Many studies also exclude individuals with psychiatric comorbidities,
such as concurrent substance abuse, eating disorders, and certain anxiety disorders (e.g., obsessive
compulsive disorder). Community, nonresearch protocol patients seeking treatment often fall into
one or more of these exclusion categories, which may limit extension of these research findings to
treatment selection in routine clinical care. Many measured and unmeasured factors contribute
to variability in results between trials, and these are discussed in greater detail below. Although
considered in the most sophisticated meta-analyses, these preceding factors severely limit the
conclusions that can be reached and the clinical practice implications derived from those reports.
The conclusions that investigators do reach are further limited by the small sample sizes in most
of the individual clinical trials and the redundant inclusion of those same small trials in extant
meta-analytic studies.
More recently, clinical scientists have shown increasing concern about the degree of placebo
response in clinical trials, an effect that may lead to an exaggerated expectation of benefit in
community clinical settings (Dunlop et al. 2012d). Placebo response identified in clinical trials is
a multifaceted phenomenon, capturing elements of expectation, time and attention, spontaneous
recovery, and regression to the mean effects (Frank & Frank 1991). Patients in clinical trials,
particularly those evaluating ADM versus placebo, are provided substantially more contact with
a treatment team than is observed in routine clinical practice; thus patients assigned to placebo
in clinical trials may experience substantial benefit. For psychotherapy treatments, competence of
treatment delivery may be greater in clinical trials, which employ specifically trained therapists,
than the level of clinical competence provided in community settings. Thus, for both ADM
and psychotherapeutic treatments there are significant concerns regarding the generalizability of
clinical trial results.
In summary, the aggregated data reported in the higher-quality meta-analytic studies and
qualitative reviews support a consistent conclusion. Namely, single-modality treatments for MDD
produce essentially identical outcomes regardless of the treatment given, though some individuals
benefit specifically from one form of treatment and not another (McGrath et al. 2013). Although
two-thirds of patients show a clinical response, only about one-third of treated MDD patients
achieve remission.

www.annualreviews.org • Combination Treatment of MDD 271

 PS65CH11-Craighead ARI 31 October 2013 12:37

Concerns with Single-Modality Treatments

Using ADM and psychotherapy as monotherapies can present significant problems and limitations.
For example, each of the ADMs has side effects, and these frequently affect the willingness of
patients to complete a course of the medicine. The medications can be expensive, especially when
they must be maintained over a long time period, such as in the treatment of chronic depression.
Both these factors as well as other variables (e.g., patient’s disorganization) result in lower-than-
effective adherence to prescribed medical regimens. Furthermore, up to 20% of patients taking
ADM for two years will experience a depressive recurrence (Hansen et al. 2008, Kocsis et al. 2007),
a phenomenon known as depressive recurrence on antidepressant therapy (DRAT) or “poop out”
(Dunlop 2013). Finally, even very effective ADM have a high relapse rate, exceeding 50%, even if
a medication is taken as a maintenance treatment for 12–15 months (Hollon et al. 2005a).
Likewise, there are issues related to delivery of short-term evidence-based psychotherapies. For
example, when a psychotherapy does not produce a remission, patients frequently experience a
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

resulting demoralization and may be unlikely to seek further treatment; thus, the MDD continues
and likely worsens. In contrast to the fairly common aphorism, “at least psychotherapy doesn’t
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

hurt patients,” the demoralized patient does suffer as a result of ineffective treatment (Craighead
et al. 2013a). Even though a clinical trial may show that a specific treatment (e.g., CBT or IPT) is
efficacious for MDD, these evidence-based therapies are not very well disseminated, thus limiting
the availability of evidence-based treatments in much of the United States. Limited dissemination
also leads many therapists to claims that they conduct evidence-based therapy, but in fact, they may
have had very minimal, if any, formal training in the therapy that is purportedly being delivered.
Consequently, psychotherapy may be incompetently delivered (e.g., CBT; DeRubeis et al. 2005b),
resulting in poorer clinical outcomes than reported in clinical trials. In a recent study, 10–15% of
community psychotherapy providers produced negative outcomes (actual deterioration) among
their patients (Kraus et al. 2011).
Evidence-based therapies need to be disseminated and delivered in a manner functionally
equivalent to those evaluated in the clinical trials that support their effectiveness. For example,
after reviewing the relevant literature, the National Health Service in England adopted CBT as
the first-line treatment for MDD. They have invested more than £400 million to train 6,000
therapists who will be competent in CBT (Cent. Ment. Health et al. 2012).
The greatest concern with all the monotherapies (both ADM and psychotherapies) was pre-
viously summarized; namely, only 30–40% of treated MDD patients remit with a monotherapy
when treatment is offered under ideal conditions with competent professionals. Despite these con-
cerns, each approach to treatment has its positive aspects. For example, ADM generally produce
faster treatment effects (Keller et al. 2000), and psychotherapy (at least CBT) seems to confer
more enduring maintenance of positive outcomes (Craighead et al. 2007, Hollon et al. 2005a).
As we discuss below, the positive effects of the two treatments have influenced researchers and
practitioners to consider treatment combinations in an effort to enhance their overall clinical
effectiveness.

COMBINATION TREATMENTS FOR MDD

Overview
As noted above, failure to achieve full recovery from MDD is the strongest predictor of a future,
recurrent episode, so improvement in short-term treatment outcomes and prevention of depressive
relapses and recurrences are greatly needed. Over the past 30 years, many studies have examined the

272 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

efficacy of combined medication and psychotherapy in the acute treatment of MDD and the effect
of combination therapy on maintenance of wellness. Outcomes from combination versus single-
modality treatment trials are mixed, although the overall weight of evidence supports combination
as superior (see Table 1). However, a key clinical challenge that remains to be solved is to identify
which individuals require combination treatment to achieve and maintain recovery from MDD.
The remainder of this article summarizes the state of knowledge about the efficacy of combi-
nation treatment for MDD, identifies the limitations of the current knowledge base, and suggests
directions for future research. The following questions are addressed: Does combination treatment
for MDD produce superior short-term outcomes relative to single treatments? Do combination
treatments for MDD produce protection against relapse/recurrence? Should combination treat-
ment be administered concurrently from the beginning of treatment, sequentially if response to
the initial treatment is inadequate, or in combination after a trial of monotherapy? Of the large
clinical trials evaluating combination treatments for MDD, which important limitations affect the
internal validity and generalizability of results to impact clinical care of depressed patients?
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

Strategies for Combination Treatments

Practitioners can utilize three strategies to employ monotherapies in combination to treat MDD.
First, the treatments may be employed concurrently from the beginning of treatment. Second,
the treatments may be employed sequentially: Either ADM or psychotherapy is offered first, and
the other is added to augment the first if the monotherapy fails or is only partially successful.
Finally, ADM and therapy may be offered within stepped treatment, which when conceptually
driven is referred to as clinical staging (see McGorry et al. 2010). Within the clinical staging
model, a combination of decisions regarding the patient’s stage of the disorder is used to guide the
treatment process. Clinical staging occurs most frequently when ADM is used initially to treat an
individual’s debilitating depressive symptoms; the treatment effects may be fairly rapid, and then
a psychosocial intervention is added to address the remaining symptoms and problems in daily
life (e.g., marital conflict, cognitive distortions, social interactions). A clear demonstration of this
process occurs when deep brain stimulation (Ramirez et al. 2013) alleviates a patient’s most severe
depressive symptoms, but some combination of evidence-based therapy (e.g., behavioral activation
and dialectical behavior therapy) is needed to teach living skills to the chronically depressed patient.
Below, we note that data regarding neural impacts of clinical interventions augur for a clinical
staging approach to combination treatments.

Rationale for Combination Treatments

There are at least three different conceptualizations of the mechanisms by which combined treat-
ment may produce better outcomes than individual treatment.

Additive model. At a study level, some individuals remit only with medication, and others remit
only with psychotherapy (in addition to those who will remit to neither and those who will
remit to either) (Schatzberg et al. 2005). Thus, by simple addition, the proportion of remitters
in a combined-therapy condition will be greater than the proportion of those treated with a
monotherapy because some of those who will not have remitted to one of the monotherapies will
have remitted to the other monotherapy. This effect produces higher rates of remission in the
combined treatment. At an individual level, additive effects may be important, particularly for
the goal of achieving remission. One of the monotherapies may produce benefit but leave certain
symptoms unaddressed (i.e., a response but not a remission). One example of this additive model

www.annualreviews.org • Combination Treatment of MDD 273

 Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org
Access provided by 179.7.195.19 on 05/08/20. For personal use only.
PS65CH11-Craighead

Table 1 Large randomized trials comparing single-modality treatments with combination treatment for acute treatment of major depressive disorder
ARI

274
%
Author Chronic Mean Duration Remission
(year) Design Patients MDE HDRS Arms (N) (weeks) Medication Dropouts definition Remission rate
Keller Randomize Recruited 35 26.9 ADM (220) 12 Nefazodone ADM: 26% HDRS ≤8 at ADM: 29%

Craighead
·
et al. from outpa- (24- CBASP (216) CBASP: weeks 10 CBASP: 33%
31 October 2013

(2000) start tients with item) ADM + 24% and 12 for ADM + CBASP:

Dunlop
MDD; CBASP (226) ADM + completers 48%
depressive CBASP: or at final p < 0.001 for
12:37

symptoms 21% visit for combination

≥2 years dropouts versus either
HDRS 24: ADM or CBASP
≥20 alone
de Jonghe Randomize Consecutive 18 20.4 ADM (57) 24 Fluoxetine, ADM: 42% HDRS <8 ADM: 22.8%
et al. from outpa- ADM + SPSP amitripty- ADM + ADM + SPSP:
(2001) start tients with (72) line, SPSP: 43.1%
MDD moclobe- 18% p = 0.02
HDRS 17: mide,
≥14 sequentially
de Jonghe Randomize Consecutive 15 18.1 SPSP (106) 24 Various SPSP: 29% HDRS <8 SPSP: 32.1%
et al. from outpa- SPSP + ADM SPSP + SPSP + ADM:
(2004) start tients with (85) ADM: 42.4%
MDD 23% p = 0.14
HDRS 17:
12–24
Schramm Randomize Inpatients 36 23.5 ADM (61) 5 Sertraline, ADM: 15% HDRS <8 ADM: 34%
et al. from with ADM + IPT amitriptyline, ADM + ADM + IPT: 49%
(2007) start MDE as (63) or IPT: 16% p = 0.105
part of amitriptyline-
MDD or N-oxide
BP II
HDRS 17:
≥16
 Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org
Access provided by 179.7.195.19 on 05/08/20. For personal use only.
PS65CH11-Craighead
ARI

Blom Randomize Clinically n.r. 21.4 ADM (47) 12–16 Nefazodone ADM: HDRS ≤8 19.3% overall
et al. from referred IPT (50) 36.2% Remission OR for
(2007) start outpa- ADM + IPT IPT: 32% ADM + IPT
tients with (49) ADM + versus ADM: 3.22
MDD IPT + PBO IPT: (1.02–10.12,
31 October 2013

HDRS 17: (47) 32.7% p = 0.045)

≥14 IPT + Remission OR for
PBO: ADM + IPT
12:37

25.5% versus IPT or IPT

+ PLA: n.s.
Lesperance Randomize Recruited 22 22.8 ADM + CM 12 Citalopram ADM + HDRS 24 ADM + CM: 36%
et al. from outpa- (75) CM: 13% ≤8 IPT + PBO +
(2007) start tients with IPT + PBO + IPT + PBO CM: 21%
MDD and CM (75) + CM: ADM + IPT +
coronary ADM + IPT 21% CM: 36%
artery + CM (67) ADM + PBO + CM: 23%
disease PBO + CM IPT + IPT versus PBO:
HDRS 24: (67) CM: 12% p = n.s.
≥20 PBO + ADM versus PBO:
CM: 30% p = 0.01
Kocsis Randomize Recruited 58 19.3 ADM (94) 12 Various ADM: 17% HDRS 24 ADM: 39.5%
et al. after outpa- (24- ADM + ADM + <8 and ADM + CBASP:
(2009) nonre- tients with item) CBASP (195) CBASP: ≥50% 38.5%
mission MDD; ADM + BSP 13% decrease ADM + BSP:
with depressive (189) ADM + from BL 31.0%
prospec- symptoms BSP: 14% p = n.s.
tive ≥2 years
ADM No min.

www.annualreviews.org • Combination Treatment of MDD

for 12 HDRS
weeks

275
Abbreviations: ADM, antidepressant medication; BL, baseline; BP, bipolar disorder; BSP, brief supportive therapy; CBASP, cognitive behavioral analysis system of psychotherapy; CM, clinical
management; HDRS, Hamilton Depression Rating Scale; IPT, interpersonal therapy; MDD, major depressive disorder; MDE, major depressive episode; n.r., not reported; n.s., not significant;
OR, odds ratio; PBO, placebo; SPSP, short psychodynamic supportive psychotherapy.
 PS65CH11-Craighead ARI 31 October 2013 12:37

in an individual is a case in which medication reduces sadness, suicidal ideation, and insomnia but
leaves persisting anhedonia and guilt, which may be resolved through CBT. Through addressing
these residual symptoms, combination treatment may convert a nonremitter to remitter status.

Synergistic model. In synergies, effects that could not be achieved by either individual interven-
tion are possible only when methods are used together. For example, ADM improve hippocampal
function through increased cell survival and dendritic growth (Encinas et al. 2006). Because the
hippocampus is crucial for new learning, patients with severe hippocampal dysfunction from their
depression may require biological enhancement of hippocampal function to be able to utilize
CBT. Similarly, highly anergic patients may require a medication to allow them to “get over the
hump” to engage in rewarding activities that are a key component of BA. In both cases, the two
treatments are required to make synergistic, meaningful gains. Alternatively, based on the partial
or full completion of CBT, patients who need ADM but, for any number of reasons, might refuse
medication initially may, as a result of CBT, experience a cognitive change that results in the
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

patient recognizing the need to take ADM. The cognitive change in this example works synergis-
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

tically to facilitate the acceptance of ADM by patients who need the medications but otherwise
might refuse them. This synergy differs from the simpler additive model and the straightforward
enhancement of adherence to medications.

Adherence model. Although a more prosaic point, adherence to treatment is crucial to the
success of both ADM and psychotherapy treatments. Psychotherapy can enhance adherence to
ADM dosing through educational and motivational impacts. Conversely, patients who may be
at risk for missing therapy appointments owing to anergia or who cannot complete homework
assignments owing to concentration deficits may be aided in these tasks through medication effects.
In addition to these three models for improved outcomes during acute treatment, combination
treatment may have effects on the long-term course of MDD. Specifically, combination treatment
may provide better protection against relapse and recurrence after remission.

Considerations for Interpreting Combination-Treatment Study Outcomes

Combination-treatment studies are challenging to implement and susceptible to a variety of inter-
nal and external validity threats that require consideration. Table 2 highlights potential sources
of bias in trials that compare the benefits of one modality of treatment versus the other or of a
combination treatment versus a single-modality treatment.
Sample selection and treatment preferences present particularly knotty issues in evaluating the
efficacy of psychotherapy in combination trials. Because all patients in clinical trials are volunteers,
a significant potential for selection bias exists. For example, there appears to be great variability
among the patient samples enrolled in different trials, and this variability can also occur between
sites in the same trial (DeRubeis et al. 2005a). Many trials have not examined patient preference
prior to randomization, and some studies show high rates of patient refusal to start a treatment
to which they have been randomized, indicating the presence of strong patient preferences (de
Jonghe et al. 2001, 2004; Dimidjian et al. 2006). Matching patients to their treatment preference
may have sizeable effects on outcomes, so differential preference rates in a randomized trial may
impact results unless patients who would refuse a particular treatment are prospectively excluded
from participation. This aspect of clinical trials has received inadequate attention in trial design
and interpretation.
The pharmacotherapy arm in a clinical trial may not reflect real-world clinical practice. Specif-
ically, patients receiving medication only in clinical trials may still receive significant amounts of

276 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

Table 2 Considerations for acute studies of combination versus individual treatments for major depressive disorder (MDD)
Question Concern
Was the psychotherapy arm an evidence-based Use of psychotherapies not proven to be effective in the treatment of MDD may
form of treatment for MDD? underestimate benefits.
Was pharmacotherapy administered in a manner 1. Was pharmacotherapy dosed adequately?
consistent with standard clinical care? 2. Was an early switch to an alternative medication allowed for patients unable
to tolerate the initial medication?
Were adequate control conditions present for 1. Including a pill-placebo arm but no control form of psychotherapy distorts
both the pharmacotherapy and the expectation bias against medication.
psychotherapy arms? 2. Wait-list controls provide no meaningful expectation of improvement and do
not represent an adequate control for psychotherapy.
Was the clinic setting of the study generally 1. Substantial sampling bias may arise from studies performed in locations
known for expertise in providing only one form where patients expect one particular form of treatment.
of treatment? 2. Study investigators’ allegiance to specific forms of treatment may have
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

substantial effects on study outcomes.

Was the trial’s duration adequate to allow for the Trials of less than 12 weeks’ duration may not allow adequate time for the
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

emergence and detection of the benefits of benefits of monotherapy with psychotherapy to emerge nor for
combination treatment? medication-induced remission to occur.
Were the rates of early termination and refusal of High rates of dropout from one treatment arm may indicate a study sample
randomization assignment different between the biased against that form of treatment.
treatment arms?
Was the study powered adequately? 1. Conclusions of “no benefit” from combination treatment may arise from an
inadequate power to detect statistically significant change.
2. Conclusions of “benefit” from combination treatment may reflect chance
results in small samples.
Were treatment preferences assessed prior to Trial outcomes may be affected by expectancy. Samples of patients in which
randomization? there is a strong preference for one treatment over another may not reflect
usual clinical practice.

time with the prescribing physician, including the provision of some aspects of supportive therapy
(Dunlop & Vaughan 2010). It is, therefore, possible that combination treatment may be more
efficacious in clinical trials than in real-world settings, where pharmacotherapy visits may be very
brief “med-check” appointments. Conversely, clinical trials involving psychotherapy typically in-
corporate well-trained therapists, who can easily communicate with the trial pharmacotherapists.
Such valuable aspects of psychotherapy provision may not apply in many real-world settings,
consequently resulting in an overestimation of the value of combined treatment in actual clinical
practice.
Expectation biases are another potential problem in multiarmed clinical trials, in which one
of the arms is a placebo pill. In such trials, patients assigned to a medication condition do not
know whether they are receiving an active treatment. In contrast, psychotherapy control con-
ditions are not typically presented to patients as “placebo psychotherapy,” but rather they are
represented as an effective form of psychotherapy, with extensive description of the treatment. In
particular, more severely depressed or impaired patients may be particularly unwilling to enroll
in trials involving potential assignment to placebo medication. In contrast, a wait-list condition
provides no benefits from attention or hope, and it is also an inadequate control condition for
psychotherapy.
Finally, the singular focus on depression rating scale changes as the measure of outcome can
lead investigators to overlook other important clinical outcomes that may emerge from combined

www.annualreviews.org • Combination Treatment of MDD 277

 PS65CH11-Craighead ARI 31 October 2013 12:37

treatment, specifically improvements in quality of life and functioning. A review of 12 natural-

istic studies of ADM-treated MDD patients followed for at least 10 years found that longitudi-
nal changes in depressive symptoms were not accompanied by changes in patients’ psychosocial
functioning (Hughes & Cohen 2009). Individual studies examining patients with chronic forms
of depression have found that combination treatment significantly improved quality of life over
medication monotherapy (Greenberg 1999, Molenaar et al. 2007, von Wolff et al. 2012). Acknowl-
edging this oversight in prior work and incorporating appropriate measures of these constructs are
essential as MDD combination-treatment outcome research progresses. Indeed, an exclusive focus
on depressive symptomatology as compared with other domains of functioning may overlook the
greatest benefit of added psychotherapy in combination treatment.

Limitations of Meta-Analysis
Meta-analysis is considered the preferred method to evaluate treatment outcomes across studies,
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

and it can be employed to overcome design concerns that may affect individual studies. Several
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

meta-analyses have examined various aspects of the efficacy of combined versus single-modality
treatments for MDD. Meta-analysis is a powerful technique to identify mean effect sizes and to
compare outcomes quantitatively across a broad range of studies, particularly when many small
studies have been conducted, as is the case in the acute treatment of MDD. Meta-analysis, however,
has several important limitations.
Specifically, studies included in a meta-analysis should be sufficiently similar in design to pro-
duce a meaningful estimate of effect size for an actual clinical intervention (Higgins & Green
2011, Lieberman et al. 2005). Heterogeneity in trial design factors listed in Table 2 may be
differentially valued by those conducting meta-analyses; these discrepant evaluations across inves-
tigators may result in differing conclusions based on different meta-analyses. For studies of MDD
combination treatment, several issues require particular consideration to determine whether the
studies included in a meta-analysis are similar enough in design to produce valid results. First,
although the additive and synergistic models of the benefits of combination treatment imply that
combination treatment should be superior to either treatment delivered singly, it is possible that
combination therapy is superior to medication-only treatment but not superior to psychotherapy
alone. Alternatively, combination therapy might be superior to psychotherapy only but may not
improve medication-treatment outcomes. Thus, meta-analyses ideally would be performed for all
combination-treatment studies with two sensitivity analyses, specifically evaluating the benefit of
combination therapy versus ADM alone and combination therapy versus psychotherapy alone.
Second, investigators need to define an adequate form of each treatment. For medication,
underdosing or not allowing a switch for medication intolerance reflects an inadequate repre-
sentation of regular psychopharmacologic practice. For experimental psychotherapy conditions
the question is even more complex: (a) What is considered an efficacious psychotherapy? Should
only evidence-based forms of psychotherapy be included, or should all forms of counseling be
included? Any form of psychotherapy may provide benefit to patients through the shared effects
common to all therapies, but additional benefit may arise for some patients specifically from the
evidence-based treatments. (b) What is considered an adequate psychotherapy dose (i.e., number
and duration of sessions)? (c) How much experience do therapists need to be considered adequate
to provide competent clinical practice? Failure to consider adequacy of treatments may bias meta-
analyses toward null findings or toward evidence of positive effects, depending on the specific
studies included, a decision not infrequently made in an arbitrary fashion.
A third consideration is the need for a measure of publication bias, such as a funnel plot or using
Duval and Tweedie’s trim and fill procedure (Duval & Tweedie 2000). For various reasons, small

278 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

studies with positive results are more likely to be published than are small studies with negative
results. Without considering the presence of this bias, meta-analyses may falsely conclude that
efficacy exists, owing to the absence of negative unpublished results from small studies. In one
meta-analysis controlling for publication bias of studies of psychological treatments for MDD,
the mean effect size estimate was reduced from 0.67 to 0.42 (Cuijpers et al. 2010a).
In sum, the previously reviewed issues dramatically affect the outcomes of meta-analyses. Thus,
when interpreting the meaning of a meta-analysis, individuals must review the selection criteria
for studies included in the analyses. The inclusion of many small studies with weak selection
criteria and poor methodological implementation, even if the results are statistically significant,
can render the outcome of a meta-analysis clinically and scientifically inconsequential.

EFFICACY OF COMBINATION TREATMENTS: ACUTE EFFECTS

Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

Medication Alone versus Combined Treatments

Access provided by 179.7.195.19 on 05/08/20. For personal use only.

Meta-analyses of ADM versus combination treatment have consistently found small benefits from
combination treatment over ADM alone. The broadest meta-analysis (Cuijpers et al. 2009a) of 25
studies, which included studies of all forms of depressive disorders, found an effect size of Cohen’s
d = 0.31. Importantly, the benefit of combination treatment was driven exclusively by studies of
patients with MDD (d = 0.40), with no effect seen in studies of dysthymia (d = 0.00). A meta-
analysis limited to 16 trials treating MDD, including a total of 1,842 patients, found a remission
rate of 33% for medication monotherapy versus 44% for combination treatment (Pampallona
et al. 2004). The odds ratio for remission with combination treatment was 1.86 [95% confidence
interval (CI): 1.38–2.52]. This analysis also suggested greater benefit for combination therapy
for studies employing treatment durations greater than 12 weeks; this effect was due partly to
a smaller percentage of patients dropping out of the combined treatment condition compared
to the monotherapy treatment condition in the longer studies. From these analyses, the number
needed to treat (NNT) with combination therapy to gain one additional remission over medication
monotherapy is 8–9. This NNT is comparable to the NNT of 7 for SSRIs versus placebo to achieve
remission in clinical trials of MDD (Arroll et al. 2009).
To delineate the specific benefit of ADM versus that of the placebo effect of a pill, Cuijpers
and colleagues (2010b) conducted a meta-analysis comparing the combination of psychotherapy
with either active ADM or a placebo pill. The effect size for combination treatment with ADM
over pill placebo was only d = 0.25 (95% CI: 0.03–0.46). Again, however, the studies included
in that analysis included very diverse groups of patients (many studies compared medically ill or
substance-abusing patients), so comparability with other meta-analyses is limited.
One large clinical trial that failed to find benefit from combination treatment over ADM
alone was the Research Evaluating the Value of Augmenting Medication with Psychotherapy
(REVAMP) (Kocsis et al. 2009). In this trial, chronically depressed patients who failed to remit
after 12 weeks of treatment with a single ADM were randomly assigned to augmentation with
(a) cognitive behavioral analysis system of psychotherapy (CBASP), (b) brief supportive therapy,
or (c) a medication change (either a switch to a second ADM or a two-drug combination). Because
the trial was enriched with medication nonremitters, investigators might have expected it to
be particularly likely to show the benefit of the psychotherapy addition to ADM. However, no
meaningful difference was found among the three treatment approaches on any efficacy measure.
There was a potential for selection bias in this trial versus other combination-treatment trials
because the patients entering REVAMP knew they would all be initially treated with ADM
monotherapy, and few study participants desired psychotherapy treatment alone (Steidtmann et al.

www.annualreviews.org • Combination Treatment of MDD 279

 PS65CH11-Craighead ARI 31 October 2013 12:37

2012). In the only other large trial permitting adjustments in the pharmacotherapy condition, the
combination of short psychodynamic supportive therapy with adjustable pharmacotherapy was
superior to pharmacotherapy alone (de Jonghe et al. 2001). The sample for that trial was highly
psychotherapy preferring; 32% of the patients refused randomization to the ADM monotherapy
arm versus only 13% who refused combination. Thus, it remains uncertain whether the combina-
tion of psychotherapy plus ADM is superior to algorithmically determined and maximized ADM
treatment.

Psychotherapy Alone versus Combined Treatments

A maximally inclusive recent meta-analysis reported a small overall effect size (Cohen’s d = 0.35;
95% CI: 0.25–0.45, p < 0.001) in the change in depression severity favoring combined treatment
versus psychotherapy alone (Cuijpers et al. 2009b). This meta-analysis incorporated studies that
employed any form of psychological intervention and patients with any form of depressive disor-
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

der, which included studies limited to specific subgroups (such as medically ill subjects). Within
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

combination treatments, non-CBT therapies produced significantly larger effect sizes than did
studies using CBT as the psychotherapy, which suggests that medication produces a bigger con-
tribution to outcomes for studies that employ non-CBT therapies. Although many factors may
have contributed to this finding, one important interpretation may be that CBT has larger antide-
pressive effects than do other psychotherapies; thus, CBT may leave less room for improvement
with ADM cotreatment than other psychotherapies. There was no evidence of publication bias in
the overall analysis, but many of the included studies were not of high quality. Similarly, Cuijpers
et al. (2009b) found essentially no benefit to psychotherapy combined with SSRIs (d = 0.10), but
slightly larger effects were found when psychotherapy was combined with tricyclic antidepressants
(d = 0.35) or other antidepressants (d = 0.47).
A meta-analysis of studies limited to patients with MDD treated in psychiatric specialty settings
and using a formal psychotherapeutic treatment also found benefit for combined treatment over
psychotherapy alone (de Maat et al. 2007). In the seven included studies, dropout rates were equal
between the combined therapy and psychotherapy alone treatment groups. The intent-to-treat
analyses found a 12% greater remission rate for combined treatment over psychotherapy (46%
versus 34%, respectively; p < 0.00007), reflecting a relative risk for remission of 1.32 (95% CI:
1.12–1.56). The remission rate difference was larger and statistically significant only in the mod-
erately depressed, not the mildly depressed patients. Similarly, statistical significance in remission
rates was present in the chronically depressed but not nonchronically depressed patients. Whether
evaluating combination therapy against medication or psychotherapy alone, results consistently
indicate that patients with chronic forms of MDD are most likely to receive enhanced acute
benefits from combination treatment (Friedman et al. 2004, von Wolff et al. 2012).

Summary of Acute-Phase Combination Treatments for MDD

Considering these meta-analyses together, we conclude that for the treatment of MDD, combina-
tion treatment is superior to ADM alone and probably superior to psychotherapy alone, with the
possible exception of CBT. There is little evidence showing that combination therapy is superior to
ADM or psychotherapy alone in mild cases of MDD. In contrast, for dysthymia, a chronic form of
low-grade depression, the addition of psychotherapy does not enhance treatment with ADM alone.
Combination treatment has more rapid effects than does psychotherapy alone. Combination treat-
ment has not definitively proven superior to pharmacotherapy administered in an adjustable man-
ner, at least for patients willing to start treatment with ADM monotherapy. There are, however,

280 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

nearly as many completed meta-analyses as there are high-quality, adequately powered studies; the
field currently needs additional adequately powered investigations rather than more meta-analyses.
Indeed, the field critically needs research studies that employ new designs that investigate which
treatments or combinations of treatments work for which patients under which conditions. This
approach was posed in principle by Paul (1969, p. 62) over 40 years ago, and it seems to be finding
its fruition in recent depression research using contemporary approaches to study personalized
medicine.

EFFICACY OF COMBINATION TREATMENTS: PREVENTION OF

RELAPSE/RECURRENCE
Repeated depressive episodes are the rule rather than the exception in patients with MDD. At least
75% of MDD patients will experience multiple depressive episodes across the life span (Hughes &
Cohen 2009, Judd 1997, Mueller et al. 1999); thus, identifying best practices to maintain wellness
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

and prevent recurrence after recovery from a depressive episode is of great importance. There is
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

growing recognition that the mechanisms involved in recovering and maintaining recovery from
a depressive episode may differ from those involving an episode’s initiation (Post 1992, Segal
et al. 2002, Sheets et al. 2013). Conceptualizing MDD in this manner suggests that different
intervention strategies may be needed during periods of acute-phase treatment versus those of
maintenance treatment, at least for some subsets of depressed patients.
Both CBT and IPT have demonstrated efficacy in preventing recurrence following successful
acute psychotherapy treatment (Dobson et al. 2008, Frank et al. 2007, Hollon et al. 2005a, Vittengl
et al. 2007), though some studies have found IPT to be less efficacious than medication when ADM
are continued through the maintenance phase (Frank et al. 1990, Reynolds et al. 1999). Such
enduring effects of psychotherapy in the absence of ongoing sessions suggest that psychotherapy
treatments change risk factors for recurrence in a manner that treatment with acute ADM does
not (Hollon et al. 2006).

Design Issues During Maintenance-Phase Interventions

Studies examining the protective effects of combination treatment in preventing depressive re-
lapses or recurrences differ in important design characteristics. The specific design issues that have
implications for interpreting results and considering how to employ combination treatments most
effectively to prevent relapse and recurrence are (a) maintenance versus discontinuation of the
ADM employed during the acute phase; (b) the degree of residual depressive symptoms present at
the beginning of the follow-up period; and (c) whether the combination treatment being evaluated
was applied during the acute-treatment phase or sequentially after remission had been achieved
with a single-modality treatment.
An important consideration in relapse/recurrence prevention studies is whether ADM are
continued during the maintenance phase. ADM treatment that is efficacious during acute treatment
protects against recurrence when it is continued during maintenance treatment, but discontinuing
ADM during maintenance increases recurrence risk (Hollon et al. 2005a); risk of recurrence
is even greater for patients with chronic or recurrent episodes (Keller et al. 1998, 2007). The
current standard of care for psychopharmacologic management of most patients with recurrent
MDD recovered with ADM is to maintain ADM indefinitely at the same dose used to achieve
recovery (APA 2010). Several head-to-head maintenance comparisons of psychotherapy alone
versus medication alone indicate psychotherapy is clearly superior in protecting against relapse
if ADM is discontinued (Blackburn et al. 1986; Hollon et al. 1992, 2005a; Kovacs et al. 1981;

www.annualreviews.org • Combination Treatment of MDD 281

 PS65CH11-Craighead ARI 31 October 2013 12:37

Simons et al. 1986), though some studies did not find this effect (Perlis et al. 2002, Shea et al.
1992). When ADM are maintained during follow-up, the data are more mixed; some studies still
indicate a benefit for CBT (Hollon et al. 2005a, Simons et al. 1986), whereas others indicate no
difference between treatment conditions (Evans et al. 1992, Frank et al. 1990).
The degree of recovery from depression required for patients to enter a trial’s maintenance
phase is another important consideration. The greater the presence of residual symptoms is after
treatment for a major depressive episode, the more elevated the risk is for depressive recurrence
compared with patients who achieve a full remission ( Judd et al. 2000, Paykel et al. 1995, Thase
et al. 1992). Any differential levels of residual symptoms within the experimental conditions in a
maintenance-phase combination trial may produce a bias in any observed differential outcomes
(Karp et al. 2004).
Finally, the timing of the provision of the combination treatment is of great importance. Com-
bination treatment can be provided from the outset of treatment, or the two treatment components
can be added sequentially. Some studies of combination treatment reporting maintenance-phase
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

outcomes simply added a naturalistic follow-up phase, with or without ADM continuation, to a
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

study designed to examine the acute-phase efficacy of combination treatment. Thus, the trial’s
primary aim was not to evaluate prevention of recurrence, but rather short-term improvements
(Evans et al. 1992, Maina et al. 2009, Miller et al. 1989), and the maintenance phase of the studies
was simply an add-on.
One of the most important threats to the validity of naturalistic follow-up of trials involving
combination treatment from the outset is the “differential sieve” phenomenon (Klein 1996). The
differential sieve problem arises when only responders or remitters to an acute-phase treatment
enter into the follow-up phase. By not considering all subjects who start treatment, the potential
for differential retention makes maintenance studies vulnerable to bias because patients at high
risk for relapse (i.e., most severely depressed or more difficult to treat) may be more likely to
enter the maintenance phase in one arm of a trial than in a comparator arm (Dunlop et al. 2012b,
Evans et al. 1992). For example, if a combination-treatment arm can push a greater proportion of
higher-risk patients into remission than the monotherapy arm does, then during follow-up there
will be a bias for greater relapses in the combination arm because it will have retained a higher
proportion of greater at-risk patients at the beginning of the maintenance phase. This concern is
partially alleviated if equal proportions of the patients in each condition from the acute-treatment
phase enter the follow-up phase; however, there could still be differential mixtures of at-risk and
less-at-risk patients, owing to differential patterns of acute response. In essence, a more powerful
acute-phase treatment can appear less effective than a comparator in maintenance treatment owing
to features of the patients it helps. Although little evidence indicates that treatment responders
differ from nonresponders in clinically meaningful ways, selection bias may still emerge from
unmeasured important factors (Hollon et al. 2006).

Design Improvements for Maintenance-Phase Interventions

Stronger evidence for the value of combined treatment to prevent recurrence can be gleaned from
trials incorporating randomized discontinuation of components of a combined treatment provided
to all patients from the outset of treatment. Patients achieving remission with combined treatment
are randomly assigned to discontinue either the psychotherapy or the ADM (blinded switch to
placebo), neither, or both after maintaining stable remission through a treatment consolidation
phase of several months. This design allows investigators to identify the relative importance of
treatment components for long-term stability. Two large trials of combined tricyclic ADM plus
IPT reached similar conclusions; namely, the best results through maintenance occur when both

282 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

aspects of treatment are continued, and maintenance of ADM alone has a greater effect on pre-
venting recurrence than does IPT alone among MDD patients initially treated with combination
therapy (Frank et al. 1990, Reynolds et al. 1999).
Sequential treatment designs provide another strong form of evidence for the role of combined
treatment. In this design, patients are randomized to additional treatment in a maintenance phase.
Mindfulness-based cognitive behavior therapy (MBCT) is a manualized, group-based, eight-
session, skills-training form of CBT (Teasdale et al. 2000), which can be added after patients
have achieved remission with ADM treatment. A meta-analysis of six similarly designed MBCT
versus treatment as usual or placebo trials found consistent, similar effect sizes in prevention of
recurrence (Piet & Hougaard 2011). Other studies of patients fully remitting (Bockting et al.
2005) or partially remitting (Fava et al. 1996, 1998; Paykel et al. 1999) to ADM monotherapy
have demonstrated the efficacy of a subsequent course of CBT in preventing recurrences. Only
one study has failed to find a benefit on relapse prevention from adding CBT to ADM therapy
for patients in remission (Perlis et al. 2002). This study was unique because at the same time that
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

CBT was added, the fluoxetine doses for all remitted patients were doubled. This dose increase,
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

combined with the relatively brief follow-up period of 28 weeks, resulted in very few recurrences
during the study. In summary, the efficacy of CBT and MBCT in preventing MDD recurrence
represents one of the strongest and most robust findings in the area of combination treatment of
MDD.

Mechanisms of Combination Relapse Prevention

Researchers do not know how treatment with CBT reduces relapse in medication-treated pa-
tients with residual symptoms. In the relevant studies, there is a minimal change in symptom
burden between patients who do and do not receive CBT for relapse prevention, which suggests
that reduction of symptoms does not explain CBT’s positive effect (Scott et al. 2000). This ef-
fect may be obtained because of the change in cognitive processing and resultant schema change
(à la Beck’s model of CBT); improvement of positive coping mechanisms, perhaps via developing
compensatory skills for poor coping patterns (Persons 1993); an increase in self-efficacy through
internalization of positive life experiences; or an increased ability to employ adaptive metacogni-
tive views of one’s cognitions and emotions (Teasdale 1997). The positive coping interpretation
receives support from the efficacy of problem-solving treatment for minor depression, which di-
rectly targets an avoidant coping style, thought to contribute to maintenance of depressive states
(Oxman et al. 2008), and the metacognitive viewpoint receives support from the MBCT data (see
Segal et al. 2002).
The sustained benefits of cognitively informed psychotherapy treatments in preventing recur-
rence may be most applicable for patients who have experienced a greater number of lifetime
episodes of MDD. Patients with three or more prior episodes demonstrate benefits of MBCT,
whereas patients with two or fewer episodes have not shown significant benefit from this treat-
ment (Ma & Teasdale 2004, Teasdale et al. 2000). Similarly, the addition of an 8-session course
of CBT prevented recurrence in patients with ≥5 prior episodes among patients in remission and
those maintained on treatment as usual (Bockting et al. 2005). In patients with more frequent
episode recurrence, mindfulness and cognitive interventions may act to disrupt the learned asso-
ciations between negative thinking patterns and depressive episodes or may interrupt ruminative
thought processes that may lead to depressed mood states (Segal et al. 2002). One must remember,
however, that the enduring effects of monotherapy CBT have been obtained with treatment-naı̈ve
patients as well as with those who have had recurrent episodes. Thus, a final conclusion regarding
the mechanisms by which the enduring effects of sequentially combined ADM and CBT operate

www.annualreviews.org • Combination Treatment of MDD 283

 PS65CH11-Craighead ARI 31 October 2013 12:37

cannot currently be reached. As we note later, the enduring effects of CBT may emerge for a specific
subset of patients regardless of whether CBT is conducted alone or in combination with ADM.

Summary of Maintenance-Phase Combination Treatments for MDD

Combination treatment extends time to recurrence or relapse in MDD patients fully or partially
remitting to acute-phase treatment. For patients who receive combination treatment during the
acute phase, the best long-term outcomes occur if both components of treatment continue through
the maintenance phase. Discontinuing ADM after its use in an acute-phase treatment carries the
greatest risk for recurrence. Thus, patients who receive combination treatment from the outset will
have a better prognosis if psychotherapy is discontinued, rather than discontinuing ADM. Patients
receiving initial treatment with ADM monotherapy are less likely to experience a recurrence if
they complete a short course of CBT after improving; this benefit is particularly clear for patients
with recurrent depression or those who continue to have residual depressive symptoms upon
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

completion of ADM acute treatment.

Access provided by 179.7.195.19 on 05/08/20. For personal use only.

IMPROVING APPLICATIONS OF COMBINATION TREATMENT

Overview
The question of whether combination treatment for MDD is superior to single-modality treat-
ment is inextricably bound with the following issue: Should combined treatment be administered
concurrently for all patients, or should it be given sequentially for those patients who do not remit
to the initial treatment? The primary advantage for concurrent initial treatment is a shortened
time to remission. Within the sequential approach, in which patients must first demonstrate in-
ability to remit from a single treatment, there is a period of 2–3 months lost while administering
an insufficiently effective treatment. The concurrent approach incurs greater financial costs in the
short run, owing to the greater intensity of treatment, though this concern may be obviated if
long-term costs are reduced through decreased relapse rates.
No controlled studies have explicitly compared concurrent versus sequential treatments. How-
ever, Frank and coworkers studied two consecutive cohorts of women treated with IPT, with
medication added either concurrently from the start of treatment or sequentially for nonremitters
to IPT monotherapy. They found significantly higher remission rates in the sequentially treated
patients (79%) compared with the concurrently treated cohort (66%, p = 0.02). This difference
was even more pronounced in the more severely depressed (HDRS ≥ 20) women (81% versus
58%, respectively; p < 0.02) (Frank et al. 2000). The criteria for participation were similar between
the two trials, but there were important differences between the studies, including the type of an-
tidepressant used and the amount of IPT provided. Although these data are suggestive, conclusive
evidence requires trials specifically designed to compare concurrent and sequential approaches.
A theoretical risk of administering combined treatment is the potential interference of one
treatment with another. The relief provided by medication may reduce motivation to engage
in psychotherapy (Rounsaville et al. 1981). Other interference effects may arise from patients’
psychological interpretation of sources of benefit. Specifically, patients may attribute gains in
treatment solely to medication effects, thereby diminishing their sense of self-efficacy in manag-
ing depressive symptoms. Relevant potential concerns have emerged from the anxiety disorders
literature, in which combination treatment led to worse long-term outcomes than did treat-
ment with CBT alone (Barlow et al. 2000). This negative interaction may arise from mood-state
or context-dependent learning effects, in which extinction learning and coping skills learned in

284 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

psychotherapy when experiencing reduced levels of anxiety (due to medication treatment) are not
fully accessible in higher-level states present after medication has been discontinued (Bouton et al.
2006). Such detrimental effects may be less likely to occur with the effective treatments for MDD,
which do not depend as much on the more exposure learning-based methods employed to treat
anxiety.

Moderators of Outcomes with Combination Treatment: Which Patients

Need Combination Treatment?
Routine provision of combination treatment to all depressed patients is not warranted, given the
increased cost of combination therapy and that many patients improve with a single-modality
treatment. Thus, there is a great need to identify the individuals for whom combination treatment
is necessary to achieve remission and sustain recovery. Efforts are increasing to identify predictor
variables that can match individual patients to the treatment modality that will provide the best
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

chance for remission. Predictors of outcome may be prognostic (i.e., nonspecifically indicating like-
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

lihood of outcome, regardless of treatment modality) or prescriptive (specifically indicating better

outcomes with one form of treatment over another) (Fournier et al. 2009). Sociodemographic and
clinical variables are the most frequently replicated predictors to date, although efforts to identify
biomarkers of treatment outcomes are currently the focus of great interest.

Sociodemographic and clinical moderators. Several demographic characteristics nonspecifi-

cally predict poor outcomes from MDD treatment. The strongest of these is low socioeconomic
status, which is consistently associated with lower likelihood of remission (Lorant et al. 2003).
The consistency of this finding across multiple large studies suggests socioeconomic status should
be included as a covariate in all outcomes analyses from clinical trials. Other well-replicated non-
specific predictors of poor outcome to monotherapies include older age (Fournier et al. 2009,
Hirschfeld et al. 1998, Sotsky et al. 1991, Thase et al. 1997) and lower intelligence and/or limited
education (Dunkin et al. 2000, Haaga et al. 1991, Trivedi et al. 2006).
Researchers have proposed several clinical factors as prescriptive variables to indicate the need
for specific treatments. Factors that could indicate the need for combination treatment are best
derived from study designs that compare a combination-treatment arm with both a medication
arm and a psychotherapy arm. Without individual arms for each modality, it is impossible to know
whether any identified predictor of outcome is specific to the combination treatment or simply
a predictor for the extra component added to the combination. Unfortunately, very few studies
have compared combination treatment individually against each modality. A second approach for
identifying individuals who need combination treatment may be derived from prognostic factors
of poor outcome in head-to-head psychotherapy versus ADM single-modality trials; factors that
are associated with poor outcome to both monotherapies may indicate the need for more intensive,
combined treatment.
The best-replicated clinical factor predictive of outcome is MDD chronicity, which predicts
poorer response across treatment modalities (Blom et al. 2007, Fournier et al. 2009, Joyce et al.
2002, Sotsky et al. 1991, Thase et al. 1994, Trivedi et al. 2006). Whether comparing combination
against ADM or psychotherapy alone, study results consistently indicate that patients with more
chronic MDD are most likely to receive enhanced acute benefits from combination treatment
(Friedman et al. 2004, von Wolff et al. 2012). A recent meta-analysis limited to patients with
chronic depression (including dysthymia and chronic MDD) found a benefit ratio (i.e., a risk ratio
for positive end points) for remission of 1.25 (0.97–1.61) for combination therapy over ADM
monotherapy (von Wolff et al. 2012). However, by combining both dysthymia and chronic MDD

www.annualreviews.org • Combination Treatment of MDD 285

 PS65CH11-Craighead ARI 31 October 2013 12:37

into one meta-analysis, this benefit ratio likely underestimated the value of combined treatment
for chronic MDD. Several large studies of combination therapy versus ADM alone for dysthymia
have failed to find benefit for combination treatment (Browne et al. 2002, de Mello et al. 2001,
Markowitz et al. 2005, Ravindran et al. 1999). Indeed, von Wolff and colleagues (2012) found
that for depressive symptoms in patients with dysthymia, the benefit ratio of combined treatment
versus ADM alone was 0.95. In contrast, for studies limited to chronic MDD, the benefit ratio
was 4.0 in favor of combination treatment.
There is a little evidence indicating that more severely depressed patients (HDRS >20) benefit
more from acute combination treatment than from ADM monotherapy (Cuijpers et al. 2009).
Gender has also not consistently predicted outcomes to different treatment modalities. Recent
adverse life events have been associated with superior outcomes for CBT over ADM (Fournier
et al. 2009) and to poorer response in an ADM-only trial (Mazure et al. 2000).
Patient preference is often used to match patients to treatment, although in clinical trials the
value of preference in predicting outcome is mixed. Some randomized trials of psychotherapy
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

versus ADM have required patients, including those with personal treatment preferences, to be
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

willing to start randomized treatment; these studies have found that preference is not associated
with outcome (Dunlop et al. 2012c, Kwan et al. 2010, Leykin et al. 2007). Attrition rates between
patients matched and mismatched by treatment preference are variable across trials; mismatch
predicts greater dropout rates in some trials (Kwan et al. 2010, Raue et al. 2009), but not others
(Dunlop et al. 2012c, Kocsis et al. 2009, Leykin et al. 2007). These variable findings indicate the
crucial contribution of how the informed consent process is performed, particularly the discussion
around randomization and the patient’s willingness to start the treatment regardless of preference.
More generally, patients preferring ADM over psychotherapy or combination treatment have
higher rates of attrition (Dunlop et al. 2012c, Steidtmann et al. 2012), which may stem from
beliefs about causes of their depression (Dunlop et al. 2012c, Steidtmann et al. 2012) or from
practical factors related to treatment (e.g., preference for ADM may derive from time or travel
constraints, which may contribute to attrition).
Two large combination trials of treatments for chronic MDD have evaluated patient preference
as a predictor (Kocsis et al. 2009, Steidtmann et al. 2012). In both trials, the majority of patients
entering the trial preferred combination therapy rather than monotherapy treatment. In the nefa-
zodone/CBASP trial, matching to treatment preference strongly predicted remission among the
few patients who preferred a monotherapy over combination (Kocsis et al. 2009). In the REVAMP
trial, patients indicated their preference for combination therapy or a monotherapy at the begin-
ning of the initial medication-only phase; this preference did not associate with outcomes among
the ADM nonremitters randomized to a 12-week course of combination treatment (CBASP or
brief supportive therapy plus ADM) or ADM adjustment only (Steidtmann et al. 2012). These
differing results speak to the importance of how trial design factors may influence the sample of
MDD patients included in trials.
Childhood maltreatment is associated with poorer outcomes to treatment with medication or
combined treatment for MDD (Nanni et al. 2012). Among chronically ill patients, patients re-
porting early childhood adversity, including parental loss, physical or sexual abuse, and neglect,
improved significantly more with combination CBASP plus nefazodone over nefazodone alone,
though the combination therapy was not superior to CBASP alone (Nemeroff et al. 2003), in-
dicating that ADM did not add to the treatment outcomes. Childhood adversity also predicted
poorer outcomes to the initial medication monotherapy phase in the REVAMP trial (Klein et al.
2009). Again, these results need further study by enrolling patients on the basis of early abuse and
then randomizing them to treatments to determine if abuse history is an actual specific marker of
treatment outcomes.

286 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

Higher levels of anxiety (or the presence of a comorbid anxiety disorder) typically predict
poorer outcomes of treatment for MDD (Fava et al. 2008, Souery et al. 2007) or longer time to
remission (Frank et al. 2011). Studies of combination-therapy versus monotherapy treatments have
not reported differential outcomes by type of treatment among anxious patients. However, among
patients with a comorbid anxiety disorder, anxiety symptoms improved considerably more among
chronically depressed patients who received combination nefazodone and CBASP compared with
those who received CBASP alone; of note, the combination was not superior to nefazodone alone
for anxiety symptoms (Ninan et al. 2002). Some studies comparing psychotherapy versus ADM
or their combination allow the concurrent use of benzodiazepines to treat insomnia or anxiety
(Frank et al. 2011), which acts to reduce the likelihood of finding differences between treatment
conditions.
Because greater medical disease burden is a poor prognostic factor for MDD treatment
(Iosifescu et al. 2003), combination treatment may have particular value in depressed medically
ill patients. However, in the single large controlled trial of combination treatment performed in
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

patients with coronary artery disease and MDD, citalopram plus IPT was not superior to citalo-
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

pram plus clinical management; both arms achieved remission rates of 36% (Lesperance et al.
2007).

Brain-based moderators. Although chronicity is currently the best clinical marker to identify
which patients need combination treatment to remit from a major depressive episode, many pa-
tients in a chronic or recurrent episode achieve remission with a single-modality treatment. Several
methods are being investigated to better stratify patients into subgroups to identify the individuals
who will need ADM, psychotherapy, or combination treatment to achieve remission (Dunlop et al.
2012a). Neuroimaging, including positron emission tomography (PET) and functional magnetic
resonance imaging (fMRI), is at the forefront of these methods, but other methods are also under
investigation (Kemp et al. 2008).
ADM and psychotherapy have been shown to act via different mechanisms in the brain to
achieve their treatment effects (DeRubeis et al. 2005b). A wide variety of studies employing neu-
roimaging have been performed in efforts to determine the neurological basis for change with
single-modality treatments, but no studies have specifically examined the effects of combined
treatment versus those of a monotherapy. Nevertheless, understanding the biological mecha-
nisms by which different single-modality treatments work may contribute to the rational use of
combination treatment in the future.
Neuroimaging has greatly enhanced our understanding of brain function in depressed patients
(Price & Drevets 2011). Symptoms of depression such as changes in appetite, sleep, and energy
are likely mediated in part through hypothalamic regions. Negative emotional experience (sadness
and anxiety) is thought to involve the amygdala, insula, and subgenual cingulate cortex. Anhedonia
and motivation are mediated largely by processing between the ventral tegmentum and ventral
striatum regions, and the basal ganglia are important for psychomotor speed. Cognitive functions
such as attention, working memory, and decision making are performed via processing in prefrontal
regions, including the dorsolateral prefrontal cortex and anterior cingulate cortex. These various
brain regions are connected within networks of activity, and the regions reflect nodes within a
network. However, a complete delineation of the scope and function of these networks has not yet
been described. One emerging conceptualization proposes that in a depressive episode excessive
negative emotion processing in the limbic regions may impact other nodes in the network, driving
cognition toward negative thoughts (pessimism, guilt, worthlessness) and disrupting attentional
systems (Wang et al. 2012); nevertheless, the directionality of network activation is not currently
clearly articulated.

www.annualreviews.org • Combination Treatment of MDD 287

 PS65CH11-Craighead ARI 31 October 2013 12:37

A commonly referenced heuristic for understanding how psychotherapy and ADM differen-
tially impact these networks in MDD is the top-down versus bottom-up distinction (Ochsner
et al. 2009). In this model, ADMs modify neurotransmitter activity to diminish activity in limbic
brain regions, including the amygdala, insula, subgenual cingulate cortex (Broadman Area 25),
and hippocampus, thereby reducing negative emotional states and freeing cortical regions from
negative bias. However, other data indicate that at least some antidepressants may have direct
effects on cortical function, which is not surprising given the diffuse distribution of serotonin and
norepinephrine pathways within the brain (Deutch & Roth 2009, Fu et al. 2004).
Psychotherapy approaches, except perhaps for BA, are proposed to act through top-down
mechanisms, implying that through actions of cognitive networks, excessive negative limbic activity
can be decreased. There are reciprocal connections between limbic regions and cortical regions
such as the anterior cingulate cortex and the orbitofrontal cortex. These regions exist within
networks that reciprocally or unidirectionally transmit information for processing. Thus, mood
may be regulated through acting on these networks at accessible nodes. The nodes are generally
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

considered limbic for medication, whereas they are more cognitive for psychotherapy (see Ochsner
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

et al. 2009 for a more detailed discussion).

Although the networks important for regulating mood have not been completely identified, this
model opens the possibility that neuroimaging may reveal different patterns of abnormal activity
and, thus, indicate the specific need for an ADM or psychotherapeutic intervention. This model
suggests that specific single-modality interventions may ultimately function most effectively
when a specific pattern of neural connectivity has been identified and a treatment that focuses
on that pattern is employed; this approach would maximize the efficacy of that monotherapy.
However, some individuals’ maladaptive patterns may be such that combination interventions
are necessary to correct the dysfunctional neural pathways associated with their MDD. The
goal of improved outcomes through combinations of different forms of treatment complements
the model of bottom-up and top-down mechanisms of action of medication and psychotherapy
treatments. It is important to recognize that although the model may be appealing, biological
evidence of its validity is scant and the model has been explored in only a few, unreplicated
studies.
Most neuroimaging studies of predictors of treatment response to date have explored single
modality forms of treatment. Very few neuroimaging studies have examined changes with ADM
versus psychotherapy treatments (Brody et al. 2001, Goldapple et al. 2004, Kennedy et al. 2007,
Siegle et al. 2012), and even fewer used a randomized design (Kennedy et al. 2007). Outcome
markers in these studies have generally reflected response or remission versus nonresponse to any
treatment (Konarski et al. 2009) rather than differential biomarkers that could guide selection of
one treatment over another.
Most reported findings of outcome prediction need replication in independent samples. Non-
specific predictors of positive outcomes, such as greater resting-state activity in the pre- and
peri-genual anterior cingulate cortex (Mayberg et al. 1997), need to be distinguished from the
specific therapeutic effect of a treatment versus nonspecific placebo effects of time, expectation,
and clinical attention. For CBT, assessment of limbic (amygdala, subgenual cingulate) reactivity to
negative mood challenge, and regulation of negative emotions by dorsolateral, ventrolateral, and
anterior cingulate cortices, may indicate which patients may benefit from CBT treatment (Siegle
et al. 2012, Wager et al. 2008). Some evidence suggests that patients who demonstrate the greatest
differences from healthy controls on emotion-processing tasks may be least likely to benefit from
CBT (Ritchey et al. 2011, Siegle et al. 2012). Significant hypoactivity of the dorsolateral prefrontal
cortex may indicate a need for ADM treatment, potentially representing an inability to activate
nodes necessary for emotion regulation (Fales et al. 2009, Fu et al. 2004, Kennedy et al. 2001).

288 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

Neuroimaging may also be valuable for identifying individuals at risk for recurrent MDD. In
a subset of unmedicated patients with MDD in remission, greater recall for negatively valenced
self-referent words was associated with greater amygdala activation over healthy controls during a
sad mood induction task (Ramel et al. 2007). Similarly, remitted MDD patients who demonstrated
activation of the medial prefrontal cortex (relative to visual cortex activation) while watching sad
film clips had higher rates of major depressive episode recurrence over 18 months of follow-up
(Farb et al. 2011). Approaches such as these may help identify which depressed patients will likely
benefit from maintenance ADM or a specific course of CBT for relapse prevention.
Only recently has neuroimaging been used to predict differential response to psychotherapy
or ADM. In a PET resting-state study, Mayberg and coworkers (McGrath et al. 2013) identified
several brain regions that differentially predicted remission and nonresponse among patients ran-
domly assigned to treatment with CBT or ADM. In particular, hypoactivity in the anterior insula
(relative to whole-brain metabolism) predicted remission with CBT and nonresponse with ADM,
whereas hyperactivity in this region predicted remission to ADM and nonresponse to CBT. The
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

better response to ADM in insula-hyperactive individuals was not absolute; several of these pa-
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

tients also did not respond to ADM. This finding conforms with the results of a meta-analysis of
neuroimaging studies of MDD treatment, in which greater activity in the anterior insula predicted
lower likelihood of response to both ADM and CBT treatments (Fu et al. 2013). That the insula
could serve a predictive role in MDD treatment selection is credible on the basis of its significant
role in several processes that are disrupted in MDD, including interoception, emotional self-
awareness, decision making, cognitive control, and integration of emotionally important social
information (Craig 2009, Critchley 2005, Farb et al. 2012).
Enthusiasm for these neuroimaging findings needs to be tempered with the recognition that,
to date, all biological predictor studies have identified potential prescriptive factors through post
hoc analyses. Proof of their utility in clinical decision making will require prospective random-
ized trials in which patients are stratified, using the potential predictors, into treatment condi-
tions. Until such data are generated, the best approach to matching individual patients to treat-
ments is to compare across studies for factors found to associate with outcomes in retrospective
analyses.

FUTURE DIRECTIONS
Future studies of combination treatments for MDD should prioritize identifying the individuals
who require combined treatment to recover and maintain wellness, rather than documenting the
average effects of combination therapy (i.e., employ a personalized medicine approach to treating
MDD). In light of expense, inconvenience, and insignificant additional gain for many patients,
application of combined treatment needs to be reserved for individuals most at risk of incomplete
benefit from a monotherapy or who possess the biological and psychological characteristics most
predictive of benefit from combination treatment. More precise identification of these factors
should be the primary focus of future research efforts.
Patient treatment preferences, beliefs about depression etiology, and treatment effects and risks
must be evaluated prior to randomization but after completion of the informed consent process;
this process may involve substantial provision of new information for participants. For effectiveness
trials, which aim to mimic real-world clinical practice, all patients should be randomized, regardless
of whether their treatment assignment would affect their intent to complete the trial. For efficacy
trials, or trials examining biological or other moderators of outcomes, patients should be asked
directly by senior research personnel during the consent process whether they are willing to accept
randomization and start treatment, regardless of their preferences; patients who do not agree (i.e.,

www.annualreviews.org • Combination Treatment of MDD 289

 PS65CH11-Craighead ARI 31 October 2013 12:37

plan to drop out if they are not randomized to their preference) should be terminated from the
study prior to randomization.
Previous treatment experiences should be recorded to the degree possible. Studies routinely
exclude patients who have failed to respond to the psychotherapy evaluated in that particular
trial. However, the exclusion should apply to patients who have received that form of treatment,
regardless of prior response. Similarly for medication, trials typically exclude patients who have
not responded to treatment with the ADM used in the trial. However, failure to respond to
one SSRI may predict failure to respond to another, so the exclusion should be based on fail-
ure to respond to the class of medications rather than on failure to respond to the individual
agent.
Placebo control arms are not necessary for trials aiming solely to identify the efficacy of com-
bined treatment. However, such trials should include arms for each of the two individual treatments
comprising the combination treatment in order to minimize selection bias for patients seeking a
particular form of treatment. Studies examining biological or psychological mediators of outcome
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

would benefit from the use of placebo arms to isolate the shared and specific mediators of change
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

in patients who benefit from treatment.

Outcomes should be broadened to include measures beyond symptom reduction. Specifically,
measures of work, family, quality of life, and social function should be included to identify whether
broader effects of combined treatment occur. Within symptom-reduction measures, outcomes can
vary substantially, depending on whether a self-report scale or clinician-rated scale is used. Self-
report measures of depression severity typically show higher rates of remission to treatment than
do clinician-administered scales (Dunlop et al. 2010, 2011; Trivedi et al. 2006). However, clinical
global impression scales may overestimate improvement compared with a patient’s assessment (de
Jonghe et al. 2001, Rush et al. 2005). Future determinants of remission should likely involve a
blending of self- and clinician-rated scales (Uher et al. 2008).

CONCLUSIONS
Single-modality treatments for MDD produce essentially identical results, on average, though
clearly some individuals benefit specifically from one form of treatment and not another. Only
about one-third of monotherapy patients achieve remission of MDD, whereas about two-thirds
show a clinical response. This finding has led to studies regarding the clinical outcomes of com-
bination treatments.
For acute treatment of MDD, there is little evidence that combination therapy is superior to
ADM or psychotherapy alone in mild cases of MDD. Combination treatment for MDD demon-
strates its greatest advantage over single-modality treatment in patients with chronic MDD. Pa-
tients with recurrent forms of MDD likely benefit from a course of CBT-based treatment after
improvement with ADM monotherapy. A simplified treatment algorithm for these recommenda-
tions is displayed in Figure 1.
Combination treatment extends time to recurrence or relapse in MDD patients fully or partially
remitting during acute-phase treatment. For patients who receive combination treatment during
the acute phase, the best long-term outcomes occur if both components of treatment continue
through the maintenance phase. Discontinuing ADM after its use in an acute-phase treatment
carries the greatest risk for recurrence. Patients receiving initial treatment with ADM monotherapy
are less likely to experience a recurrence if they complete a short course of CBT after improving;
this benefit is clear, particularly for patients with recurrent depression or those who continue to
have residual depressive symptoms upon completion of acute treatment.

290 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

Nonchronic MDD Chronic MDD Pure dysthymia

ADM or PT Combination ADM

Remission Nonremission Remission Nonremission

≥3 episodes <3 episodes Maintenance Medication combinations

Alternative treatments
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

MBCT or CBTB
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

Figure 1
Evidence-based treatment algorithm for depressive disorders. For nonchronic MDD, monotherapy
treatment with either ADM or psychotherapy is appropriate, with combination treatment indicated if the
monotherapy treatment fails to produce remission. For chronic MDD, combination treatment is indicated
from the start of treatment. For both forms of MDD, remitting patients with highly recurrent illness
(≥3 lifetime episodes) may benefit from additional sessions of MBCT or CBTB directed toward relapse
prevention. Patients with dysthymia should initially be treated with ADM, progressing to combination
treatments if remission is not attained. Abbreviations: ADM, antidepressant medication; CBTB, cognitive
behavior therapy with booster sessions; MBCT, mindfulness-based cognitive therapy; MDD, major
depressive disorder; PT, psychotherapy.

Great promise is offered by the identification of baseline markers that differentially predict
remission to individual or combination treatments. Further work needs to identify markers of
maintenance of effects achieved during acute treatment.

DISCLOSURE STATEMENT
Dr. Craighead is a board member of Hugarheill ehf, an Icelandic company dedicated to the
prevention of depression, and he receives book royalties from John Wiley & Sons. He is a con-
sultant to the George West Mental Health Foundation that oversees Skyland Trail, a residential
treatment facility in Atlanta, Georgia. Dr. Dunlop has received honoraria for consulting work
performed for Bristol- Myers Squibb, MedAvante, Pfizer, and Roche. He has received research
support from AstraZeneca, Bristol-Myers Squibb, Evotec, Forest, GlaxoSmithKline, Novartis,
Ono Pharmaceuticals, Pfizer, Takeda, and Transcept.

ACKNOWLEDGMENTS
The authors acknowledge the support of NIH grant MH-080880 (W.E.C.) and MH-086690
(B.W.D.) as well as the support of grants (W.E.C.) from the Brock Family Foundation and the
Realan Foundation. The authors express their appreciation to Linda W. Craighead, Jacqueline
Larson, and Allison Meyer for reading earlier versions of this manuscript and to Jacqueline Larson
for assistance in the technical preparation of the manuscript.

www.annualreviews.org • Combination Treatment of MDD 291

 PS65CH11-Craighead ARI 31 October 2013 12:37

LITERATURE CITED
Am. Psychiatr. Assoc. (APA). 2010. Practice Guideline for the Treatment of Patients with Major Depressive Disorder.
Arlington, VA: APA. 3rd ed.
Am. Psychiatr. Assoc. (APA). 2013. Diagnostic and Statistical Manual of Mental Disorders. Washington, DC:
APA. 5th ed.
Angst J, Gamma A, Sellaro R, Lavori PW, Zhang H. 2003. Recurrence of bipolar disorders and major depres-
sion. A life-long perspective. Eur. Arch. Psychiatry Clin. Neurosci. 253:236–40
Arató M, Demeter E, Rihmer Z, Somogyi E. 1988. Retrospective psychiatric assessment of 200 suicides in
Budapest. Acta Psychiatr. Scand. 77(4):454–56
Arroll B, Elley CR, Fishman T, Goodyear-Smith FA, Kenealy T, et al. 2009. Antidepressant versus placebo
for depression in primary care. Cochrane Database Syst. Rev. 3:CD007954
Barlow DH, Gorman JM, Shear MK, Woods SW. 2000. Cognitive-behavioral therapy, imipramine, or their
combination for panic disorder: a randomized controlled trial. JAMA 283:2529–36
Barth J, Paul J, Härter M, Bengel J. 2005. Inpatient psychotherapeutic treatment for cardiac patients with
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

depression in Germany: short-term results. Psychosoc. Med. 2:1–8

Beach SRH, Sandeen EE, O’Leary KD. 1990. Depression in Marriage: A Model for Etiology and Treatment.
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

New York: Guilford

Beck AT, Rush AJ, Shaw BF, Emery G. 1979. Cognitive Therapy of Depression. New York: Guilford
Beck AT, Steer RA, Brown GK. 1996. Manual for the Beck Depression Inventory–II. San Antonio, TX: Psychol.
Corp.
Blackburn I, Eunson K, Bishop S. 1986. A two-year naturalistic follow-up of depressed patients treated with
cognitive therapy, pharmacotherapy and a combination of both. J. Affect. Disord. 10(1):67–75
Blom MB, Spinhoven P, Hoffman T, Jonker K, Hoencamp E, et al. 2007. Severity and duration of depression,
not personality factors, predict short term outcome in the treatment of major depression. J. Affect. Disord.
104:119–26
Bockting CH, Schene AH, Spinhoven P, Koeter MW, Wouters LF, et al. 2005. Preventing relapse/recurrence
in recurrent depression with cognitive therapy: a randomized controlled trial. J. Consult. Clin. Psychol.
73(4):647–57
Bouton ME, Westbrook RF, Corcoran KA, Maren S. 2006. Contextual and temporal modulation of extinction:
behavioral and brain mechanisms. Biol. Psychiatry 60:352–60
Brody AL, Saxena S, Stoessel P, Gillies LA, Fairbanks LA, et al. 2001. Regional brain metabolic changes
in patients with major depression treated with either paroxetine or interpersonal therapy: preliminary
findings. Arch. Gen. Psychiatry 58(7):631–40
Browne G, Steiner M, Roberts J, Gafni A, Byrne C, et al. 2002. Sertraline and/or interpersonal psychotherapy
for patients with dysthymic disorder in primary care: 6-month comparison with longitudinal 2-year
follow-up of effectiveness and costs. J. Affect. Disord. 68:317–30
Cent. Ment. Health, Dep. Health, Mind, NHS Confed. Ment. Health Netw., et al. 2012. No Health Without
Mental Health: Implementation Framework. London: Dep. Health, Natl. Health Serv. https://www.
gov.uk/government/uploads/system/uploads/attachment_data/file/216870/No-Health-Without-
Mental-Health-Implementation-Framework-Report-accessible-version.pdf
Craig DB. 2009. How do you feel–now? The anterior insula and human awareness. Nat. Rev. Neurosci. 10(1):59–
70
Craighead WE, Craighead L, Ritschel L, Zagoloff A. 2013a. Behavior therapy and cognitive-behavioral ther-
apy. In Handbook of Psychology, Vol. 8: Clinical Psychology, pp. 291–319. Hoboken, NJ: Wiley. 2nd ed.
Craighead WE, Miklowitz DJ, Craighead LW, eds. 2013b. Psychopathology: History, Diagnosis, and Empirical
Foundations. New York: Wiley
Craighead WE, Sheets ES, Brosse AL, Ilardi SS. 2007. Psychosocial treatments for major depressive disorder.
In A Guide to Treatments That Work, ed. PE Nathan, JM Gorman, pp. 289–307. New York: Oxford Univ.
Press. 3rd ed.
Critchley HD. 2005. Neural mechanisms of autonomic, affective, and cognitive integration. Comp. Neurol.
493(1):154–66

292 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

Cuijpers P, Dekker J, Hollon SD, Andersson G. 2009a. Adding psychotherapy to pharmacotherapy in the
treatment of depressive disorders in adults: a meta-analysis. J. Clin. Psychiatry 70(9):1219–29
Cuijpers P, Smit F, Bohlmeijer E, Hollon SD, Andersson G. 2010a. Efficacy of cognitive-behavioural therapy
and other psychological treatments for adult depression: meta-analytic study of publication bias. Br. J.
Psychiatry 196(3):173–78
Cuijpers P, van Straten A, Andersson G, van Oppen P. 2008. Psychotherapy for depression in adults: a meta-
analysis of comparative outcome studies. J. Consult. Clin. Psychol. 76(6):909–22
Cuijpers P, van Straten A, Hollon SD, Andersson G. 2010b. The contribution of active medication to combined
treatments of psychotherapy and pharmacotherapy for adult depression: a meta-analysis. Acta Psychiatr.
Scand. 121(6):415–23
Cuijpers P, van Straten A, Warmerdam L, Andersson G. 2009b. Psychotherapy versus the combination of
psychotherapy and pharmacotherapy in the treatment of depression: a meta-analysis. Depress. Anxiety
26(3):279–88
de Jonghe F, Hendriksen M, van Aalst G, Kool S, Peen J, et al. 2004. Psychotherapy alone and combined with
pharmacotherapy in the treatment of depression. Br. J. Psychiatry 185(1):37–45
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

de Jonghe F, Kool S, van Aalst G, Dekker J, Peen J. 2001. Combining psychotherapy and antidepressants in
the treatment of depression. J. Affect. Disord. 64:217–29
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

de Maat S, Dekker J, Schoevers R, de Jonghe F. 2007. Relative efficacy of psychotherapy and combined therapy
in the treatment of depression: a meta-analysis. Eur. Psychiatry 22(1):1–8
de Mello MF, Myczcowisk LM, Menezes PR. 2001. A randomized controlled trial comparing moclobemide
and moclobemide plus interpersonal psychotherapy in the treatment of dysthymic disorder. J. Psychother.
Pract. Res. 10(2):117–23
DeRubeis RJ, Hollon SD, Amsterdam JD, Shelton RC, Young PR, et al. 2005a. Cognitive therapy versus
medications in the treatment of moderate to severe depression. Arch. Gen. Psychiatry 62(4):409–16
DeRubeis RJ, Siegle GJ, Hollon SD. 2005b. Cognitive therapy versus medication for depression: treatment
outcomes and neural mechanisms. Nat. Rev. Neurosci. 9(10):788–96
Deutch AY, Roth RH. 2009. Neurochemical systems in the central nervous system. In Neurobiology of Mental
Illness, ed. DS Charney, EJ Nestler, pp. 12–28. New York: Oxford Univ. Press. 3rd ed.
Dimidjian S, Hollon SD, Dobson KS, Schmaling KB, Kohlenberg RJ, et al. 2006. Randomized trial of behav-
ioral activation, cognitive therapy, and antidepressant medication in the acute treatment of adults with
major depression. J. Consult. Clin. Psychol. 74(4):658–70
Dobson KS, Hollon SD, Dimidjian S, Schmaling KB, Kohlenberg RJ, et al. 2008. Randomized trial of be-
havioral activation, cognitive therapy, and antidepressant medication in the prevention of relapse and
recurrence in major depression. J. Consult. Clin. Psychol. 76:468–77
Driessen E, Cuijpers P, de Maat SCM, Abbass AA, de Jonghe F, Dekker JJM. 2010. The efficacy of short-term
psychodynamic psychotherapy for depression: a meta-analysis. Clin. Psychol. Rev. 30:25–36
Dunkin JJ, Leuchter AF, Cook IA, Kasl-Godley JE, Abrams M, Rosenberg-Thompson S. 2000. Executive
dysfunction predicts nonresponse to fluoxetine in major depression. J. Affect. Disord. 60(1):13–23
Dunlop BW. 2013. Depressive recurrence on antidepressant treatment (DRAT): 4 next step options.
Curr. Psychiatry 12(5):54–55
Dunlop BW, Binder EB, Cubells JF, Goodman MG, Kelley ME, et al. 2012a. Predictors of remission in de-
pression to individual and combined treatments (PReDICT): study protocol for a randomized controlled
trial. Trials 13:106
Dunlop BW, Holland P, Bao W, Ninan PT, Keller MB. 2012b. Recovery and subsequent recurrence in
patients with recurrent major depressive disorder. J. Psychiatr. Res. 46(6):708–15
Dunlop BW, Kelley ME, Mletzko TC, Velasquez CM, Craighead WE, Mayberg HS. 2012c. Depression be-
liefs, treatment preference, and outcomes in a randomized trial for major depressive disorder. J. Psychiatr.
Res. 46(3):375–81
Dunlop BW, Li T, Kornstein SG, Friedman ES, Rothschild AJ, et al. 2010. Correlation between patient and
clinician assessments of depression severity in the PREVENT study. Psychiatry Res. 177:177–83
Dunlop BW, Li T, Kornstein SG, Friedman ES, Rothschild AJ, et al. 2011. Concordance between clinician
and patient ratings as predictors of response, remission and recurrence in major depressive disorder.
J. Psychiatr. Res. 45:96–103

www.annualreviews.org • Combination Treatment of MDD 293

 PS65CH11-Craighead ARI 31 October 2013 12:37

Dunlop BW, Thase ME, Wun CC, Fayyad R, Guico-Pabia CJ, et al. 2012d. A meta-analysis of fac-
tors impacting detection of antidepressant efficacy in clinical trials: the importance of academic sites.
Neuropsychopharmacology 37(13):2830–36
Dunlop BW, Vaughan CL. 2010. Survey of investigators’ opinions on the acceptability of interactions with
patients participating in clinical trials. J. Clin. Psychopharmacol. 30:323–27
Duval S, Tweedie R. 2000. Trim and fill: a simple funnel-plot-based method of testing and adjusting for
publication bias in meta-analysis. Biometrics 56:455–63
Egede LE. 2006. Disease-focused or integrated treatment: diabetes and depression. Med. Clin. North Am.
90(4):627–46
Encinas JM, Vaahtokari A, Enikolopov G. 2006. Fluoxetine targets early progenitor cells in the adult brain.
Proc. Natl. Acad. Sci. USA 103:8233–38
Evans MD, Hollon SD, DeRubeis RJ, Piasecki JM, Grove WM, et al. 1992. Differential relapse following
cognitive therapy and pharmacotherapy for depression. Arch. Gen. Psychiatry 49(10):802–8
Fales CL, Barch DM, Rundle MA, Mintun MA, Mathews J, et al. 2009. Antidepressant treatment normalizes
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

hypoactivity in dorsolateral prefrontal cortex during emotional interference processing major depression.
J. Affect. Disord. 112:206–11
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

Farb NAS, Anderson AK, Bloch RT, Segal ZV. 2011. Mood linked responses in medial prefrontal cortex
predict relapse in patients with recurrent unipolar depression. Biol. Psychiatry 70(4):366–72
Farb NAS, Segal ZV, Anderson AK. 2012. Attentional modulation of primary interoceptive and exteroceptive
cortices. Cereb. Cortex 23(1):114–26
Fava GA, Grandi S, Zielezny M, Rafanelli C, Canestrari R. 1996. Four-year outcome for cognitive behavioral
treatment of residual symptoms in major depression. Am. J. Psychiatry 153(7):945–47
Fava GA, Rafanelli C, Grandi S, Conti S, Belluardo P. 1998. Prevention of recurrent depression with cognitive
behavioral therapy: preliminary findings. Arch. Gen. Psychiatry 55(9):816–20
Fava M, Rush AJ, Alpert JE, Balasubramani GK, Wisniewski SR, et al. 2008. Difference in treatment outcome
in outpatients with anxious versus nonanxious depression: STAR∗ D report. Am. J. Psychiatry 165(3):342–
51
Fergusson DM, Woodward LJ. 2002. Mental health, educational, and social role outcomes of adolescents with
depression. Arch. Gen. Psychiatry 59:225–31
Fournier JC, DeRubeis RJ, Shelton RC, Hollon SD, Amsterdam JD, Gallop R. 2009. Prediction of response
to medication and cognitive therapy in the treatment of moderate to severe depression. J. Consult. Clin.
Psychol. 77(4):775–87
Frank E, Cassano GB, Rucci P, Thompson WK, Kraemer HC, et al. 2011. Predictors and moderators of
time to remission of major depression with interpersonal psychotherapy and SSRI pharmacotherapy.
Psychol. Med. 41(1):151–62
Frank E, Grochocinski VJ, Spanier CA, Buysse DJ, Cherry CR. 2000. Interpersonal psychotherapy and
antidepressant medication: evaluation of a sequential treatment strategy in women with recurrent major
depression. J. Clin. Psychiatry 61(1):51–57
Frank E, Kupfer DJ, Buysse DJ, Swartz HA, Pilkonis PA, et al. 2007. Randomized trial of weekly, twice-
monthly, and monthly interpersonal psychotherapy as maintenance treatment for women with recurrent
depression. Am. J. Psychiatry 164:761–67
Frank E, Kupfer DJ, Perel JM, Cornes C, Jarrett DB, et al. 1990. Three-year outcomes for maintenance
therapies in recurrent depression. Arch. Gen. Psychiatry 47:1093–99
Frank E, Prien RF, Jarrett RB, Keller MB, Kupfer DJ, et al. 1991. Conceptualization and rationale for consensus
definitions of terms in major depressive disorder. Remission, recovery, relapse, and recurrence. Arch. Gen.
Psychiatry 48:851–55
Frank JD, Frank JB. 1991. Persuasion and Healing: A Comparative Study of Psychotherapy. Baltimore, MD: Johns
Hopkins Univ. Press. 3rd ed.
Friedman MA, Detweiler-Bedell JB, Leventhal HE, Horne R, Keitner GI, Miller IW. 2004. Combined
psychotherapy and pharmacotherapy for the treatment of major depressive disorder. Clin. Psychol. Sci.
Pract. 11(1):47–68

294 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

Fu CHY, Steiner H, Costafreda SG. 2013. Predictive neural markers of clinical response in depression: a
meta-analysis of functional and structural neuroimaging studies of pharmacological and psychological
therapies. Neurobiol. Dis. 52:75–83
Fu CHY, Williams SC, Cleare AH, Brammer MJ, Walsh ND et al. 2004. Attenuation of the neural response
to sad faces in major depression by antidepressant treatment: a prospective, event-related functional
magnetic resonance imaging study. Arch. Gen. Psychiatry 61:877–89
Gelenberg AJ, Freeman MP, Markowitz JC, Rosenbaum JF, Thase ME, et al. 2010. Practice Guideline for the
Treatment of Patients with Major Depressive Disorder. Arlington, VA: Am. Psychiatr. Assoc. 3rd ed.
Goldapple K, Segal Z, Garson C, Lau M, Bieling P, et al. 2004. Modulation of cortical-limbic pathways in major
depression: treatment-specific effects of cognitive behavior therapy. Arch. Gen. Psychiatry 61(1):34–41
González HM, Vega WA, Williams DR, Tarraf W, West BT, Neighbors HW. 2010. Depression care in the
United States: too little for too few. Arch. Gen. Psychiatry 67(1):37–46
Greenberg RP. 1999. Common psychosocial factors in psychiatric drug therapy. In The Heart and Soul
of Change: Common Factors in Effective Psychotherapy, Medicine, and Human Services, ed. MA Hubble,
BL Duncan, S Miller, pp. 297–328. Washington, DC: Am. Psychiatr. Assoc.
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

Haaga DA, DeRubeis RJ, Stewart BL, Beck AT. 1991. Relationship of intelligence with cognitive therapy
outcome. Behav. Res. Ther. 29(3):277–81
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

Hamilton M. 1960. A rating scale for depression. J. Neurol. Neurosurg. Psychiatry 23:56–62
Hankin BL, Abramson LY. 1999. Development of gender differences in depression: description and possible
explanations. Ann. Med. 31(6):372–79
Hansen R, Gaynes B, Thieda P, Gartlehner G, Deveaugh-Geiss A, et al. 2008. Meta-analysis of major depres-
sive disorder relapse and recurrence with second-generation antidepressants. Psychiatr. Serv. 59(10):1121–
30
Higgins JPT, Green S, eds. 2011. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0
(updated March 20, 2011). Cochrane Collab. http://www.cochrane-handbook.org
Hirschfeld RM, Russell JM, Delgado PL, Fawcett J, Friedman RA, et al. 1998. Predictors of response to acute
treatment of chronic and double depression with sertraline or imipramine. J. Clin. Psychiatry 59(12):669–
75
Hollon SD, DeRubeis RJ, Seligman MEP. 1992. Cognitive therapy and the prevention of depression. Appl.
Prev. Psychol. 1(2):89–95
Hollon SD, DeRubeis RJ, Shelton RC, Amsterdam JD, Salomon RM, et al. 2005a. Prevention of relapse
following cognitive therapy versus medications in moderate to severe depression. Arch. Gen. Psychiatry
62(4):417–22
Hollon SD, Jarrett RB, Nierenberg AA, Thase ME, Trivedi M, Rush AJ. 2005b. Psychotherapy and medication
in the treatment of adult and geriatric depression: which monotherapy or combined treatment? J. Clin.
Psychiatry 66(4):455–68
Hollon SD, Stewart MO, Strunk D. 2006. Enduring effects for cognitive behavior therapy in the treatment
of depression and anxiety. Annu. Rev. Psychol. 57:285–315
Hughes S, Cohen D. 2009. A systematic review of long-term studies of drug treated and non-drug treated
depression. J. Affect. Disord. 118:9–18
Iosifescu DV, Nierenberg AA, Alpert JE, Smith M, Bitran S, et al. 2003. The impact of medical comorbidity
on acute treatment in major depressive disorder. Am. J. Psychiatry 160(2):2122–27
Jacobson NS, Martell CR, Dimidjian S. 2001. Behavioral activation treatment for depression: returning to
contextual roots. Clin. Psychol. Sci. Pract. 8(3):255–70
Joyce PR, Mulder RT, Luty SE, Sullivan PF, McKenzie JM, et al. 2002. Patterns and predictors of remission,
response and recovery in major depression treated with fluoxetine or nortriptyline. Aust. N. Z. J. Psychiatry
36(3):384–91
Judd LL. 1997. The clinical course of unipolar major depressive disorders. Arch. Gen. Psychiatry 52:989–91
Judd LL, Paulus MJ, Schettler PJ, Akiskal HS, Endicott J, et al. 2000. Does incomplete recovery from first
lifetime major depressive episode herald a chronic course of illness? Am. J. Psychiatry 157(9):1501–4
Karp JF, Buysse DJ, Houck PR, Cherry C, Kupfer DJ, Frank E. 2004. Relationship of variability in residual
symptoms with recurrence of major depressive disorder during maintenance treatment. Am. J. Psychiatry
161:1877–84

www.annualreviews.org • Combination Treatment of MDD 295

 PS65CH11-Craighead ARI 31 October 2013 12:37

Keller MB, Kocsis JH, Thase ME, Gelenberg AJ, Rush AJ, et al. 1998. Maintenance phase efficacy of sertraline
for chronic depression: a randomized controlled trial. JAMA 280:1665–1672
Keller MB, McCullough JP, Klein DN, Arnow B, Dunner DL, et al. 2000. A comparison of nefazodone,
the cognitive behavioral-analysis system of psychotherapy, and their combination for the treatment of
chronic depression. N. Engl. J. Med. 342(20):1462–70
Keller MB, Trivedi MH, Thase ME, Shelton RC, Kornstein SG, et al. 2007. The Prevention of Recurrent
Episodes of Depression with Venlafaxine for Two Years (PREVENT) study: outcomes from the 2-year
and combined maintenance phases. J. Clin. Psychiatry 68(8):1246–56
Kemp AH, Gordon E, Rush AJ, Williams LM. 2008. Improving the prediction of treatment response in
depression: integration of clinical, cognitive, psychophysiological, neuroimaging and genetic measures.
CNS Spectr. 13(12):1066–86
Kennedy SH, Evans KR, Krüger S, Mayberg HS, Meyer JH, et al. 2001. Changes in regional brain glucose
metabolism measured with positron emission tomography after paroxetine treatment of major depression.
Am. J. Psychiatry 158(6):899–905
Kennedy SH, Giacobbe P. 2007. Treatment resistant depression—advances in somatic therapies. Ann. Clin.
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

Psychiatry 19(4):279–87
Kennedy SH, Konarski JZ, Segal ZV, Lau MA, Bieling PJ, et al. 2007. Differences in brain glucose metabolism
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

between responders to CBT and venlafaxine in a 16-week randomized controlled trial. Am. J. Psychiatry
164(5):778–88
Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. 2005. Lifetime prevalence and age-
of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch. Gen.
Psychiatry 62(6):593–602
Khan A, Faucett J, Lichtenberg P, Kirsch I, Brown WA. 2012. A systematic review of comparative efficacy of
treatments and controls for depression. PLoS ONE 7(7):e41778
Klein DF. 1996. Preventing hung juries about therapy studies. J. Consult. Clin. Psychol. 64(1):81–87
Klein DN, Arnow BA, Barkin JL, Dowling F, Kocsis JH, et al. 2009. Early adversity in chronic depression:
clinical correlates and response to pharmacotherapy. Depress. Anxiety 26:701–10
Klein DN, Black SR. 2013. Persistent depressive disorder. Major depression. See Craighead et al. 2013b,
pp. 334–55
Klerman GL, Weissman MM, Rounsaville BJ, Chevron ES. 1984. Interpersonal Psychotherapy of Depression.
New York: Basic Books
Kocsis JH, Leon AC, Markowitz JC, Manber R, Arnow B, et al. 2009. Patient preference as a moderator of
outcome for chronic forms of major depressive disorder treated with nefazodone, cognitive behavioral
analysis system of psychotherapy, or their combination. J. Clin. Psychiatry 70(3):354–61
Kocsis JH, Thase ME, Trivedi MH, Shelton RC, Kornstein SG, et al. 2007. Prevention of recurrent episodes
of depression with venlafaxine ER in a 1-year maintenance phase from the PREVENT Study. J. Clin.
Psychiatry 68(7):1014–23
Konarski JZ, Kennedy SH, Segal ZV, Lau MA, Bieling PJ, et al. 2009. Predictors of nonresponse to cognitive
behavioural therapy or venlafaxine using glucose metabolism in major depressive disorder. J. Psychiatry
Neurosci. 34(3):175–80
Kovacs M, Rush A, Beck AT, Hollon SD. 1981. Depressed outpatients treated with cognitive therapy or
pharmacotherapy: a one-year follow-up. Arch. Gen. Psychiatry 38(1):33–39
Kraus DR, Castonguay L, Boswell JF, Nordberg SS, Hayes JA. 2011. Therapist effectiveness: implications for
accountability and patient care. Psychother. Res. 21(3):267–76
Kwan BM, Dimidjian S, Rizvi SL. 2010. Treatment preference, engagement, and clinical improvement in
pharmacotherapy versus psychotherapy for depression. Behav. Res. Ther. 48(8):799–804
Lesperance F, Frasure-Smith N, Koszycki D, Laliberte M-A, van Zyl LT, et al. 2007. Effects of citalopram and
interpersonal psychotherapy on depression in patients with coronary artery disease. JAMA 297:367–79
Lewinsohn PM, Rohde P, Klein DN, Seeley JR. 1999. Natural course of adolescent major depressive disorder:
I. Continuity into young adulthood. J. Am. Acad. Child Adolesc. Psychiatry 38:56–63
Leykin Y, DeRubeis RJ, Gallop R, Amsterdam JD, Shelton RC, Hollon SD. 2007. The relation of patients’
treatment preferences to outcome in a randomized clinical trial. Behav. Ther. 38(3):209–17

296 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

Lieberman JA, Greenhouse J, Hamer RM, Krishnan KR, Nemeroff CB, et al. 2005. Comparing the effects
of antidepressants: consensus guidelines for evaluating quantitative reviews of antidepressant efficacy.
Neuropsychopharmacology 30(3):445–60
Loeb DF, Ghushchyan V, Huebschmann AG, Lobo IE, Bayliss EA. 2012. Association of treatment modality
for depression and burden of comorbid chronic illness in a nationally representative sample in the United
States. Gen. Hosp. Psychiatry 34(6):588–97
Lorant V, Deliège D, Eaton W, Robert A, Philippot P, Ansseau M. 2003. Socioeconomic inequalities in
depression: a meta-analysis. Am. J. Epidemiol. 157:98–112
Ma SH, Teasdale JD. 2004. Mindfulness-based cognitive therapy for depression: replication and exploration
of differential relapse prevention effects. J. Consult. Clin. Psychol. 72(1):31–40
Maina G, Rosso G, Bogetto F. 2009. Brief dynamic therapy combined with pharmacotherapy in the treatment
of major depressive disorder: long-term results. J. Affect. Disord. 114:200–7
Markowitz JC, Kocsis JH, Bleiberg KL, Christos PJ, Sacks M. 2005. A comparative trial of psychotherapy and
pharmacotherapy for ‘pure’ dysthymic patients. J. Affect. Disord. 89:167–75
Mayberg HS, Brannan SK, Mahurin RK, Jerabek PA, Brickman JS, et al. 1997. Cingulate function in depres-
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

sion: a potential predictor of treatment response. Neuroreport 8:1057–61

Mazure CM, Bruce ML, Maciejewski PK, Jacobs SC. 2000. Adverse life events and cognitive-personality char-
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

acteristics in the prediction of major depression and antidepressant response. Am. J. Psychiatry 157(6):896–
903
McGorry PD, Nelson B, Goldstone S, Yung AR. 2010. Clinical staging: a heuristic and practical strat-
egy for new research and better health and social outcomes for psychotic and related mood disorders.
Can. J. Psychiatry 55(8):486–96
McGrath CL, Kelley ME, Holtzheimer PE, Dunlop BW, Craighead WE, et al. 2013. Toward a neuroimaging
treatment selection biomarker for major depressive disorder. JAMA Psychiatry 70:821–29
Miller IW, Norman WH, Keitner GI. 1989. Cognitive-behavioral treatment of depressed inpatients: six- and
twelve-month follow-up. Am. J. Psychiatry 146:1274–79
Molenaar PJ, Dekker J, Van R, Hendricksen M, Vink A, Schoevers RA. 2007. Does adding psychotherapy to
pharmacotherapy improve social functioning in the treatment of outpatient depression? Depress. Anxiety
24:553–62
Montgomery SA, Åsberg M. 1979. A new depression scale designed to be sensitive to change. Br. J. Psychiatry
134:382–89
Mueller TI, Leon AC, Keller MB, Solomon DA, Endicott J, et al. 1999. Recurrence after recovery from major
depressive disorder during 15 years of observational follow-up. Am. J. Psychiatry 156:1000–6
Murray CJL, Lopez AD. 1997. Global mortality, disability, and the contribution of risk factors: Global Burden
of Disease Study. Lancet 349:1436–42
Nanni V, Uher R, Danese A. 2012. Childhood maltreatment predicts unfavorable course of illness and treat-
ment outcome in depression: a meta-analysis. Am. J. Psychiatry 169(2):141–51
Nemeroff CB, Heim CM, Thase ME, Klein DN, Rush AJ, et al. 2003. Differential responses to psychother-
apy versus pharmacotherapy in patients with chronic forms of major depression and childhood trauma.
Proc. Natl. Acad. Sci. USA 100(24):14293–96
Nezu AM, Nezu CM, D’Zurilla TJ. 2013. Problem-Solving Therapy. New York: Springer
Ninan PT, Rush AJ, Crits-Christoph P, Kornstein SG, Manber R, et al. 2002. Symptomatic and syndromal
anxiety in chronic forms of major depression: effect of nefazodone, cognitive behavioral analysis system
of psychotherapy, and their combination. J. Clin. Psychiatry 63(5):434–41
Ochsner KN, Ray RR, Hughes B, McRae K, Cooper JC, et al. 2009. Bottom-up and top-down processes in
emotion generation: common and distinct neural mechanisms. Psychol. Sci. 20(11):1322–31
Oxman TE, Hegel MT, Hull JG, Dietrich AJ. 2008. Problem-solving treatment and coping styles in primary
care for minor depression. J. Consult. Clin. Psychol. 76(6):933–43
Pampallona S, Bollini P, Tibalbi G, Kupelnick B, Munizza C. 2004. Combined pharmacotherapy and psy-
chological treatment for depression: a systematic review. Arch. Gen. Psychiatry 61(7):714–19
Paul GL. 1969. Behavior modification research: design and tactics. In Behavior Therapy: Appraisal and Status,
ed. CM Franks, pp. 29–62. New York: McGraw-Hill

www.annualreviews.org • Combination Treatment of MDD 297

 PS65CH11-Craighead ARI 31 October 2013 12:37

Paykel ES, Ramana R, Cooper Z, Hayhurst H, Kerr J, Barocka A. 1995. Residual symptoms after partial
remission: an important outcome in depression. Psychol. Med. 25(6):1171–80
Paykel ES, Scott J, Teasdale JD, Johnson AL, Garland A, et al. 1999. Prevention of relapse in residual
depression by cognitive therapy: a controlled trial. Arch. Gen. Psychiatry 56(9):829–35
Perlis RH, Nierenberg AA, Alpert JE, Pava J, Matthews JD, et al. 2002. Effects of adding cognitive therapy to
fluoxetine dose increase on risk of relapse and residual depressive symptoms in continuation treatment
of major depressive disorder. J. Clin. Psychopharmacol. 22(5):474–80
Persons JB. 1993. Outcome of psychotherapy for unipolar depression. In Handbook of Effective Psychotherapy,
ed. TR Giles, pp. 305–23. New York: Plenum
Piet J, Hougaard E. 2011. The effect of mindfulness-based cognitive therapy for prevention of relapse in
recurrent major depressive disorder: a systematic review and meta-analysis. Clin. Psychol. Rev. 31:1032–40
Post RM. 1992. Transduction of psychosocial stress into the neurobiology of recurrent affective disorder.
Am. J. Psychiatry 149:999–1010
Price JL, Drevets W. 2011. Neural circuits underlying the pathophysiology of mood disorders. Trends Cogn.
Sci. 16(1):61–71
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

Ramel W, Goldin PR, Eyler LT, Brown GG, Gotlib IH, McQuaid JR. 2007. Amygdala reactivity and mood-
congruent memory in individuals at risk for depressive relapse. Biol. Psychiatry 61(2):231–39
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

Ramirez CL, Ritschel LA, Mayberg HS. 2013. Adjunctive treatment with behavioral activation to enhance functional
recovery following deep brain stimulation for depression. Poster accepted for presentation at Assoc. Behav.
Cogn. Ther., Nov., Nashville, TN
Raue PJ, Schulberg HC, Heo M, Klimstra S, Bruce ML. 2009. Patients’ depression treatment preferences and
initiation, adherence, and outcome: a randomized primary care study. Psychiatr. Serv. 60(3):337–43
Ravindran AV, Anisman H, Merali Z, Charbonneau Y, Telner J, et al. 1999. Treatment of primary dysthymia
with group cognitive therapy and pharmacotherapy: clinical symptoms and functional impairments.
Am. J. Psychiatry 156(10):1608–17
Reynolds CF III, Frank E, Perel JM, Imber SD, Cornes C, et al. 1999. Nortriptyline and interpersonal
psychotherapy as maintenance therapies for recurrent major depression: a randomized controlled trial in
patients older than 59 years. JAMA 281:39–45
Rich CL, Fowler RC, Fogarty LA, Young D. 1988. San Diego Suicide Study: III. Relationships between
diagnoses and stressors. Arch. Gen. Psychiatry 45(6):589–92
Ritchey M, Dolcos F, Eddington KM, Strauman TJ, Cabeza R. 2011. Neural correlates of emotion processing
in depression: changes with cognitive behavioral therapy and predictors of treatment response. J. Psychiatr.
Res. 45(5):577–87
Ritschel LA, Gillespie CF, Arnarson EO, Craighead WE. 2013. Major depression. See Craighead et al. 2013b,
pp. 285–333
Rohde P, Lewinsohn PM, Klein DN, Seeley JR, Gau JM. 2012. Key characteristics of major depressive disorder
occurring in childhood, adolescence, emerging adulthood, and adulthood. Clin. Psychol. Sci. 1:41–53
Rounsaville BJ, Klerman GL, Weissman MM. 1981. Do psychotherapy and pharmacotherapy for depression
conflict? Empirical evidence from a clinical trial. Arch. Gen. Psychiatry 38:24–29
Rush AJ, Kraemer HC, Sackeim HA, Fava M, Trivedi MH, et al. 2006. Report by the ACNP Task Force on
response and remission in major depressive disorder. Neuropsychopharmacology 31(9):1841–53
Rush AJ, Trivedi MH, Carmody TJ, Ibrahim HM, Markowitz JC, et al. 2005. Self-reported depressive symp-
tom measures: sensitivity to detecting change in a randomized, controlled trial of chronically depressed,
non-psychotic outpatients. Neuropsychopharmacology 30:405–16
Rush AJ, Trivedi MH, Ibrahim HM, Carmody TJ, Arnow B, et al. 2003. The 16-item Quick Inventory
of Depressive Symptomatology (QIDS) Clinician Rating (QIDS-C) and Self- Report (QIDS-SR): a
psychometric evaluation in patients with chronic major depression. Biol. Psychiatry 54:573–83
Schatzberg AF, Rush AJ, Arnow BA, Banks PLC, Blalock JA, et al. 2005. Chronic depression. Medication
(nefazodone) or psychotherapy (CBASP) is effective when the other is not. Arch. Gen. Psychiatry 62(5):513–
20
Schramm E, van Calker D, Dykierek P, Lieb K, Kech S, et al. 2007. An intensive treatment program of
interpersonal psychotherapy plus pharmacotherapy for depressed inpatients: acute and long-term results.
Am. J. Psychiatry 164(5):768–77

298 Craighead · Dunlop

 PS65CH11-Craighead ARI 31 October 2013 12:37

Scott J, Teasdale JD, Paykel ES, Johnson AL, Abbott R, et al. 2000. Effects of cognitive therapy on psychological
symptoms and social functioning in residual depression. Br. J. Psychiatry 177:440–46
Segal ZV, Williams JG, Teasdale JD. 2002. Mindfulness-Based Cognitive Therapy For Depression: A New Approach
To Preventing Relapse. New York: Guilford
Shea MT, Elkin I, Imber SD, Sotsky SM, Watkins JT, et al. 1992. Course of depressive symptoms over
follow-up. Findings from the National Institute of Mental Health Treatment of Depression Collaborative
Research Program. Arch. Gen. Psychiatry 49(10):782–87
Sheets ES, Craighead LW, Brosse AL, Hauser M, Madsen JW, Craighead WE. 2013. Prevention of recurrence
of major depression among emerging adults by a group cognitive-behavioral/interpersonal intervention.
J. Affect. Disord. 147:425–30
Siegle GJ, Thompson WK, Collier A, Berman SR, Feldmiller J, et al. 2012. Toward clinically useful neu-
roimaging in depression treatment. Arch. Gen. Psychiatry 69(9):913–24
Simons AD, Murphy GE, Levine JL, Wetzel RD. 1986. Cognitive therapy and pharmacotherapy for depres-
sion: sustained improvement over one year. Arch. Gen. Psychiatry 43:43–48
Sotsky SM, Glass DR, Shea MT, Pilkonis PA, Collins JF, et al. 1991. Patient predictors of response to
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

psychotherapy and pharmacotherapy: findings in the NIMH Treatment of Depression Collaborative

Research Program. Am. J. Psychiatry 148(8):997–1008
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

Souery D, Oswald P, Massat I, Bailer U, Bollen J, et al. 2007. Clinical factors associated with treatment
resistance in major depressive disorder: results from a European multicenter study. J. Clin. Psychiatry
68:1062–70
Spielmans GI, Berman MI, Usitalo AN. 2011. Psychotherapy versus second-generation antidepressants in the
treatment of depression: a meta-analysis. J. Nerv. Ment. Dis. 199(3):142–49
Steidtmann D, Manber R, Arnow BA, Klein DN, Markowitz JC, et al. 2012. Patient treatment preference as a
predictor of response and attrition in treatment for chronic depression. Depress. Anxiety 29(10):896–905
Teasdale JD. 1997. The relationship between cognition and emotion: the mind-in-place in mood disorders.
In Science and Practice of Cognitive Behavior Therapy, ed. DM Clark, CG Fairburn, pp. 67–93. Oxford, UK:
Oxford Univ. Press
Teasdale JD, Segal ZV, Williams JM, Ridgeway VA, Soulsby JM, Lau MA. 2000. Prevention of re-
lapse/recurrence in major depression by mindfulness-based cognitive therapy. J. Consult. Clin. Psychol.
68(4):615–23
Thase ME, Buysse DJ, Frank E, Cherry CR, Cornes CL, et al. 1997. Which depressed patients will respond
to interpersonal psychotherapy? The role of abnormal EEG sleep profiles. Am. J. Psychiatry 154(4):502–9
Thase ME, Reynolds CF, Frank E, Simon AD. 1994. Response to cognitive-behavioral therapy in chronic
depression. J. Psychother. Pract. Res. 3(3):204–14
Thase ME, Simons AD, McGeary J, Cahalane JF, Hughes C, et al. 1992. Relapse after cognitive behavior ther-
apy of depression: potential implications for longer courses of treatment. Am. J. Psychiatry 149(8):1046–52
Trivedi MH, Rush AJ, Wisniewski SR, Nierenberg AA, Warden D, et al. 2006. Evaluation of outcomes with
citalopram for depression using measurement-based care in STAR∗ D: implications for clinical practice.
Am. J. Psychiatry 163(1):28–40
Uher RR, Farmer AA, Maier WW, Rietschel MM, Hauser JJ, et al. 2008. Measuring depression: comparison
and integration of three scales in the GENDEP study. Psychol. Med. 38(2):289–300
Vittengl JR, Clark LA, Dunn TW, Jarrett RB. 2007. Reducing relapse and recurrence in unipolar depression: a
comparative meta-analysis of cognitive-behavioral therapy’s effects. J. Consult. Clin. Psychol. 75(3):475–88
von Wolff A, Hölzel LP, Westphal A, Härter M, Kriston L. 2012. Combination of pharmacotherapy and psy-
chotherapy in the treatment of chronic depression: a systematic review and meta-analysis. BMC Psychiatry
12:61
Wager TD, Barrett LF, Bliss-Moreau E, Lindquist K, Duncan S, et al. 2008. Prefrontal-subcortical pathways
mediating successful emotion regulation. Neuron 59:1037–50
Wang L, Hermens DF, Hickie IB, Lagopoulos J. 2012. A systematic review of resting-state functional-MRI
studies in major depression. J. Affect. Disord. 142:6–12
Waraich P, Goldner EM, Somers JM, Hsu L. 2004. Prevalence and incidence studies of mood disorders: a
systematic review of the literature. Can. J. Psychiatry 49(2):124–38

www.annualreviews.org • Combination Treatment of MDD 299

 PS65CH11-Craighead ARI 31 October 2013 12:37

Wells KB, Stewart A, Hays RD, Burnam MA, Rogers W, et al. 1989. The functioning and well-being of
depressed patients. Results from the Medical Outcomes Study. JAMA 262(7):914–19
World Health Organ. (WHO). 2004. Global Burden of Disease: 2004 update. Geneva: WHO. http://www.
who.int/healthinfo/global_burden_disease/GBD_report_2004update_full.pdf
Zimmerman M, Martinez J, Attiullah N, Friedman M, Toba C, et al. 2012. Further evidence that the cutoff to
define remission on the 17-item Hamilton Depression Rating Scale should be lowered. Depress. Anxiety
29(2):159–65
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

300 Craighead · Dunlop

 PS65-FrontMatter ARI 13 November 2013 20:27

Annual Review of
Contents
Psychology

Volume 65, 2014

Prefatory
I Study What I Stink At: Lessons Learned from a Career in Psychology
Robert J. Sternberg p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1

Stress and Neuroendocrinology

Oxytocin Pathways and the Evolution of Human Behavior
C. Sue Carter p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p17
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

Genetics of Behavior
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

Gene-Environment Interaction
Stephen B. Manuck and Jeanne M. McCaffery p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p41

Cognitive Neuroscience
The Cognitive Neuroscience of Insight
John Kounios and Mark Beeman p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p71

Color Perception
Color Psychology: Effects of Perceiving Color on Psychological
Functioning in Humans
Andrew J. Elliot and Markus A. Maier p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p95

Infancy
Human Infancy. . . and the Rest of the Lifespan
Marc H. Bornstein p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 121

Adolescence and Emerging Adulthood

Bullying in Schools: The Power of Bullies and the Plight of Victims
Jaana Juvonen and Sandra Graham p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 159
Is Adolescence a Sensitive Period for Sociocultural Processing?
Sarah-Jayne Blakemore and Kathryn L. Mills p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 187

Adulthood and Aging

Psychological Research on Retirement
Mo Wang and Junqi Shi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209

Development in the Family

Adoption: Biological and Social Processes Linked to Adaptation
Harold D. Grotevant and Jennifer M. McDermott p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 235

vi
 PS65-FrontMatter ARI 13 November 2013 20:27

Individual Treatment
Combination Psychotherapy and Antidepressant Medication Treatment
for Depression: For Whom, When, and How
W. Edward Craighead and Boadie W. Dunlop p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 267

Adult Clinical Neuropsychology

Sport and Nonsport Etiologies of Mild Traumatic Brain Injury:
Similarities and Differences
Amanda R. Rabinowitz, Xiaoqi Li, and Harvey S. Levin p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 301

Self and Identity

The Psychology of Change: Self-Affirmation and Social
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

Psychological Intervention
Geoffrey L. Cohen and David K. Sherman p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 333
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

Gender
Gender Similarities and Differences
Janet Shibley Hyde p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 373

Altruism and Aggression

Dehumanization and Infrahumanization
Nick Haslam and Steve Loughnan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 399
The Sociocultural Appraisals, Values, and Emotions (SAVE) Framework
of Prosociality: Core Processes from Gene to Meme
Dacher Keltner, Aleksandr Kogan, Paul K. Piff, and Sarina R. Saturn p p p p p p p p p p p p p p p p 425

Small Groups
Deviance and Dissent in Groups
Jolanda Jetten and Matthew J. Hornsey p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 461

Social Neuroscience
Cultural Neuroscience: Biology of the Mind in Cultural Contexts
Heejung S. Kim and Joni Y. Sasaki p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 487

Genes and Personality

A Phenotypic Null Hypothesis for the Genetics of Personality
Eric Turkheimer, Erik Pettersson, and Erin E. Horn p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 515

Environmental Psychology
Environmental Psychology Matters
Robert Gifford p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 541

Contents vii
 PS65-FrontMatter ARI 13 November 2013 20:27

Community Psychology
Socioecological Psychology
Shigehiro Oishi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 581

Subcultures Within Countries

Social Class Culture Cycles: How Three Gateway Contexts Shape Selves
and Fuel Inequality
Nicole M. Stephens Hazel Rose Markus, and L. Taylor Phillips p p p p p p p p p p p p p p p p p p p p p p p p p 611

Organizational Climate/Culture
(Un)Ethical Behavior in Organizations
Linda Klebe Treviño, Niki A. den Nieuwenboer, and Jennifer J. Kish-Gephart p p p p p p p 635
Annu. Rev. Psychol. 2014.65:267-300. Downloaded from www.annualreviews.org

Job/Work Design
Access provided by 179.7.195.19 on 05/08/20. For personal use only.

Beyond Motivation: Job and Work Design for Development, Health,

Ambidexterity, and More
Sharon K. Parker p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 661

Selection and Placement

A Century of Selection
Ann Marie Ryan and Robert E. Ployhart p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 693

Personality and Coping Styles

Personality, Well-Being, and Health
Howard S. Friedman and Margaret L. Kern p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 719

Timely Topics
Properties of the Internal Clock: First- and Second-Order Principles of
Subjective Time
Melissa J. Allman, Sundeep Teki, Timothy D. Griffiths, and Warren H. Meck p p p p p p p p 743

Indexes

Cumulative Index of Contributing Authors, Volumes 55–65 p p p p p p p p p p p p p p p p p p p p p p p p p p p 773

Cumulative Index of Article Titles, Volumes 55–65 p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 778

Errata
An online log of corrections to Annual Review of Psychology articles may be found at
http://psych.AnnualReviews.org/errata.shtml

viii Contents