Innovative Psychological Treatments for Depression

Of the existing treatments for unipolar depression, CBT remains one of the best established. Although CBT has been labeled as a unitary approach, multiple minor modifications of the general model have been developed and evaluated. CBT has been shown to be as efficacious as antidepressant medications, and each of these treatments has been shown to be superior to placebo in the treatment of patients with more severe depressions (7). This finding was confirmed in a recent patient-level meta-analysis involving over 1,700 patients treated in randomized controlled trials (RCTs) comparing the two monotherapies (8). CBT has been criticized by some as lacking long-term follow-up; however, CBT has an enduring effect after treatment termination not found for antidepressant medications (9). There are indications that this enduring effect might even be superior to prior effects of continuing antidepressant medications with patients whose depression remitted (10). Adding CBT appears to produce a modest 10% increment in rates of recovery over medication alone, but this increment is heavily moderated; nonchronic patients with more severe depressions showed a rather large 30% increment when CBT was added, whereas nonchronic patients with less severe depressions did not need this addition, and chronic patients regardless of severity did not benefit from it (incremental benefits were negligible for each of these latter two groups) (11). DeRubeis and colleagues (12) have developed a novel strategy for generating algorithms that can identify the best treatment for a given patient. Application of this personalized advantage index (PAI) to a recent RCT indicated that about a quarter of the patients randomly assigned to each monotherapy (CBT or antidepressant medications) would have improved more if they had received the other treatment and that overall outcomes would have been improved by an amount equal to the advantage of antidepressant medications over placebo if each patient had received optimal treatment (12). The PAI can be applied to any treatment and could be used to improve the efficiency of health care delivery and to increase the power and specificity of tests of mechanism. However, this specialist high-intensity CBT approach is maximized when conducted according to a treatment manual and delivered by trained and competent practitioners who receive ongoing supervision (13). When delivered in such robust ways, CBT results in high efficacy in terms of recovery (14); however, results can be more disappointing when CBT is introduced without a focus on consistently delivered service. There is increasing interest in whether and how CBT can be delivered via low-intensity interventions such as through books (bibliotherapy) and online.

Behavioral activation is a therapy that represents a partial return to the principles of behavioral medicine as developed in the 1960s and 1970s (15), and the model has been extended to include added concepts such as having the patient choose to do activities that he or she values, as well as activities that provide a sense of pleasure, achievement, and closeness to others. The model also emphasizes the importance of engaging in routine/mundane activities such as washing clothes, doing household chores, and performing self-care routines as a way to prevent further problems from developing. The focus is still on behavior more than on cognition, but the model instead emphasizes engagement with potential reinforcers in the environment. Avoidance is a key target for change. The model argues that when people are distressed it is a relief to them not to have to do their usual activities; however, the less people do, the worse they feel, and the worse they feel, the less they do, contributing to a vicious cycle of avoidance, which is the target for change. In a study conducted at its home institution, behavioral activation was found to be as efficacious as antidepressant medications and superior to both CBT and placebo among patients with more severe depressions (16) and as enduring as CBT (with each superior to prior antidepressant medications) following treatment termination (17).

A subsequent study conducted in Iran found behavioral activation superior to antidepressant medications (albeit at about half the maximum dosage) (18), and another in the United Kingdom found it superior to treatment as usual (19). Trials are currently under way comparing behavioral activation to CBT in primary care settings in the United Kingdom (20) and rural India (21), and an as yet unpublished trial indicates that behavioral activation reduced depressive symptoms with no risk of side effects for women given antenatal diagnoses of depression (personal communication, Dimidjian S, November 29, 2015).

Behavioral activation appears to be less complicated to learn than CBT, and if its apparent enduring effect proves to be robust, it may supplant that more established intervention. It is possible to train nonspecialist nurses to deliver behavioral activation (19), and as with CBT, it can be successfully delivered in both high-intensity and low-intensity ways (22, 23). Again, there is an increasing focus on delivering behavioral activation approaches in low-intensity ways via the use of worksheets, books, and online therapies, with high success.

MBCT involves an integration of meditation training with cognitive therapy that is often provided in a group format. MBCT is well established as a preventive intervention and is often provided to patients who have first been brought to remission with antidepressant medications or other interventions (24). MBCT is currently recommended to prevent relapse in people who have had two or more episodes of depression; however, there is a lack of evidence on whether mindfulness approaches can aid in the treatment of milder depression or anxiety. MBCT’s mindfulness component may carry less stigma than conventional psychosocial interventions, and it may prove to be of particular use in reducing risk for pre- and postpartum depression among pregnant women who prefer not to take medications (25). Despite earlier misgivings, recent trials suggest that MBCT also might be efficacious in the treatment of acute depression (26) and that it can be used to prevent the onset of depression among at least some at-risk adolescents (27). However, again there have been difficulties involved in delivering the approach effectively in everyday settings—that is, when it is delivered without clarity of training, consistent delivery, and adequate supervision, and short follow-up times have been an issue for many studies (28, 29).

Acceptance and commitment therapy is one of the newer “third wave” behavior therapies (along with behavioral activation and MBCT) that emphasize context and experiential aspects of psychological experience. It was developed to focus more on complicated long-standing treatment-resistant disorders and has not been as often applied to patients with diagnosed depressions, although change in the symptoms of depression often has been assessed across time. A recent meta-analysis found nearly 40 trials that assessed depression across the course of treatment in samples with a variety of different disorders; in those trials, acceptance and commitment therapy produced large within-group reductions in symptoms and moderate reductions relative to minimal treatment controls (but not active controls) (30). For example, only one study (conducted in Iran) focused on the treatment of patients who met criteria for major depressive disorders, and cell sizes in that trial were quite small (31). Although acceptance and commitment therapy has not been studied as much as the other interventions for diagnosed depression, the available findings indicate it shows promise with respect to efficacy, if not specificity.

Until recent years, IPT had one of the most consistent records of efficacy in controlled trials, demonstrating both efficacy and specificity across a number of trials (32). That may be partly because it has not been all that widely adopted and because most of the controlled trials were thus conducted by advocates expert in its implementation. That situation has changed somewhat in recent years. A trial coming out of New Zealand found that IPT was less efficacious than CBT in the treatment of patients with more severe depressions (33) or personality disorders (34), and another trial done in Canada found IPT less efficacious than antidepressant medications (35). On the other hand, an even more recent trial from the Netherlands found no differences between IPT and CBT (36), with indications of differential response (moderation) on the part of different patients (37). At this time it seems fair to say that IPT remains one of the best established treatments for depression, although it is still not widely practiced outside of certain areas.

Dynamic psychotherapy remains one of the most widely practiced but least evaluated psychosocial interventions. In the past decade there has been an increased emphasis on brief interventions, often on those approaches characterized by adherence to treatment manuals and focused goals. There have been two recent trials worth noting. In the first, the effects of supportive-expressive therapy, a form of brief dynamic psychotherapy, did not differ from those of antidepressant medications (sertraline with nonresponders switched to venlafaxine at midtreatment) or placebo across 16 weeks of treatment (38). Within the larger sample, black men did better in supportive-expressive therapy than they did in either of the two pill conditions, and white women did better in either active intervention than they did in the placebo condition. In a separate trial conducted in the Netherlands, short-term dynamic psychotherapy did not differ from CBT, showing noninferiority to CBT for the continuous measures of depression but not for the somewhat disappointingly low categorical remission rates (22.7% overall) (39). It is not clear what conclusions can be drawn from these studies. On the one hand, dynamic psychotherapy was not inferior to two of the best-established interventions in the field (antidepressant medications and CBT), but on the other hand, in what were essentially null findings, neither study demonstrated either efficacy or specificity for dynamic psychotherapy. If dynamic psychotherapy were a novel medication, the U.S. Food and Drug Administration (FDA) would not have counted either study’s results as a positive indication sufficient for marketing. The first trial did include a nonspecific control condition, but antidepressant medications only exceeded placebo in about half the trials submitted to the FDA to win marketing approval for more recent serotonergic medications (40), and in the second trial, CBT was conducted by a group with no prior track record with that approach. Neither consideration inspires confidence in the generalizability of the comparisons.

In summary, evidence-based interventions typically share three characteristics: They have a clear structure, focus on problems relevant to the person, and build on a relationship with a practitioner. This relationship has traditionally been extensive in terms of time and frequency of sessions (up to one-hour sessions for 12–20 weeks). A typical mindfulness course recommends 26 hours of therapy (29), and over 20 hours of CBT have been recommended for treatment-resistant depression (41). However, can a proportion of people recover equally well with shorter and more focused interventions?

Low-intensity approaches have been developed and tested across an increasing range of disorders, including depression, anxiety, panic, and pain and fatigue. The majority of evaluated interventions to date have used CBT, reflecting its structured approach and psychoeducational skills-based content, which make it especially appropriate for low-intensity delivery. Increasingly, MBCT, acceptance and commitment therapy, and behavioral activation approaches also are becoming available via low-intensity interventions. Other approaches like brief psychodynamic therapy also can be delivered in this way (43). CBT-based resources (both book and online) are more effective at treating depression than are approaches using psychoeducation alone (44).

Low-intensity delivery has at its heart the concept that resource materials deliver key components of the therapy. This might include delivery via print (bibliotherapy) and digital (computerized CBT, or cCBT) media. Both approaches lead to equivalent outcomes (44). Online approaches can provide scalability to encompass a large number of people; also, members of the public are more likely to endorse computer-oriented approaches than book-based approaches, and both are viewed more positively when offered with therapist support (i.e., as guided self-help) than when unsupported (unguided) (45). Some wider issues are also important when choosing which resources to use. Both books and online resources vary significantly in terms of accessibility and readability, and the typical reading ability required for bibliotherapy approaches excludes significant numbers of potential readers (46). It has been suggested that a learning assessment should be conducted in addition to the clinical assessment to identify choice of low-intensity resources (47).

For depression, the efficacy of CBT-based (self-help) resources improves significantly when accompanied by support from a practitioner. It appears this support can be delivered effectively by either experts or nonexperts in CBT, and the focus of the support does not need to include additional therapy components (44). Support contacts play a key role in helping encourage and motivate the patient to use the resources and overcome blocks and low motivation in applying what is learned. It appears there is no difference in outcome when support is delivered face to face or by telephone. Online support (e-mail or chat) is likely to be as efficacious. It is uncertain whether text-message–based chat is as effective, although it may have a role in reminding users to read or test out resources.

An overview of studies suggests that computer-delivered and book-delivered CBTs appear equally effective and that modality selection should be informed by patient choice (44). A review of studies comparing low- and high-intensity interventions found equivalent outcomes in both the short term and the longer term (48). There are many uncertainties, however, and a recent large, well-conducted study with an active control group found no advantage of cCBT (free or licensed) over general practitioner usual care (49). However, that study provided little in the way of active support for the treatment packages, which perhaps explains the result.

A key issue in determining whether low-intensity interventions are working is whether fidelity to the underlying evidence-based model can be attained. Critics question whether such approaches provide only an inexpensive and poorer-quality form of service delivery. This criticism is not borne out by the data or by national treatment guidelines. Second, it is not fully clear whether some subgroups of people do better with more flexible support from a skilled and trained practitioner than from a generic support worker. In practice, at least in the United Kingdom, a majority of accredited CBT (high-intensity) specialists also supplement their work with self-help resources (50). Trials are needed to clarify whether this added expertise and knowledge of the CBT model lead to improved patient outcomes. Provisional data (44) suggest that they do not; however, this review examined the combined results of studies rather than identifying whether specific groups of patients (e.g., those with more complex or chronic presentations) required added input.

Many clinicians view CBT-based resources such as cCBT as being less effective than seeing a specialist practitioner (51). They identify patients who might do well with cCBT as those with milder presenting symptoms. This view is not supported by findings that show larger benefits in clinical samples than in samples of those who are less unwell (44). Also, severity did not predict who failed to engage with or benefit from low-intensity interventions (45). It is also unclear whether age affects outcomes.