Psychotherapy for Bipolar Disorder in Adults: A Review of the Evidence

Recurrent affective illnesses characterized by both depression and mania have been described since the time of Hippocrates. The modern era of bipolar disorder, however, began in 1949 with the introduction of lithium. John Cade’s discovery of a chemical compound that effectively and specifically treats bipolar disorder (1) revolutionized not only the management of bipolar disorder but also altered how the illness was viewed by researchers and clinicians. What was once conceptualized as an illness caused by unbalanced humours (2) became recognized as a biologic illness amenable to biochemical intervention. Because many with bipolar disorder responded so well to lithium (at least compared with previous interventions), attention turned toward finding and optimizing pharmacologic treatments for bipolar disorder. Positive experiences with major and minor tranquilizers led further credence to this approach. For the next 30 years, little attention was paid to psychosocial treatments for bipolar disorder (3) as it was considered a “problem solved” (E. Frank, personal communication, 2013).

Although rarely the focus of systematic inquiry, psychotherapy was routinely offered to patients suffering from bipolar disorder during the 20th century. Most of these treatments were based on the prevailing psychotherapeutic paradigm of the era (i.e., psychoanalysis), which relied on transference and development of insight to bring about change. Not surprisingly, psychoanalysis had little to offer manic patients who, by definition, suffer from marked impairments in insight (4). Although psychodynamic psychotherapy played a greater role in the management of bipolar depression and bipolar spectrum disorders, early practitioners of psychotherapy for bipolar disorder concluded that “[w]hereas it appeared to work with the schizophrenic, it was not generally successful with the manic-depressive” (5) (p.158). These treatment failures further reinforced the concept that bipolar disorder was best treated primarily–if not exclusively–with pharmacotherapy.

Toward the end of the 20th century, it became increasingly apparent that medication offered only partial relief from bipolar disorder. Treatment with pharmacologic interventions alone was associated with disappointingly low rates of remission (6, 7), high rates of recurrence (8, 9), residual symptoms (10), and psychosocial impairment (11). Gradually, the field moved from conceptualizing bipolar disorder as a disorder requiring only medication to an illness that, like many chronic disorders, is best treated using a combination of pharmacotherapy and psychotherapy (12, 13).

At face value, psychotherapy for bipolar disorder makes a lot of sense. Bipolar disorder is characterized by a high degree of psychosocial impairment (11, 14–16), low rates of medication adherence (17–19), interpersonal dysfunction (11, 20–22), and cognitive impairment (23–25). Each of these domains is reasonably addressed by psychotherapeutic interventions—especially when delivered in combination with pharmacotherapy. Indeed, beginning in the 1990s, a series of clinical trials demonstrating the efficacy and effectiveness of bipolar-specific psychotherapies for the treatment of bipolar disorder appeared in the literature. Unlike the psychodynamic therapies of the previous decades that focused on intrapsychic conflicts and acquisition of insight, contemporary bipolar-specific psychotherapies utilize more directive and symptom-focused strategies such as encouragement of medication adherence, provision of psychoeducation, involvement of family members, development of strategies for relapse prevention, exploration of the reciprocal relationship between mood and either cognitions or interpersonal relationships, and establishment of regular sleep-wake cycles. The goal of this review is to examine the evidence supporting psychotherapy as an efficacious approach to treating bipolar disorder and to summarize the strategies for illness management that have been proven to be helpful for individuals with bipolar disorder.

Methods

We conducted a literature search of outcome studies for psychosocial interventions for adults with bipolar disorder published between 1995 and 2013. One study published in 2014 was included in this review because it was published online in 2012 (26). We searched major databases: PubMed (U.S. National Library of Medicine and National Institutes of Health), Psych INFO (American Psychological Association), and OvidSP (Wolters Kluwer). We also examined the bibliographies of major review articles on the topic to search for missing references. Search terms included combinations of bipolar disorder, mania, depression, psychotherapy, adult, and psychosocial. We restricted our review to randomized controlled trials (RCTs), treatments for adults, and manuscripts published in English. We excluded RCTs of interventions directed primarily at caregivers of individuals with bipolar disorder because the focus of this review is psychotherapy for adult patients.

Results

Table 1 summarizes the published RCTs of psychotherapies for the treatment of adults with bipolar disorder by type of psychotherapy. The trials are listed once in Table 1, even if they test psychotherapies from multiple categories. When trials compare psychotherapies of different modalities, it is described under the first category listed that summarizes information about at least one type of psychotherapy included in the trial. (e.g., an RCT comparing individual cognitive-behavioral therapy and group psychoeducation appears under the group psychoeducation heading only). We identified 35 reports of 28 RCTs testing individual or group psychosocial interventions for individuals with bipolar disorder. In all but one trial (27), psychotherapy was administered as an adjunct to pharmacologic treatments. In most cases, the addition of psychotherapy to pharmacotherapy was associated with improved outcomes for bipolar disorder compared with the control condition [but for exceptions, see, Scott et al. (28), Meyer and Hautzinger (29), Parikh et al. (30), Crowe et al. (31), and de Barros Pellegrinelli et al. (32)]. Below we briefly describe individual and group models of psychotherapy tested for bipolar disorder and the evidence supporting their efficacy/effectiveness. (See Table 2 for brief descriptions of each type of psychotherapy.)

Individual Psychoeducation

Individual psychoeducation (PE) typically consists of 6–21 structured sessions that focus on the provision of information about illness etiology, treatments, course and outcomes, strategies to identify prodromes/early warning signs of relapse, and illness-coping strategies (33–35).

Perry and colleagues conducted the first trial of individual PE, compared with treatment as usual (TAU), for individuals with bipolar I and II disorder with two or more relapses in the preceding 12 months (N=69). Those assigned to individual PE had significant reductions in time to manic relapses (log rank 7.04, df=1, p=0.008) and number of manic relapses over 18 months (median difference of 30%). Although the difference did not reach any statistical significance, there was a trend toward fewer depressive episodes over time in the PE group (p=0.05) (34).

Zaretsky and colleagues compared seven sessions of individual PE to seven sessions of PE followed by 13 additional sessions of cognitive-behavioral therapy (CBT) as treatments for euthymic or minimally symptomatic individuals with bipolar disorder. Among the randomized subjects (N=79), those who received CBT plus PE had significantly greater decreases in Hamilton Depression Rating Scale (HAM-D) scores from baseline to posttreatment than those assigned to PE alone (F_1,44=3.87, p=0.055); but there was no difference between groups in the number of hospitalizations or psychosocial functioning (36).

Rea and colleagues compared 21 sessions of individual PE to family-focused therapy (FFT) in 53 patients with bipolar disorder who were recently hospitalized for mania and partially stabilized on pharmacotherapy. Over 2 years, those assigned to FFT had fewer relapses than those in individual PE (χ²=5.04, p<0.05) and fewer hospitalizations (χ²=3.87, p<0.05) but did not differ on likelihood of a first relapse (χ²=0.50, p>0.10) (37).

Collectively, these studies suggest that individual PE alone offers a benefit over TAU, especially as a strategy for preventing manic episodes, but that more intensive interventions such as CBT and FFT confer additional advantages over and above PE alone.

Group Psychoeducation

Colom and Vieta developed a 21-session group PE intervention that focuses on 1) awareness of the disorder (symptoms, classification, etiologies, course, and prognosis), 2) medication adherence (information about classes of medications, alternative therapies, withdrawal syndromes, and risks of nonadherence), 3) avoiding substance abuse, 4) early detection of new episodes (detection of prodromes, warning signs of relapse, and relapse prevention planning), and 5) stress management (regularity of habits and problem-solving) (38).

Colom and colleagues in Barcelona randomized euthymic individuals with bipolar disorder (N=120) to 21 sessions of either the PE group or an unstructured support group. Over a 2-year period, an assignment to group PE was associated with a longer time to any mood episode (log rank=9.3, p<0.003), lower hospitalization rates (Mann-Whitney U=−2.69, p<0.01), and lower cumulative mean number of hospitalizations (U=−2.23, p<0.05) compared with the unstructured support group (39). At a 5-year follow-up, recurrence rates were lower among those assigned to the PE group compared with the unstructured group (log rank=10.0, p<0.01) (40).

Candini and colleagues (41) compared 21 sessions of the PE group, using the Colom and Vieta model, to TAU in a sample of euthymic patients receiving care in Italy. The results from this trial (N=102) compared favorably to the Barcelona trial, with fewer hospitalizations (U=934, p=0.001) and longer time to hospitalization (log rank=13.5, p<0.001) over a 1-year period among those assigned to group PE compared with TAU (41).

In contrast to these earlier trials, de Barros Pellegrinelli and colleagues (32) in Brazil found no advantage of the PE group (administered as a 16- rather than 21-session intervention) compared with TAU over a 1-year period. They found that on the HAM-D, both groups had increased depressive symptoms at endpoint (p=0.014) with no differences between groups. Only one subject was hospitalized, and there were no differences between groups on the number of depressive or manic relapses (32). Although the Brazilian trial failed to confirm the positive findings of the Spanish and Italian trials, their sample size was much smaller (N=55 versus 120 and 102, respectively), raising the possibility of type II error (i.e., failure to reject the null hypothesis because of inadequate sample size).

Parikh and colleagues (30) compared a six-session PE group to 20 sessions of individual cognitive-behavioral therapy (CBT) (N=204) for adults with bipolar disorder who were either euthymic or had minimal symptoms. Over 72 weeks, both groups had significant reductions in mood symptoms as measured by the Longitudinal Interval Follow-Up Evaluation (LIFE) (p<0.01 for both groups) with no significant differences between the groups. The time to manic or depressive recurrence was not significantly different between groups (30).

Torrent and colleagues (42) randomly assigned 239 euthymic individuals with bipolar disorder to a PE group (N=82), functional remediation (FR) (N=77), or TAU (N=80) for 21 weeks. The PE group was based on the model of Colom and Vieta (see above). FR consists of 21 weekly 90-minute group sessions that use written and oral exercises to address neurocognitive issues such as attention, memory, and executive functions as well as functioning in daily routines. The primary outcome measure was global psychosocial functioning, as measured by change in the Functioning Assessment Short Test from baseline to endpoint. Analyses revealed significant functional improvement over the 21 weeks of treatment (last observation carried forward) in all groups and as well as differences between groups (Pillai’s Trace=0.065, F=6.51, df=2, p=0.002). Those assigned to FR had significantly greater improvement in functioning than those assigned to TAU (p=0.001), but fell short of statistical significance when compared with the PE group (p=0.056). Effect sizes within groups showed a large effect for FR (d=0.93), a moderate effect size for PE (d=0.41), and a small effect size for TAU (d=0.22) (42), suggesting that FR was a more potent treatment for impaired functioning than either PE or TAU, but that PE was somewhat helpful.

In many respects group PE is an ideal “base” treatment for bipolar disorder. Groups are more cost-effective and resource-friendly than individual treatments. Knowledge about bipolar disorder is essential to illness management, and a manualized PE group is well-suited to standardized transmission of this information. However, PE, although necessary for individuals with bipolar disorder, may not be sufficient. Indeed, these studies demonstrate that treatment with a PE group (both 21- and six-session formats) confer benefits for those with bipolar disorder including longer time to recurrence, decreased rates of hospitalization, and improved symptoms over time. The advantages are less apparent, however, when a PE group is compared with a more active comparator than TAU. For instance, outcomes with six-session PE did not differ from 20-session individual CBT. Similarly, both 21-session PE and FR groups were associated with improvement in global functioning, although those assigned to FR fared even better than those assigned to PE. These studies raise the possibility that a stepped care approach to bipolar disorder may be indicated, i.e., treat patients with the less costly/burdensome group PE prior to adding CBT or functional remediation for those who do not achieve an adequate benefit with PE alone.

Individual Cognitive or Cognitive-Behavioral Therapy

Cognitive-behavioral therapy (CBT) and cognitive therapy (CT) for bipolar disorder build on Beck's CT for depression (43), a skills-based treatment that helps individuals recognize and modify the link between maladaptive thoughts and moods. Through the use of thought records, mood diaries, and activity scheduling, patients learn to modify automatic negative thoughts, remove distorted thinking, and interrupt cycles of mania and depression. As described by Lam and colleagues, CT for bipolar disorder adds additional modules of psychoeducation, strategies for coping with prodromes, activities for regulating sleep and routines, and approaches to managing long-term sequelae of the illness (44). Several groups have elaborated variants of CBT and CT for working with bipolar disorder including Basco and Rush (45), Scott (46), Otto et al. (47), and Lam et al. (48).

Lam and colleagues randomized 103 patients with histories of frequent relapses (at least two mood episodes in the prior 2 years or three episodes in the prior 5 years) to individual CT or TAU. CT was administered as 12–18 sessions over 6 months followed by two booster sessions. Over 12 months, the risk for relapse was significantly lower in the CT group compared with TAU (0.40 with 95% confidence interval 0.21–0.74; p=0.004) (49). Over 30 months, those assigned to CT also had lower rates of relapse compared with TAU (64% versus 84%) and a longer time to depressive relapse than TAU (hazard ratio=0.38 with 95% confidence interval=0.19–0.75; p<0.006); however, the difference was not significant for manic/hypomanic episodes (50). The addition of CT to TAU was deemed to be cost-effective because of associated reductions in service use (51).

Ball and colleagues (52) randomized 52 patients with histories of frequent relapses (at least one mood episode over the prior 18 months) to individual CT or TAU. After 6 months, those assigned to CT had significantly greater improvement on the Beck Depression Inventory (BDI) than those assigned to TAU (t=2.71, p=0.009). In regression models, there was a significant time effect for BDI over 12 months (F=5.14, df=7.3, p<0.00) but there were no time × group interactions, meaning that both groups’ depression symptoms improved but there were no differences between groups. There were no significant differences in groups on time to relapse, although there was a trend toward a difference in time to depressive relapse with those assigned to CT remaining well longer (hazard ratio=0.38, 95% confidence interval=0.14 to 1.03; p=0.057) (52).

Scott and colleagues (28) randomly assigned 253 individuals with bipolar disorder to 22 sessions of individual CBT or TAU in a multisite pragmatic trial. There were no significant differences in groups on time to relapse, although among those with fewer than 12 previous episodes, assignment to CBT was associated with significantly fewer relapses than TAU (p=0.04) (28).

The systematic treatment enhancement program for bipolar disorder (STEP-BD), is a multisite trial, compared the effects of three intensive psychosocial treatments—CBT, Interpersonal and Social Rhythm Therapy (IPSRT), or family-focused therapy (FFT)—with a brief (three session) psychoeducational control condition (CC) for the acute treatment of bipolar depression (N=293). Patients who received one of the three intensive psychotherapies had significantly higher rates of recovery (64% versus 52%) and recovered more rapidly (median=113 days ±78.2 versus 146 days ±80.0) than patients in the control condition, after controlling for site, family involvement, and bipolar disorder I versus bipolar disorder II (53). Compared with controls, individuals receiving intensive psychotherapy showed a significantly greater improvement in functioning as well as symptoms over time (54).

Meyer and Hautzinger (29) randomly assigned 76 individuals with bipolar disorder to 20 sessions of either CBT or supportive psychotherapy (ST). According to the investigators, “both treatment conditions included information (e.g., symptoms, etiology, and medication) and mood monitoring in the form of a mood diary. The difference between therapies was such that in the ST condition, therapists adopted a client-centerd focus, meaning that whatever problems the patient presented were dealt with by providing emotional support and general advice. Participants were followed for 33 months, and there were no differences between groups in relapse rates or mood symptoms over time (29). (See also trials by Zaretsky et al. (36) and Parikh et al. (30) described above.)

Individual CBT or CT has been evaluated in seven acute and maintenance trials (see Table 1). In general, CBT out-performs TAU but is comparable to more intensive psychotherapies. When compared with TAU (49, 52), CBT/CT was associated with lower relapse rates and lower depression severity over time. When tested against a very brief control condition (“collaborative care,” which consisted of three sessions of PE) in STEP-BD, CBT was again shown to be the “better” treatment on measures of symptoms (55) and functioning (54). However, CBT appears to have comparable effects to other manualized therapies of adequate duration. Meyer and Hautzinger compared CBT to supportive psychotherapy, finding no differences in relapse rates or mood symptoms over 2-year follow-up (29). Similarly, as discussed above, Parikh and colleagues found no differences between CBT and six-session group PE (30). And, in STEP-BD, the active therapies did not separate from one another (55). An exception to this observation, Zaretsky and colleagues found that CBT plus PE was associated with fewer days depressed and fewer medication changes over 1 year than PE alone (36). Thus, CBT is clearly better than TAU, but it appears to be equivalent to other bipolar disorder-specific treatments.

The Scott et al. trial (28) raises somewhat different issues. This was an effectiveness study, conducted in routine practice settings. As interventions move from academic to real world settings, they typically appear to lose efficacy, resulting in an “efficacy-effectiveness gap” (56). Perhaps not surprisingly, under pragmatic conditions, outcomes with CBT did not differ from TAU—although treatment with CBT was associated with increased time to recurrence among those with fewer than 12 prior mood episodes.

Group Cognitive and Cognitive-Behavioral Therapy

Williams and colleagues compared an eight-session group Mindfulness-Based Cognitive Therapy (MBCT) to wait list control (WLC) for individuals with both bipolar disorder (N=17) and unipolar disorder (N=53) (57). An assignment to MBCT was associated with greater reductions in depressive symptoms than WLC: for BDI scores, there was a main effect for time (F_1,41=6.07; p=0.018) and a significant time × condition interaction (F_1,41=8.05; p=0.007) with those receiving MBCT having greater reductions in scores compared with WLC. On the Beck Anxiety Inventory, there was a significant three-way time × group × condition interaction (F_1,41=7.55; p=0.009) such that those with bipolar disorder who received MBCT had fewer anxiety symptoms posttreatment compared with both WLC and those with unipolar depression (57).

Costa and colleagues (58) randomly assigned euthymic or mildly symptomatic individuals with bipolar disorder to either cognitive-behavioral group therapy (CBGT; N=27) or TAU (N=14), and followed them for 6 months. CBGT consisted of 14 2-hour sessions focusing on both psychoeducation and cognitive and behavioral skills for illness management. An assignment to CBGT was associated with significantly greater declines in depression scores on the BDI compared with TAU (R²=0.909, p=0.002) but no differences in Young Mania Rating Scale (YMRS) scores over time (58). Those who received CBGT also experienced significantly greater improvements in quality of life as measured by the Medical Outcomes Study 36-item Short-Form Health Survey (all subscales except functional capacity) from baseline to week 14 (59).

Gomes and colleagues (60) randomly assigned euthymic individuals with bipolar disorder (N=50) to either 18 sessions of CBGT or TAU. Over 12 months, there were no differences between groups on time until relapse or number of relapses (64% in CBGT versus 56% in TAU). The median time to first relapse was longer for patients treated with CBGT compared with TAU (66 weeks versus 31 weeks; Mann-Whitney=−2.554; p=0.011) (60).

Gonzalez Isasi and colleagues (26) randomly assigned patients with refractory bipolar disorder (frequent relapses, rapid cycling, suicide attempts, or persistent symptoms despite pharmacotherapy) to either 20 one-and-a-half hour sessions of CBGT (N=20) or TAU (N=20) and followed them for 5 years. Those assigned to CBGT had fewer hospitalizations than TAU at the 12-month evaluation (t=2.71, p=0.015). The CBGT group had lower depression and anxiety scores on the BDI and State Trait Anxiety Inventory, respectively, at the 6-month (p=0.006; p=0.019), 12-month (p=0.001; p<0.001), and 5-year (p<0.001, p<0.001) evaluation time points. At 5-year follow-up, 89% of patients in the TAU group and 20% of patients in the CBGT group showed persistent affective symptoms (BDI>7, YMRS>6) and/or difficulties in social-occupational functioning (χ²=18.03, p=0.001) (26).

CBGT, although not as widely tested as individual CBT, has demonstrated efficacy for bipolar disorder. Each of the four studies summarized above showed superior outcomes to TAU on measures of depressive symptoms (26, 60), anxiety symptoms (26, 57), functioning (57, 59), and time to relapse (60). These studies suggest that CBGT is a useful adjunct to pharmacotherapy for the management of bipolar disorder. It will be interesting to see how CBGT fares when tested against more intensive psychotherapies.

Family Therapy

Family-focused therapy (FFT), as described by Miklowitz and colleagues, is a 21-session intervention that is conducted conjointly with a patient and family member (parent, sibling). Treatment focuses on psychoeducation, communication enhancement training, and problem-solving skills training (61). Miklowitz and colleagues compared FFT to three-session crisis management (CM) in 101 adults who had been recently hospitalized for bipolar disorder and partially stabilized on pharmacotherapy. Over a 2-year follow-up period, those assigned to FFT had longer delays prior to relapse (χ²=8.71, p=0.003). A hazard ratio of 0.38 reflects a threefold higher rate of survival without a mood episode in the FFT group over 2 years (52%) than the CM group (17%). Affective (both manic and depressive symptoms) symptoms were derived from the Schedule for Affective Disorders and Schizophrenia–Change Version (SADS-C). A repeated measure mixed model analysis of variance revealed a significant time × treatment interaction favoring FFT on the SADS–C-derived mood symptoms (F_7,549=2.81; p=0.007), showing that those assigned to FFT had greater reductions in mood symptoms than those assigned to CM (61).

Solomon, Miller, and colleagues randomly assigned individuals with bipolar disorder (N=92), currently in a mood episode, to individual family therapy plus pharmacotherapy, multifamily group therapy plus pharmacotherapy, or pharmacotherapy alone. Individual family therapy involved one therapist meeting with a single patient and the patient’s family members, with the content of each session and number of sessions determined by the therapist and family. Multifamily group psychotherapy involved two therapists meeting together for six sessions with multiple patients and their respective family members, with the content of each session preset. Fifty-eight percent of subjects (53/92) recovered from their intake mood episode, and there were no differences between groups on the proportion recovering or the time to recovery. However, there were significant treatment conditions by family impairment interactions (p<0.05): in patients from families with high levels of impairment, the addition of a family intervention (family therapy or multifamily group) resulted in a significantly improved course of illness, particularly with respect to the number of depressive episodes (p<0.01) and the proportion of time spent in a depressive episode (p<0.01). Patients receiving either family intervention had approximately half the number of depressive episodes and spent half as much time depressed as those receiving pharmacotherapy alone (62). Of those suffering a recurrence over 28 months, a significantly smaller proportion of patients assigned to adjunctive multifamily group therapy required hospitalization, compared with those receiving adjunctive individual family therapy or pharmacotherapy alone (χ²=6.53, df=2, p<0.04) (63). (Also see studies of Rea and colleagues (37) and STEP-BD (55) described above.)

For patients with family members who are willing and able to participate in treatment, family therapy is an excellent option. Compared with psychoeducation, FFT hastens recovery (55) and confers additional protection against recurrence (37, 55). Families with greater levels of impairment may derive additional benefit from family therapy when delivered either as individual or multifamily group therapy (63).

Interpersonal and Social Rhythm Therapy

Interpersonal and Social Rhythm Therapy (IPSRT) was developed by Frank and colleagues to help patients with bipolar disorder address interpersonal problems and regulate their social rhythms (64). Social rhythms are those daily activities that help to set (or disrupt) underlying biologic rhythms. Examples of social rhythms include sleep and wake schedules, mealtimes, start time for work/school, and daily exercise. IPSRT postulates that improved regularity of social rhythms will help to entrain/regulate underlying biologic rhythms, thereby addressing the putative link between circadian rhythm disruption and genesis of mood episodes in bipolar disorder (65, 66). IPSRT combines the well-established principles of interpersonal psychotherapy (IPT) for unipolar depression (67) with a behavioral strategy designed to regularize daily routines (social rhythm therapy), and psychoeducation to enhance adherence to medication regimens. IPSRT focuses on: 1) the identification and management of affective symptoms; 2) the link between mood and life events; 3) the maintenance of regular daily rhythms as elucidated by the Social Rhythm Metric (68); 4) the identification and management of potential precipitants of rhythm dysregulation, with special attention to interpersonal triggers; and 5) the facilitation of mourning the lost healthy self (69).

Frank and colleagues (70) evaluated IPSRT as a maintenance treatment for individuals with bipolar disorder (N=175). Acutely ill patients were treated with medication and randomly assigned to either IPSRT or intensive clinical management (CM). Once stabilized [defined as 4 weeks of symptom scores averaging ≤7 on the HAM-D and ≤7 on the Bech-Rafaelsen Mania Scale (70) while on a stable medication regimen], patients were reassigned to either IPSRT or CM (in conjunction with the medication regimen that led to stabilization) for 2 years of monthly maintenance treatment. Those assigned to IPSRT in acute treatment survived significantly longer without a new affective episode during the 2-year maintenance phase than those assigned to CM (p=0.01; hazard ratio=0.35) (71). An assignment to IPSRT was also associated with more rapid improvement in occupational functioning (72).

Swartz and colleagues (27) conducted a small RCT comparing IPSRT (N=14) to quetiapine (N=11), flexibly dosed from 25–300 mg daily, in unmedicated individuals meeting DSM-IV criteria for bipolar disorder II disorder, currently depressed. Over 12 weeks, both groups showed significant declines in the 25-item HAM-D (F_1,21= 44, p<0.0001) and YMRS (F_1,21=20, p=0.0002) scores over time but no differences between groups were detected (27). (Also see description of STEP-BD study (55) above.)

IPSRT has demonstrated efficacy as both an acute (55) and maintenance treatment for bipolar disorder (71). Interestingly, it appears that administering IPSRT during the acute phase of treatment confers the greatest advantage to patients (55). IPSRT also shows promise as monotherapy (i.e., without medication) for bipolar disorder II depression (27). Although IPSRT is typically administered as an individual psychotherapy, it has been implemented in routine practice settings as a group intervention as well (73).

Integrated Care Management

Integrated Care Management (ICM) utilizes the strategies of case management in conjunction with psychotherapy to optimize outcomes for individuals with bipolar disorder. Using a chronic care model of disease management, ICM is defined by Bauer and colleagues as “an organization of care that emphasizes the patient’s development of illness management skills and supports provider capability and availability in order to engage patients in timely, joint decision making about their illness” (74). Thus, ICM encompasses both patient- and provider-level interventions.

Bauer and colleagues randomized 306 individuals recently discharged from the hospital and receiving care at 11 U.S. Department of Veterans Affairs (VA) Hospitals to either TAU or collaborative care. Those assigned to collaborative care participated in weekly group psychoeducation using the Life Goals Program (75), which focuses on the identification of personal symptom profiles, early warning symptoms, and triggers of illness relapse. The patients also had access to evidence-based pharmacotherapy delivered by a psychiatrist with expertise in bipolar disorder and care management from a nurse specialist who coordinated services. Primary clinical outcome was weeks in any episode. Those assigned to collaborative care had 6.2 fewer weeks in any affective episode over 3 years of follow-up compared with TAU (95% CI, −0.3 to −12.5, p<0.05). Over all social role dysfunction as measured by the Social Adjustment Scale decreased significantly more over 3 years in the collaborative care group relative to TAU (p=0.003) (74, 76).

Simon and colleagues (77) tested an integrated care intervention in individuals with bipolar disorder enrolled in four group-model behavioral health clinics of a managed care organization, Group Health Cooperative. Patients hospitalized in the prior 12 months (N=441) were randomly assigned to either ICM or TAU and followed for 1 year. ICM consisted of a group psychoeducation program, monthly telephone monitoring of symptoms and medications, and feedback to the mental health treatment team. Those assigned to ICM had significantly lower mean mania rating scores on the Longitudinal Interval Follow-up Evaluation’s (LIFE) 6-point Psychiatric Status Rating over time (Z=2.24, p=0.025), but there was no difference between groups on change in LIFE mania scores over time. The intervention group showed a significantly greater decline in depression ratings over time compared with TAU (Z=1.98, p=0.048) (77).

Crowe and colleagues (31) conducted a pragmatic trial of nurse-led specialist supportive care (SSC), delivered in a community mental health clinic. SSC is designed to promote and maintain a therapeutic relationship between patient and clinician, improve adherence to pharmacotherapy, and develop problem-solving tools to help them better manage bipolar disorder. It includes psychoeducation and strategies to promote self-esteem and self-efficacy. SSC was delivered weekly for 2 months and then every 2 weeks for seven additional months. Patients with bipolar disorder, currently in a mood episode, were randomly assigned to either SSP (N=36) or TAU (N=42). Forty-two percent of those assigned to SSP (15/36) declined the intervention. Depression and mania symptoms were assessed using the SCL-90, and there were no differences in changes in these measures over time between the two groups. Notably, this study was powered to detect relatively large effect sizes (>0.7), which suggests the possibility of type II error (31).

Large effectiveness trials of integrated care interventions in both private (Group Health Cooperative) and public (VA) were associated with decreased time in episodes (74, 76) and decline in depressive symptoms (77). These studies show that psychosocial treatments delivered in the context of a larger set of interventions designed to improve medication adherence and facilitate communication with care providers leads to reduced illness burden. A smaller trial of SSC was probably underpowered to detect the modest effects that occur in routine practice settings.

Other Group Interventions

Below we briefly describe trials of group interventions that cannot be categorized as one of the types of psychotherapies previously discussed.

Weiss and colleagues (78) compared integrated group therapy (N=31) to group drug counseling (N=31) as adjunctive treatments for individuals with comorbid bipolar disorder and substance dependence. Integrated group therapy consists of 20 weekly sessions utilizing a cognitive-behavioral approach to relapse prevention focusing on the similarities between bipolar disorder and substance dependence. Group drug counseling is also a 20-week program but focuses only on substance use-related problems and goals. Those assigned to integrated group therapy had half as many days of substance use compared with group drug counseling (5.3 [SD=6.6] versus 10.0 [SD=9.1] days/month) with significantly shorter time from treatment initiation to first abstinent month (hazard ratio=2.12, z=2.23, p<0.03). In contrast to the positive effects on substance use, there were no group differences in proportion of weeks in a mood episode, and those receiving integrated group therapy had more depressive and (hypo)manic symptoms than those in group drug counseling (p<0.001) (78). A follow-up study evaluating a shorter (12 session) version of integrated group therapy (N=31) versus group drug counseling (N=30) showed trends favoring integrated group therapy, with greater reductions in substance use during follow-up and a greater decline in risk of mood episodes during treatment (79).

Castle and colleagues (80) randomized euthymic individuals with bipolar disorder (N=84) to either a manualized MAPS group (described in the following sentence) intervention or to TAU. The MAPS acronym reflects the strategies employed in treatment: monitoring and mood activities (M), assessing prodromes (A), preventing relapse (P), and setting specific, measurable, achievable, realistic, time-framed (SMART) goals (S). The MAPS group intervention consists of 12 weekly sessions followed by monthly booster sessions. Over 9 months, there was a significantly decreased risk of relapse in those participating in the MAPS group compared with TAU [log rank χ²(1)=4.31, p=0.04] (80).

Van Dijk and colleagues (81) randomly assigned a small number of individuals with bipolar disorder (N=26) to treatment with a dialectical behavior therapy (DBT)-based PE group or TAU. The group intervention consisted of 12 weekly 90-minute sessions, which taught DBT skills, mindfulness techniques, and general psychoeducation. Individuals assigned to group intervention demonstrated a trend toward reduced depressive symptoms posttreatment (F=3.43, p=0.078), and significant improvement in measures of mindfulness (F=9.41, p=0.006). Furthermore, group attendees had fewer emergency room visits and mental health-related admissions in the 6 months following treatment than those assigned to TAU (t=4.6, p<0.0001) (81).

Weiss and colleagues’ groups show that an integrated treatment group is better than a group focused on addiction only for individuals with bipolar disorder and co-occurring substance use disorders—although the benefit of treatment appears to impact substance use rather than mood symptoms (78, 79). Both MAPS and DBT groups utilize strategies from both PE and CBT. Derived from treatments that have demonstrated efficacy for bipolar disorder, it is not surprising that they also show a favorable impact on bipolar disorder when compared with TAU (80, 81)

Discussion

Psychotherapy, when added to medication for the treatment of bipolar disorder, consistently shows advantages over medication alone as a treatment for bipolar disorder. Whether delivered in a group or individual format, those who receive bipolar disorder-specific psychotherapy fare better than those who do not. Meta-analyses of psychotherapy for bipolar disorder confirm these findings (82, 83). Psychotherapy hastens the recovery from depressive episodes and prevents new mood episodes. It also helps to improve functioning and quality of life. Given the relatively modest risks associated with psychotherapy (i.e., loss of confidentiality) and robust benefits, psychosocial treatments should be considered an important component of bipolar disorder illness management.

We have previously argued that psychotherapies for bipolar disorder may have a more robust impact on depression compared with mania (84). In general, psychotherapy trials show a bigger impact on depressive symptoms than manic symptoms. This may be related to the fact that many bipolar disorder psychotherapies, originally developed for treatment of unipolar depression and later adapted for management of bipolar disorder (for example, CBT and IPT), have a bigger impact on depression. It may also be explained at least in part by the fact that depressive symptoms are much more prevalent than [hypo]manic symptoms (85, 86), and, unless patients are specifically recruited on the basis of elevated mania symptoms, there is likely to be a “ceiling effect” on the change in mania scores. It makes sense, for instance, that in acute depression trials, studies show more of an effect on the depressive than manic pole of the illness. However, even when subjects are recruited in a euthymic state, some studies show more of an impact on depressive symptoms [compare (36)]. Exceptions to this observation include an ICM trial that found reduced time in manic or hypomanic episodes in those receiving ICM and having a high symptom burden at baseline, but no impact on depressive symptoms (87), and an individual PE study finding reductions in the time to mania but not depression (34). These studies suggest the possibility that more intensive interventions targeting more severely ill patients may have preferential effects on mania.

Interestingly, there is considerable overlap among the bipolar disorder-specific psychotherapies. Just as nonspecific factors contribute to a large percentage of the variance in outcomes for all psychotherapies (see Markowitz’ supportive psychotherapy article in the current issue of FOCUS [88] for a discussion of this topic) there are several core strategies that are common to most, if not all, of the efficacious treatments for bipolar disorder. These core strategies include psychoeducation and self-rated mood charts. (See Table 3 for a description of strategies used across psychotherapeutic approaches.)

How then, should a patient decide which bipolar disorder-specific psychotherapy is best for him or her? Although active treatments consistently separate from TAU, most RCTs that compare two evidence-based psychotherapies show little difference between them, suggesting that any of the bipolar disorder-specific psychotherapies will help. Unfortunately, the availability of evidence-based psychotherapy in routine practice settings has not kept pace with the increasing demand for these services (89). Thus, the choice of treatment may be primarily driven by the availability of trained therapists and preference for individual versus group treatment. Stepped psychotherapy care for bipolar disorder, i.e., administering brief interventions that deliver core components of psychotherapy followed by longer treatments of increasing specificity for those who do not benefit fully from shorter treatments, may help the field to allocate efficiently relatively scarce psychotherapy resources, improve outcomes, and ensure that as many people as possible have access to bipolar disorder-specific psychotherapies. Studies of this approach, however, are needed.