Interpersonal Psychotherapy and Cognitive-Behavioral Therapy as an Integrated Treatment Approach for Co-Occurring Bipolar I and Social Anxiety Disorder

1 Theoretical and Research Basis for Treatment

Anxiety disorders commonly co-occur among individuals with bipolar disorder (BD; Merikangas et al., 2007). A recent epidemiological study among 1,600 adults with BD found that 60% of the sample met diagnostic criteria for at least one comorbid lifetime anxiety disorder (Sala et al., 2012), although other studies have found comorbidity rates as high as nearly 75% (Merikangas et al., 2007). The presence of a comorbid anxiety disorder has been associated with more severe symptoms among adults with BD, compared with bipolar patients without comorbid anxiety. These include earlier onset of bipolar symptoms, greater severity of manic and depressive symptoms, increased risk of suicide, higher rate of unemployment, reduced level of close and meaningful relationships, lower quality of life (QOL), and higher prevalence of alcohol and substance use disorders (Albert et al., 2008; Otto & Miklowitz, 2006; Sala et al., 2012; Simon et al., 2004). In addition, the presence of a comorbid anxiety disorder is associated with reduced efficacy of mood-stabilizer medication and poorer patient adherence to psychiatric pharmacotherapy (Perlis et al., 2010). Given the high rates of anxiety comorbidity and associated poorer outcomes among adults with BD, there is a clear need for effective adjunctive treatments for comorbid anxiety in this clinical population.

Social anxiety disorder (SAD), characterized by excessive fear and avoidance of social situations, has one of the highest rates of comorbidity with BD among all of the anxiety disorders. Epidemiological data from the National Comorbidity Survey indicate that approximately 47% to 52% of adults with Bipolar I Disorder (BD-I) meet criteria for co-occurring SAD (Merikangas et al., 2007). These rates are higher than estimated comorbidity rates of BD-I with generalized anxiety disorder (39%-42%), panic disorder (29%-33%), and posttraumatic stress disorder (31%-39%; Merikangas et al., 2007).

The seemingly high rates of SAD in BD-I adults may not be surprising, given that individuals with BD report fewer social interactions and smaller social networks, and are less likely to meet social milestones such as marriage, compared with healthy controls (Bauwens, Pardoen, Staner, Dramaix, & Mendlewicz, 1998; Bauwens et al., 1991; Judd & Akiskal, 2003). Notably, in a study of 164 adults with BD, nearly one third of participants had observer-rated deficits in social competence, with severity of depressive symptoms being the strongest predictor of poorer social functioning (Depp et al., 2010). Therefore, the high comorbidity of SAD in individuals with BD may be due, at least partially, to a vicious cycle in which symptoms of BD result in poorer social competence and functioning, which then promote perceptions of social inadequacy, learned fear of social situations, and social avoidance. In turn, social avoidance reduces one’s social network and ability to improve social functioning, leading to worsening of BD symptoms, thereby maintaining this cycle. Behavioral withdrawal is common to both depressive features of BD and SAD. When the two exist concurrently, symptoms of social withdrawal result from fear of negative social evaluation or interpersonal consequences.

Although there are efficacious psychosocial and pharmacological treatments for SAD, several limitations exist when generalizing these results to individuals with BD. First, the frontline pharmacological treatments for SAD, the selective serotonin reuptake inhibitors (SSRIs), are contraindicated for individuals with BD due to elevated risk for mania (Provencher, Guimond, & Hawke, 2012; Sidor & MacQueen, 2011). Second, although cognitive-behavioral therapy (CBT) and, to a lesser extent, interpersonal psychotherapy (IPT) have shown to be efficacious psychological treatments for SAD, the majority of prior randomized controlled trials have excluded individuals with BD (Heimberg, 2002; Lipsitz, 2012).

Because pharmacotherapy options for SAD may be more limited for individuals with BD, it is desirable to test whether psychological treatments known to be effective for SAD in non-bipolar adults (i.e., CBT, IPT) are also efficacious for adults with BD. Both CBT and Interpersonal and Social Rhythm Therapy (IPSRT), a form of IPT specifically designed for BD, have previously shown to be efficacious adjunctive treatments to pharmacotherapy for BD (Frank et al., 2005). However, investigators have only recently begun to examine effects of psychosocial interventions for BD on secondary anxiety symptoms.

Preliminary research suggests that CBT, whether built into a BD-focused intervention or conducted sequentially (e.g., BD-focused CBT, anxiety-focused CBT), produces moderate effects on comorbid anxiety symptoms (Mueser et al., 2004; Williams et al., 2008). Furthermore, recent analyses found that the presence of a comorbid anxiety disorder was unrelated to treatment response to BD-focused CBT or psychoeducation, and that both CBT and psychoeducation were moderately effective in reducing secondary anxiety symptoms (Hawke, Velyvis, & Parikh, 2013). In turn, comorbid BD did not predict response to group-based CBT for social anxiety, although those with comorbid SAD and BD were found to have greater residual social anxiety symptoms after treatment compared with patients with SAD alone (Fracalanza, McCabe, Taylor, & Antony, 2014).

The presence of a lifetime comorbid anxiety disorder was found to be a treatment moderator in a randomized controlled trial comparing three arms of psychotherapy (CBT, family therapy, and IPSRT) with a three-session, collaborative-care plus pharmacotherapy condition (Deckersbach et al., 2014). Whereas psychotherapy was equivalently effective for individuals with and without a comorbid anxiety disorder, those with comorbid anxiety responded much better to psychotherapy compared with collaborative care, whereas those without comorbid anxiety responded similarly to psychotherapy and collaborative care (Deckersbach et al., 2014). These results suggest that BD individuals with comorbid anxiety may need a higher level of care than those without comorbid anxiety but respond similarly on receiving treatment. Unfortunately, the sample size in this study was insufficient to compare the relative efficacies of the active psychotherapies against one another.

Although preliminary research suggests psychosocial approaches may indeed be efficacious for BD individuals with comorbid anxiety, several limitations exist. First, as noted by Hawke et al. (2013) and with few exceptions (e.g., Fracalanza et al., 2014), several studies failed to use anxiety-specific or anxiety disorder–specific measures in assessing the efficacy of psychological treatments for comorbid anxiety disorders in BD. Therefore, in the case of comorbid SAD, there are few data examining changes in social anxiety symptomatology during treatment. Second, although CBT can reduce secondary anxiety symptoms in BD patients, individuals with BD still evidence more severe social anxiety symptoms post treatment compared with non-BD patients (Fracalanza et al., 2014), suggesting these individuals may benefit from additive treatment.

Acknowledging the interpersonal difficulties BD individuals frequently demonstrate (e.g., Depp et al., 2010), it may be prudent to combine IPT techniques with CBT to enhance social anxiety symptom reduction with these patients. Prior work, in fact, has found an additive benefit for the integration of CBT and IPT in the treatment of other mental disorders. For instance, in a case study of a female patient with bulimia nervosa, Hendricks and Thompson (2005) found that CBT techniques were most helpful in decreasing binging behaviors, whereas IPT techniques further reduced purging behaviors, as these behaviors were often preceded by interpersonal crises. However, to our knowledge, no prior investigation has tested an additive treatment of CBT + IPT for secondary social anxiety symptoms in BD.

The common and unique theoretical perspectives of CBT and IPT offer a complementary treatment approach to comorbid BD and SAD. CBT, with a focus on altering maladaptive cognitions and behaviors, may best address intraindividual factors maintaining social anxiety symptoms. The focus on the interpersonal context of social anxiety symptoms, through the lens of IPT, is well-suited to target maladaptive social and environmental factors that may be maintaining or exacerbating symptoms. This complementary blend of intraindividual and interpersonal factors may offer a broader and thus more efficacious treatment approach than either modality independently.

In the following case study, we present a male client with BD-I who had secondary social anxiety symptoms. Following a baseline assessment, treatment consisted of CBT plus medication management, followed by CBT, IPT, plus medication management. This case featured weekly and bimonthly outcome measures of the client’s social anxiety and depression symptoms, respectively, during treatment and at 1-month and 3-month follow-up phases.

2 Case Introduction

Anthony (a pseudonym) was a 29-year-old White male who presented at a community-based outpatient mental health clinic after being referred for treatment by his former psychiatrist. Anthony was referred to this clinic to reduce his commute to receive treatment services. He had previously been diagnosed with BD-I 1 year earlier. Anthony lived with his girlfriend and young child in a nearby town and worked in the computer software business. He was well-educated, having received a master’s degree from a regional university, appeared to have above-average intelligence, and presented as insightful, friendly, and motivated for treatment.

3 Presenting Complaints

Anthony had a significant history of mania, depression, social anxiety, and substance abuse. Although not endorsing current symptoms of mania, he reported previous manic episodes, the most recent lasting several months and ending approximately 6 months prior to treatment at this clinic. Manic symptoms were assessed through administration of the Mood Disorders Questionnaire (Hirschfeld et al., 2000), a well-validated screening instrument for BD.

When experiencing a manic episode, Anthony endorsed grandiose and racing thoughts, irritability, decreased need for sleep, impulsivity, risk taking, hypersexuality and increased energy, goal-directed activity, and spending. He also endorsed psychotic symptoms that would occur during his manic episodes, including delusions of reference and beliefs that he was able to predict the future, had a psychic connection to world and religious leaders, and was working for a secret government agency.

Anthony reported current symptoms of mild depression and a history of severe depression. At the time of intake, he confirmed low mood, anhedonia, feelings of guilt, irritability, fatigue, difficulty with sleep, and reduced sex drive. Anthony did not report current suicidal ideation but reported a recent hospitalization several months prior for suicidal intent, plan, and means. He had not previously attempted suicide. Anthony’s depression symptoms appeared to have partially remitted from the time of hospitalization to the time of intake at this clinic.

In addition to symptoms of BD, Anthony endorsed symptoms of SAD, including a fear and avoidance of being the center of attention, being criticized by others, public speaking, and interacting with people in unstructured social settings. He indicated a core fear of negative social evaluation from others because they would think him “weird” or “incompetent.” By report, his social anxiety symptoms appeared to have worsened in the preceding 2 years and were currently the primary problem area for which he sought treatment.

4 History

Anthony reported his manic symptoms emerged when he was in his mid-20s and predated his substance use. Anthony reported that his substance usage began approximately 1 year following the development of manic symptoms. Prior to his bipolar symptoms and substance abuse, Anthony had obtained a master’s degree and was successfully employed. He did not endorse any history of trauma or psychiatric problems during childhood or adolescence. He verified multiple hospitalizations for mania, most recently for acute mania 2 years before, as well as one hospitalization for suicidality. He also reported severe opioid and substance abuse that lasted for approximately 2 years. In total, Anthony underwent three hospitalizations for substance detoxification, one for opioid dependence and two for alcohol dependence. At the time of intake, Anthony reported complete sobriety from opioids for nearly 2 years and sobriety from alcohol for nearly 1 year. Notably, his manic and depressive symptoms occurred independent of substance use, as his symptoms predated his substance abuse and continued after sobriety. He was actively involved in Alcoholics Anonymous (AA), where he had a regular sponsor, and attended three meetings per week throughout the duration of treatment. Anthony stated he attended AA meetings 3 times per week for approximately 1 year prior to beginning treatment.

Approximately 1 year ago, Anthony was diagnosed with BD-I and began psychopharmacological treatment at that time. At the time of intake, he was prescribed lithium carbonate (300 mg capsule q.i.d.), lamotrigine (200 mg tablet), Antabuse (250 mg tablet), and risperidone (2 mg tablet). Anthony reported that he had been on these medication dosages for approximately 1 year and endorsed high medication adherence. He indicated further that he had received supportive and psychodynamic psychotherapy at this time, but not CBT or IPT, which was confirmed from a record review of his prior treatment.

5 Assessment

6 Case Conceptualization

Anthony had not experienced manic episodes in the past 6 months, which appeared to be managed well by his mood-stabilizer medication. In addition, he reported sobriety maintenance and there was no evidence of current substance usage. Based on observation and patient self-report, Anthony’s therapist decided to target Anthony’s social anxiety symptoms in treatment, as these appeared to be most functionally impairing. Clinical presentation suggested that Anthony would benefit from CBT to challenge negative thoughts and reduce social avoidance behaviors. Elements of IPT were indicated to assist Anthony with social communication and interpersonal problem solving. The combination of CBT and IPT was intended to improve his anxious cognitions that contributed to avoidance behaviors and to help him develop specific interpersonal communication skills that could improve problematic interpersonal situations. These processes, in turn, were hypothesized to diminish his social anxiety. Specifically, we hypothesized that Anthony would demonstrate significant reductions in self-reported social anxiety and depression symptoms. We expected reductions to be observed in both the CBT and CBT + IPT conditions, with a positive additive effect for the latter condition. In addition to symptom reduction, we hypothesized Anthony would report higher levels of QOL at the end of treatment and at follow-up, compared with a control group of BD patients.

7 Course of Treatment and Assessment of Progress

CBT + IPT

Following eight sessions of CBT, the second additive treatment phase of CBT + IPT was conducted for the remaining 15 sessions. During this phase, Anthony and his therapist continued to review core CBT skills (e.g., cognitive restructuring, exposure therapy), but combined them with IPT techniques. IPT is theoretically based in attachment theory and was originally developed as a treatment for depression, but has since been adapted to treat social anxiety (Lipsitz, 2012). IPT focuses on improving interpersonal functioning, resolving interpersonal conflicts, and increasing social support as methods to promote symptom relief (Weissman, 2006). For Anthony, the emphasis of IPT treatment was his relationship with his girlfriend, but it also included interactions with his father and other family members.

Specific IPT techniques implemented in Anthony’s treatment were communication analysis, review of interpersonal incidents and conflicts, recognition and communication of affect, role-playing, and problem solving. For instance, during several sessions Anthony described recent interpersonal conflicts with his girlfriend, and he and his therapist examined the specific communication patterns involved in the incident, identified Anthony’s contribution to communication difficulties, and specified how he could more clearly communicate and express emotion to his partner. Furthermore, Anthony and the therapist often role-played specific scenarios to practice newly developed skills, notably affect expression and problem solving.

As noted, Anthony participated in baseline assessment, 8 weekly sessions of CBT, and 15 weekly sessions of CBT + IPT. He attended every session, and thus the outcome data at each time point reflect weekly scores. The one exception was a 3-week period between the 18th and 19th time points during which treatment was temporarily suspended due to therapist absence.

Figure 1 shows that Anthony’s baseline scores on the SPIN at baseline were well above the clinical cutoff score and indicative of severe levels of social anxiety symptoms. Although there was some slight variability, his baseline scores remained at similarly high levels. During the first 3 weeks of CBT, Anthony’s scores showed the steepest decline between Sessions 4 and 5, when his SPIN score fell from 42 to 26. This reduction corresponded to the introduction of cognitive restructuring techniques, which Anthony enjoyed, particularly the metaphor of “thinking like a scientist.” He demonstrated good ability to identify his cognitive biases and then apply evidence to evaluate the negative automatic thought. Anthony’s SPIN scores decreased further with exposure therapy, eventually dropping below the clinical cutoff. With the addition of IPT, there was in initial decrease in Anthony’s symptoms; however, his SPIN scores subsequently increased, which coincided with a period of time that included the unexpected death of his mother and loss of his job. With ongoing treatment (CBT + IPT), Anthony improved further by virtue of his SPIN scores remaining below the clinical cutoff. His social anxiety had decreased again at the 1-month and 3-month follow-ups.

OPEN IN VIEWER

Figure 1.

SPIN scores during baseline (BL).

In addition to the data presented in Figure 1, we quantified the difference in Anthony’s SPIN scores at baseline and his final scores at 1-month and 3-month follow-ups. For reference, a 35% reduction in SPIN scores is considered a “minimal” response, 53% as “strong” response, and 63% reduction as “very strong” response (Connor et al., 2000). Anthony’s three baseline data points averaged 43, and his 1-month follow-up score was 11, yielding a 74.42% reduction in SPIN scores. Anthony’s 3-month follow-up corresponded to a 90.70% reduction. In summary, the changes in SPIN scores between the baseline phase and follow-up time points would be classified as a “very strong” clinical response.

Anthony’s baseline and follow-up SPIN scores were also compared with averages observed in a sample of 19 BD patients with comorbid SAD who received group CBT for social anxiety (Fracalanza et al., 2014). His baseline average of 43 was within one standard deviation of the pre-treatment mean of this sample (M = 49.16, SD = 9.27), suggesting the severity of his baseline social anxiety was similar to this patient group, whereas his post-treatment SPIN score of 13 was well below the post-treatment mean (M = 33.79, SD = 12.53).

Anthony’s BDI-II scores are shown in Figure 2. His scores remained below clinical levels at baseline, throughout treatment, and follow-up. These results were consistent with his verbal report of stable mood. The fluctuations in scores across time appeared to be influenced by the periodic personal and financial stressors noted earlier. Anthony’s final BDI-II score at 3-month follow-up is considered well below the clinical range.

OPEN IN VIEWER

Figure 2.

BDI-II scores during baseline (BL).

Figure 3 presents the therapist’s ratings of Anthony’s manic symptoms. To reiterate, scores below 0 indicate “slight decline” (−1) and “decompensation” (−2). Compared with the baseline phase and following two CBT sessions, the therapist’s ratings were consistently 1 or 2, indicating “slight improvement” and “progress” respectively.

OPEN IN VIEWER

Figure 3.

Therapist ratings of mania symptoms during baseline (BL).

Anthony completed the QLI at 1-month and 3-month follow-ups. Per scoring instructions, his satisfaction scores were first recoded by subtracting 3.5 from each item to center the scale at 0. Next, recoded satisfaction scores were multiplied by the raw importance scores for each pair of items. At the 1-month follow-up, Anthony’s total score of 209.50 was divided by the 33 items to create an Overall QOL score of 21.35. Out of a maximum score of 30 for each subscale, Anthony’s highest rated QLI domains were Family QOL (28.80), followed by Psychological/Spiritual QOL (23.29), and Health/Functioning (22.92). Anthony rated his Socioeconomic QOL as significantly lower (12.44) than these other domains. Anthony’s QLI ratings remained stable at 3-month follow-up, with the exception of his Socioeconomic QOL, which increased significantly (27.81) from his 1-month follow-up (12.44), bringing his Overall QOL score higher (26.86).

To facilitate interpretation, Anthony’s scores were compared with mean QLI scores of BD outpatients (N = 37; 51% male) obtained in a prior study (Atkinson, Zibin, & Chuang, 1997). Although Anthony’s self-rated Socioeconomic QOL was slightly lower than this group mean at 1-month follow-up (M = 15.3, SD = 6.2), his other domain scores were significantly higher compared with this group at 1-month follow-up, including Overall QOL (M = 14.9, SD = 5.2), Psychological/Spiritual QOL (M = 14.1, SD = 6.6), Family QOL (M = 18.3, SD = 6.8), and Health/Functioning QOL (M = 14.6, SD = 5.6). Anthony’s QLI scores at 3-month follow-up remained significantly higher than the reference group, including his Socioeconomic QOL. Refer to Figure 4 for a graphical display of these comparisons.

OPEN IN VIEWER

Figure 4.

Comparison of QLI scores with BD reference group (Atkinson et al., 1997).

8 Complicating Factors

There were minimal complicating factors in Anthony’s treatment. He attended all sessions as scheduled, arrived on time, and was compliant with treatment recommendations. In addition, Anthony quickly developed a rapport with his therapist and was open to providing ongoing feedback about the therapy and his progress. He appeared to benefit from the collaborative therapeutic relationship and evidence-based components of CBT. For instance, within the first few sessions, he reported to his therapist that he found the focus on “thinking like a scientist” as highly effective in reducing his anxiety in social situations. He also reported high compliance with homework assignments, a critical component of effective CBT treatment.

9 Access and Barriers to Care

Anthony’s treatment sessions were covered by his health insurance. There were no difficulties qualifying his treatment or other barriers to care.

10 Follow-Up

During the 1-month follow-up, Anthony continued to meet with the therapist weekly. No new treatment skills were introduced at this time. However, previously introduced CBT and IPT skills were reviewed as needed. Given the chronic nature of Anthony’s symptoms, it was not deemed ethical or clinically appropriate to suspend treatment for 1 month. Thus, the reader should place some caution when interpreting 1-month follow-up data, as Anthony continued to receive treatment. Anthony subsequently received a job, relocated out of the clinic area, but completed 3-month follow-up measures and returned data via postal mail. He did not receive psychotherapy between his 1-month and 3-month follow-up assessments although he continued to take prescribed medication at the same dosage during this time.

11 Treatment Implications of the Case

Anthony demonstrated a clinically significant reduction in his social anxiety symptoms during and following treatment. Cognitive restructuring appeared to be particularly efficacious, notwithstanding that his social anxiety symptoms decreased below clinical levels with exposure therapy. This finding supports prior research indicating exposure therapy as a crucial treatment component in CBT for social anxiety (Heimberg, 2002). We note further that Anthony’s social anxiety symptoms continued to improve during the CBT + IPT phase, and with one exception, remained below the clinical cutoff score throughout treatment. Furthermore, he had significantly lower rated social anxiety symptoms than a reference group of BD patients that received group CBT for social anxiety. One implication from these findings is that integrating CBT and IPT techniques could promote greater improvement than CBT alone. However, replication with similar clients is certainly needed to provide greater support. It is encouraging that Anthony continued to do well in treatment despite unanticipated life stressors.

Anthony’s depressive symptoms were below the clinical cutoff at baseline and consistent with his verbal report of mild and partially remitted depressive symptoms. Although fluctuations in his depressed mood were observed, his BDI-II scores remained below clinical levels throughout the duration of treatment. Thus, there was not an observed deterioration in mood-related symptoms despite an improvement in social anxiety symptoms. It is not surprising that Anthony’s depressive symptoms did not appear to respond to treatment because they were not specifically targeted and his symptom presentation was mild at intake. We suggest that a more transdiagnostic approach to targeting anxiety and depressive symptoms (e.g., Barlow et al., 2010) may have elicited change in both anxiety and depression. Indeed, one such transdiagnostic approach, the Unified Protocol (UP), appears promising for BD patients with comorbid anxiety (Ellard et al., 2012).

An important clinical outcome in this case was that Anthony not only demonstrated symptom reduction but also reported relatively high QOL in several domains at the conclusion of treatment and compared with a reference group of other BD outpatients (Atkinson et al., 1997). Although we cannot definitively report Anthony’s QOL improved as a result of his treatment, it is noteworthy that his subjective QOL ratings were significantly higher than outpatients receiving usual care, the specifics of which were not reported. This outcome suggests an additional benefit of the integration of CBT + IPT over usual care in patient QOL. Objective indictors of QOL for Anthony included getting married and receiving a new job offer near the end of treatment. In contrast, only 5% of the reference group was employed full-time, and only 14% were married at the time of the study (Atkinson et al., 1997).,

The significant decrease in Anthony’s social anxiety and depressive symptoms at the 3-month follow-up, as compared with the end of treatment and 1-month follow-up assessments, could be explained by him continuing to use CBT and IPT skills after treatment. In effect, greater practice and refinement of these skills may have resulted in symptom reduction. Also, Anthony relocated to a more desired area and began a new job following the 1-month follow-up. These life circumstances could also have positively affected his social anxiety and depression symptoms. Empirical support for this latter hypothesis was the significant increase in subjective ratings of socioeconomic QOL at 3-months follow-up, as compared with the 1-month follow-up, with relatively stable ratings in other QOL domains. This positive and significant effect of improved financial circumstances suggests that additional treatment services directly targeting finances (e.g., balancing a budget, obtaining financial support services, job coaching) may be a crucial yet underutilized treatment component for BD patients.

There are several limitations to consider when interpreting results. First, caution should be exercised when attempting to generalize findings from a single case. Of particular note, Anthony was a motivated and intelligent client, which may not be representative of other patients receiving treatment services. Replication would be needed prior to making treatment recommendations. Furthermore, Anthony presented with stable mood throughout treatment and was in active recovery from substance abuse. He continued to attend AA meetings 3 times per week throughout treatment. One could argue that this additional structure offered therapeutic benefits, both in terms of ability to practice exposure exercises and provision of social support, which would extend beyond what other patients undergoing this treatment may receive. Finally, Anthony reported high medication adherence throughout treatment. Thus, medication adherence, which was a central part of his treatment and often addressed within the context of CBT for BD, was not emphasized in his case. Patients with BD who are acutely manic and/or abusing substances would thus require mood stabilization and/or substance detoxification treatment prior to receiving this treatment.

Second, this case study did not allow for a direct comparison of CBT versus IPT and it is difficult to conclude which treatment would be more efficacious. Although Anthony’s social anxiety symptoms decreased further during CBT + IPT, this effect may have been a function of time and/or treatment intensity. There is some evidence in non-BD populations that CBT alone is more effective for social anxiety than IPT alone (Dagoo et al., 2014), but this is yet to be tested in BD populations. Future studies could improve on this design by alternating the introduction of IPT and CBT techniques (e.g., IPT and then IPT + CBT) across patients to determine whether a clear pattern of superiority emerges.

Third, we did not measure Anthony’s subjective QOL during baseline. Therefore, conclusions cannot be made about Anthony’s changes in QOL as a function of treatment. Future work could assess QOL at multiple time points during treatment and follow-up to assess whether changes in QOL are seen and whether they correlate with symptom reduction.

Fourth, manic symptoms were not assessed with a standardized instrument, such as the Young Mania Rating Scale (Young, Biggs, Ziegler, & Meyer, 1978), which would have increased confidence in the assessment of Anthony’s manic symptoms over time in addition to clinician ratings through progress notes. Furthermore, ratings conducted by an independent observer, as opposed to the treating therapist, would likely reduce bias in ratings.

Finally, as noted earlier, Anthony continued to receive treatment at the 1-month follow-up. He did not receive psychotherapy between the 1-month and 3-month follow-up assessments; hence, the 3-month outcome could be considered a more valid post-treatment measure. Future work examining changes at longer-term follow-up intervals of 1 to 2 years would be beneficial.

12 Recommendations to Clinicians and Students

We advise students and clinicians to reference the research literature when considering treatment options for individuals with BD and co-occurring anxiety disorders. As cited previously, social anxiety has a high rate of comorbidity among persons with BD and will likely be a focus of treatment. However, trainees and clinicians must be keenly aware of treatments that appear to be most effective for either BD or SAD but contraindicated for the other. Note that in the present case, we selected CBT as a first-line treatment for social phobia, followed by IPT that has been effective with BD. Of course, there may be an interactive effect from these and similar treatments, which must be monitored and carefully evaluated, especially if pharmacotherapy is instituted contemporaneously.

Including multiple measures, as illustrated with Anthony, is another recommendation for assessing treatment effects across more than one clinical domain. For example, we were able to evaluate CBT + IPT on specific symptoms of social anxiety, depression, therapist ratings of mania, and self-reported QOL. Such measurement enables clinicians to confidently proceed with treatment by documenting positive behavior changes without corresponding symptom exacerbation.

One additional recommendation is selecting and implementing treatment methods that have high client appeal and acceptability. In the present case, Anthony commented favorably several times about the “thinking” approach and communicative skill-building operations that were integral to CBT and IPT. He certainly entered treatment with motivation, but we speculate that his compliance within and between sessions, as well as during follow-up, was influenced by a high positive regard for the chief components of treatment.