Lifetime anxiety disorder and current anxiety symptoms associated with hastened depressive recurrence in bipolar disorder

Journal of Affective Disorders 219 (2017) 165–171

Contents lists available at ScienceDirect

Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad

Research paper

Lifetime anxiety disorder and current anxiety symptoms associated with hastened depressive recurrence in bipolar disorder

MARK

Saloni Shaha, Jane P. Kima, Dong Yeon Parkb, Hyun Kimc, Laura D. Yuena, Dennis Doa, ⁎ Bernardo Dell’Ossoa,d, Farnaz Hooshmanda, Shefali Millera, Po W. Wanga, Terence A. Kettera, a

Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, CA, USA Department of Psychiatry, Seoul National Hospital, Seoul, South Korea c Department of Psychiatry, Ilsan Paik Hospital, Inje University School of Medicine, Goyang, South Korea d Department of Psychiatry, University of Milan; Fondazione IRCCS Ca’Granda, Ospedale Maggiore Policlinico, Milan, Italy b

A R T I C L E I N F O

A B S T R A C T

Keywords: Current anxiety symptoms Lifetime anxiety disorder Bipolar disorder Longitudinal Depression

Aims: To assess differential relationships between lifetime anxiety disorder/current anxiety symptoms and longitudinal depressive severity in bipolar disorder (BD). Methods: Stanford BD Clinic outpatients enrolled during 2000–2011 were assessed with the Systematic Treatment Enhancement Program for BD (STEP-BD) Affective Disorders Evaluation and followed with the STEP-BD Clinical Monitoring Form while receiving naturalistic treatment for up to two years. Baseline unfavorable illness characteristics/current mood symptoms and times to depressive recurrence/recovery were compared in patients with versus without lifetime anxiety disorder/current anxiety symptoms. Results: Among 105 currently recovered patients, lifetime anxiety disorder was significantly associated with 10/ 27 (37.0%) demographic/other unfavorable illness characteristics/current mood symptoms/current psychotropics, hastened depressive recurrence (driven by earlier onset age), and a significantly (> two-fold) higher Kaplan-Meier estimated depressive recurrence rate, whereas current anxiety symptoms were significantly associated with 10/27 (37.0%) demographic/other unfavorable illness characteristics/current mood symptoms/ current psychotropics and hastened depressive recurrence (driven by lifetime anxiety disorder), but only a numerically higher Kaplan-Meier estimated depressive recurrence rate. In contrast, among 153 currently depressed patients, lifetime anxiety disorder/current anxiety symptoms were not significantly associated with time to depressive recovery or depressive recovery rate. Limitations: American tertiary BD clinic referral sample, open naturalistic treatment. Conclusions: Research is needed regarding differential relationships between lifetime anxiety disorder and current anxiety symptoms and hastened/delayed depressive recurrence/recovery – specifically whether lifetime anxiety disorder versus current anxiety symptoms has marginally more robust association with hastened depressive recurrence, and whether both have marginally more robust associations with hastened depressive recurrence versus delayed depressive recovery, and related clinical implications.

1. Introduction Bipolar disorder (BD) is a common, severe chronic mental illness, characterized by recurrent episodes of depression, mood elevation, and mixtures thereof (Miller et al., 2014). Lifetime prevalence of bipolar spectrum disorders in the United States is 4.4%, with the substantial heterogeneity of BD making diagnosis and treatment challenging,

resulting in substantial human, societal, and economic costs (Dilsaver, 2011; Merikangas et al., 2007). Accordingly, bipolar spectrum disorders have been associated with multi-dimensional impairment, challenging psychiatric comorbidities, compromised quality of life, and social stigma (Geddes and Miklowitz, 2013; Judd et al., 2005). Despite preventive treatment, mood episode recurrence occurs in more than one-third of BD patients within one year, almost two-thirds

Abbreviations: BD, bipolar disorder; BD I, bipolar I disorder; BD II, bipolar II disorder; CGI-BP-OS, Clinical Global Impression for Bipolar Disorder-Overall Severity; CI, confidence interval; df, degrees of freedom; HR, hazard ratio; MINI, Mini International Neuropsychiatric Interview; SCID for DSM-IV, Structured Clinical Interview for the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders; SD, standard deviation; SPSS, Statistical Package for the Social Sciences; STEP-BD, Systematic Treatment Enhancement Program for Bipolar Disorder ⁎ Correspondence to: Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, 401 Quarry Road, Room 2124, Stanford, CA 94305-5723 USA. E-mail address: [email protected] (T.A. Ketter). http://dx.doi.org/10.1016/j.jad.2017.05.007 Received 29 December 2016; Received in revised form 28 March 2017; Accepted 6 May 2017 Available online 08 May 2017 0165-0327/ © 2017 Elsevier B.V. All rights reserved.

Journal of Affective Disorders 219 (2017) 165–171

S. Shah et al.

comorbid (including lifetime anxiety disorder) psychiatric disorder diagnoses were determined by clinician consensus of results of the ADE and MINI (which was administered by trained research staff and assessed lifetime anxiety disorders) as well as available medical records. Clinical status at each follow-up visit was determined by the STEP-BD Clinical Monitoring Form (Sachs et al., 2002) while patients received naturalistic treatment (with monthly modal visit frequency) for up to 2 years. For this observational study that required only minimal patient effort (i.e., completing the STEP-BD ADE and MINI baseline assessments and the STEP-BD Clinical Monitoring Form longitudinal assessments), the base population and the recruited population were identical. Presence or absence of lifetime anxiety disorder/current anxiety symptoms was determined from subjective report (as assessed by the STEP-BD Affective Disorders Evaluation and MINI), clinical assessment, and medical records. Current anxiety symptoms severity was assessed in relationship to baseline anxiety symptoms pervasiveness according to the STEP-BD ADE, and thus reflected any anxiety symptoms in the ten days prior to enrollment for the primary analysis, and anxiety symptoms thresholded for occurring on at least four or seven of the ten days prior to enrollment for secondary analyses. Other current mood symptoms (such as anhedonia) were similarly assessed according to the STEP-BD ADE and quantified (thresholded for any in the ten days prior to enrollment for the primary analysis, and occurring on at least four or seven of the ten days prior to enrollment for secondary analyses). As described below, clinical characteristics of participants were evaluated and prospective clinical course of participants meeting diagnostic criteria for either a current major depressive episode or current recovery at enrollment were assessed. We used the STEP-BD definitions of recovery, which entailed sustained (≥8 weeks) euthymia/remission (fewer than three DSM-IV threshold-level symptoms of mood elevation or depression) and recurrence, which entailed development of a new DSM-IV syndromal mood elevation episode (hypomanic, manic, mixed) or major depressive episode, as assessed on the STEP-BD ADE and Clinical Monitoring Form, respectively. Thus, patients with subsyndromal mood symptoms (at least three DSM-IV threshold-level symptoms of mood elevation or depression but not meeting DSM-IV criteria for a syndromal mood elevation episode or major depressive episode, i.e. STEP-BD “continued symptoms” or “roughening” clinical status) were not included among recovered or syndromal episode patients. All patients were assessed with these STEPBD instruments rather than symptom rating scales. The STEP-BD protocol and the subsequent similar Stanford-specific Assessment, Monitoring, and Centralized Database protocol were approved by the Stanford University Administrative Panel on Human Subjects, and patients provided verbal and written informed consent prior to participation. Trained medical and research staff collected data on 6 demographic parameters and 22 illness characteristics/current mood symptoms/ current psychotropic use. The demographic parameters assessed were: (A) Age (in years); (B) Gender; (C) Race/Ethnicity; (D) Education; (E) Marital status; and (F) Employment status. The illness characteristics/ current mood symptoms/current psychotropic use assessed were: (1) Lifetime anxiety disorder; (2) Lifetime alcohol/substance use disorder; (3) Lifetime eating disorder; (4) Lifetime personality disorder; (5) BD II subtype; (5A) Lifetime psychosis (which is commonly associated with BD I); (5B) Lifetime prior psychiatric hospitalization (which is also commonly associated with BD I); (6) ≥ One first-degree relative with mood disorder; (7) Onset age (in years); (8) Childhood (age < 13 years) onset; (9) Illness duration (in years); (10) Long Illness duration (≥15 years); (11) Episode accumulation (≥10 prior mood episodes); (12) Lifetime suicide attempt; (13) Rapid cycling in prior year; and (14) CGIBP-OS; as well as current (i.e., in the prior 10 days) mood symptoms (15) Sadness; (16) Anhedonia; (17) Euphoria; (18) Irritability; (19) Anxiety, and psychotropic use (20) Mood stabilizers; (21) Antipsychotics; and (22) Antidepressants. In addition, DSM-5 depres-

within two years (Perlis et al., 2006), and nearly three-quarters within five years (Gitlin et al., 1995). Accordingly, although slightly more than half of Systematic Treatment Enhancement Program for BD (STEP-BD) patients who had a syndromal mood episode at enrollment recovered (i.e., became euthymic for ≥8 weeks), approximately half of these had mood episode recurrence within two years, with twice as many depressive versus mood elevation (manic, hypomanic, or mixed) recurrences (Perlis et al., 2006). Thus, depressive episodes in BD are prevalent and highly recurrent, even while patients are receiving optimized measurement-, evidence-, and guideline-based treatment (Perlis et al., 2006). Depressive compared to mood elevation episodes not only have higher incidence, but also greater multi-dimensional impairment, highlighting the importance of delaying depressive recurrence and hastening depressive recovery (Di Marzo et al., 2006; Merikangas et al., 2011; Perlis et al., 2006). Thus, identification and mitigation of clinical variables associated with hastened depressive recurrence and delayed depressive recovery is crucial to effective BD management. Anxiety in BD is highly prevalent (Pavlova et al., 2015), and associated with depression (Thompson et al., 2010; McIntyre et al., 2006), more intense psychiatric symptoms (McIntyre et al., 2006), and poorer affective (Tohen et al., 2007) and quality of life (KauerSant'Anna et al., 2007) outcomes. Lifetime anxiety disorder in BD has been associated with multiple unfavorable illness characteristics including earlier BD onset age, less time euthymic (Zutshi et al., 2006), and more subsyndromal symptoms (MacQueen et al., 2003), mood episode accumulation (i.e., ≥10 prior mood episodes) (Tamam and Ozpoyraz, 2002), substance disorder comorbitity, suicidality (Simon et al., 2004), and mood stabilizer resistance (Zutshi et al., 2006). Current anxiety disorder comorbidity in BD has also been associated with multiple unfavorable illness characteristics including earlier BD onset age, less time euthymic (Simon et al., 2004), and more prior-year depressive episodes (Bauer et al., 2005), severe depression, impaired global functioning (Lee and Dunner, 2008), eating (MacQueen et al., 2003) and substance use (McIntyre et al., 2006) disorder comorbidity, suicidality (Lee and Dunner, 2008), and, treatment-resistance (Henry et al., 2003). Anxiety disorders in BD have been associated with worse longitudinal outcomes (Boylan et al., 2004), including longer time to remission (Feske et al., 2000) and lower likelihood of timely recovery from depression and hastened depressive recurrence (Otto et al., 2006). Accordingly, among National Institute of Mental Health (NIMH) STEPBD patients, lifetime and current anxiety disorder(s)/symptoms were associated with significantly less time euthymic, whereas current (but not lifetime) anxiety disorder was associated with hastened depressive recurrence (Otto et al., 2006; Simon et al., 2004). Also, among NIMH Collaborative Depression Study BD (and unipolar depression) participants, number/severity of current anxiety symptoms (but not presence of pre-existing anxiety disorder) was associated with spending more time depressed (Coryell et al., 2012). Differential impacts of lifetime anxiety disorder and current anxiety symptoms upon delayed depressive recovery in BD remain to be established (Otto et al., 2006). We examined longitudinal relationships between lifetime anxiety disorder/ current anxiety symptoms and BD course. 2. Methods We included outpatients with bipolar I disorder (BD I) or bipolar II disorder (BD II) referred to the Stanford University BD Clinic between 2000 and 2011. Patients were assessed with the STEP-BD Affective Disorders Evaluation (Sachs et al., 2003), which included the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (First et al., 1996) mood disorders module, as well as the anxiety disorder screening questions from the Mini International Neuropsychiatric Interview (MINI) (Sheehan et al., 1998), and Clinical Global Impression for Bipolar Version-Overall Severity (CGI-BP-OS) score (Spearing et al., 1997). Bipolar and 166

Journal of Affective Disorders 219 (2017) 165–171

S. Shah et al.

prior to mood elevation recurrence in assessing time to mood elevation recurrence (Tohen et al., 1990). Cox regression analyses with HRs and 95% confidence intervals assessed potential mediators of significant longitudinal associations. To select parameters for entry into mediator models, univariate Cox proportional hazard analyses were performed for all significant clinical correlates of lifetime anxiety disorder/current anxiety symptoms in recovered BD patients, respectively. Thus, parameters with p < 0.05 were entered into a forward-stepwise procedure, and retained in the model if p < 0.05. Cox proportional hazard analyses with time-dependent covariates were used to further characterize associations between lifetime anxiety disorder/current anxiety symptoms and depressive recurrence. A two-tailed significance threshold of p < 0.05 was used, with no adjustment for multiple comparisons.

sive mixed symptoms were assessed in currently depressed patients. Statistical analyses were performed using Statistical Package for Social Sciences (SPSS) Version 23, Release 23.0.0 software (IBM Corporation, Armonk, NY) on an Apple MacBook Pro (Apple Inc., Cupertino, California). For the main analyses, baseline demographics and illness characteristics/current mood symptoms/current psychotropic use were compared in currently recovered (i.e., euthymic ≥8 weeks) patients and in currently depressed (i.e., with a current major depressive episode) patients with versus without lifetime anxiety disorder/current anxiety symptoms. Unpaired t-tests and Fisher's Exact tests were used for comparisons of continuous variables and categorical variables, respectively. Linear and logistic regressions were used to covary for potential demographic and clinical confounds. For lifetime anxiety disorder/current anxiety symptoms in currently recovered or depressed patients, relationships with times to recurrence and recovery were assessed using 2-year Kaplan-Meier survival analyses (Log-Rank tests) as primary metrics, and 2-year Cox Proportional Hazard Ratios (HRs) and 95% confidence intervals (CIs) as secondary metrics. Additional secondary metrics included observed and KaplanMeier estimated recurrence/recovery rates for significant longitudinal depressive associations and first year and second year Cox Proportional HRs and 95% CIs using Cox proportional hazard analyses with timedependent covariates. We used the standard approaches of censoring patients with mood elevation prior to depressive recurrence in assessing time to depressive recurrence, and censoring patients with depressive

3. Results Table 1 provides demographics, illness characteristics, current mood symptoms, and current psychotropic use in 105 currently recovered BD outpatients (stratified by presence/absence of lifetime anxiety disorder/current anxiety symptoms and in aggregate) from 503 participants enrolled in the study. At study entry, participants were commonly recovered (euthymic ≥8 weeks, N=105, 20.9%), even more often had syndromal depression (N=153, 30.4%), but substantially less often had syndromal mood elevation (N=54, 10.7%).

Table 1 Demographics, illness characteristics, and current mood symptoms in recovered bd patients in aggregate and with versus without lifetime anxiety disorder/current anxiety symptoms.

N (%) Demographics A. Age (years, mean ± SD) B. Female (%) C. Caucasian (%) D. College degree (%) E. Married (current, %) F. Full-time employment (current, %) Comorbid disorders (lifetime, %) 1. Anxiety 2. Alcohol/substance Use 3. Eating 4. Personality Other illness characteristics 5. Bipolar II disorder (%) 5A. Psychosis (lifetime, %) 5B. Psychiatric hospitalization (lifetime, %) 6. ≥One 1° degree relative w mood disorder (%) 7. Onset age (years, mean ± SD) 8. Childhood (age < 13 years) Onset (%) 9. Illness Duration (years, mean ± SD) 10. Long illness Duration (≥ 15 years, %) 11. Episode accumulation (≥ 10, lifetime, %) 12. Suicide Attempt (lifetime, %) 13. Rapid Cycling (prior year, %) 14. CGI-BP-OS (current, mean ± SD) Current mood symptoms (any in prior 10 days, %) 15. Sadness 16. Anhedonia 17. Euphoria 18. Irritability 19. Anxiety Current psychotroptic use (%) 20. Mood stabilizers 21. Antipsychotics 22. Antidepressants

Recovered w lifetime anxiety disorder

Recovered wo lifetime anxiety disorder

Recovered w current anxiety symptoms

Recovered wo current anxiety symptoms

All recovered

49 (46.7)

56 (53.3)

38 (36.2)§§

67 (63.8)

105 (100.0)

33.8 ± 11.3 65.3 77.6 57.1 40.8 30.6

38.1 ± 15.4 46.4 75.5 67.9 32.7 37.0

37.5 ± 11.8 71.1§ 75.7 59.5 50.0§ 31.6

35.3 ± 14.7 46.3 76.9 64.6 28.8 35.4

36.1 ± 13.7 55.2 76.5 62.7 36.5 34.0

100.0**** 53.1 16.3* 8.2

0.0 48.2 3.6 8.9

73.7§§§§ 55.3 13.2 13.2

31.3 47.8 7.5 6.0

46.7 50.5 9.5 8.6

51.0* 51.0 44.9 57.1 17.3 ± 6.7* 16.3 17.6 ± 13.3 50.0 64.4* 33.3 12.8 2.5 ± 0.7***

28.6 42.9 48.2 37.5 21.3 ± 10.2 10.9 17.3 ± 14.4 46.2 38.3 20.4 7.7 1.9 ± 0.8

50.0 44.7 47.4 60.5§ 17.8 ± 7.5 13.2 20.6 ± 14.0 58.3 74.3§§§ 37.8 10.8 2.7 ± 0.4§§§§

32.8 47.8 46.3 38.8 20.3 ± 9.5 13.6 15.6 ± 13.5 41.9 36.8 20.0 9.7 1.9 ± 0.7

39.0 46.7 46.7 46.7 19.4 ± 8.9 13.5 17.4 ± 13.8 48.0 51.1 26.5 10.1 2.2 ± 0.8

24.5 22.4* 24.5** 46.9* 57.1****

14.3 7.1 3.6 26.8 17.9

34.2§§ 28.9§§ 28.9§§§ 57.9§§§ 100.0§§§§

10.4 6.0 4.5 23.9 0.0

19.0 14.3 13.3 36.2 36.2

79.6 44.9 46.9**

74.5 29.1 17.9

86.8 39.5 39.5

71.0 34.8 26.9

73.3 36.2 31.4

*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 vs. Recovered wo Lifetime Anxiety Disorder. § p < 0.05, §§p < 0.01, §§§p < 0.001, §§§§p < 0.0001 vs. Recovered wo Current Anxiety Symptoms. CGI-BP-OS indicates Clinical Global Impression for Bipolar Disorder-Overall Severity; SD indicates standard deviation; w indicates with; wo indicates without. Boldface font indicates parameters with statistically significant relationships with lifetime anxiety disorder or current anxiety symptoms. Missing data: 12.4% for ≥10 prior episodes, 0.0–5.7% for other parameters.

167

Journal of Affective Disorders 219 (2017) 165–171

S. Shah et al.

Among depressed patients, lifetime anxiety disorder overlapped substantially with current anxiety symptoms, which were seen in 81.0% of depressed patients overall and in 83.1% and 74.3% of depressed patients with and without lifetime anxiety disorder, respectively. Currently depressed patients with versus without lifetime anxiety disorder had significantly higher rates of lifetime anxiety disorder (by definition), childhood onset, and current irritability but a significantly lower rate of prior psychiatric hospitalization. However, the latter appeared to be secondary to the non-significantly lower rate of BD I (i.e., non-significantly higher rate of BD II, 65.3% versus 48.6%) in depressed patients with versus without lifetime anxiety disorder. Depressed patients with versus without lifetime anxiety disorder did not differ significantly with respect to depressive mixed features or any demographic parameter or other illness characteristic/current mood symptom/psychotropic in Table 1. Similarly, with a prevalence of 81.0%, current anxiety symptoms among depressed patients were extremely common, significantly more common than lack of current anxiety symptoms, which was seen in only 19.0% of, and significantly more common than current anxiety symptoms among recovered patients, which were seen in only 36.2% of recovered patients. Among depressed patients, current anxiety symptoms overlapped substantially with lifetime anxiety disorder, which was seen in 77.1% of depressed patients overall and in 79.0% and 69.0% of depressed patients with and without current anxiety symptoms, respectively. Currently depressed patients with versus without current anxiety symptoms had significantly higher rates of lifetime personality disorder and current anhedonia, irritability, and anxiety (by definition), but a significantly lower rate of prior psychiatric hospitalization. However, the latter appeared to be secondary to the non-significantly lower rate of BD I (i.e., non-significantly higher rate of BD II, 64.5% versus 48.3%) in depressed patients with versus without current anxiety symptoms. Depressed patients with versus without current anxiety symptoms did not differ significantly with respect to depressive mixed features or any demographic parameter or other illness characteristic/current mood symptom/psychotropic in Table 1.

3.1. Demographics & illness characteristics/current mood symptoms/ current psychotropic use in recovered patients with versus without lifetime anxiety disorder/current anxiety symptoms Lifetime anxiety disorder was common among recovered patients, with a prevalence of 46.7%, but was even more common among depressed patients (77.1%). Among recovered patients, rates of having versus lacking lifetime anxiety disorder were not significantly different (46.7% vs. 53.3%), and lifetime anxiety disorder overlapped with (but was not identical to) current anxiety symptoms, which were seen in 36.2% of recovered patients overall and in 57.1% and 17.9% of recovered patients with and without lifetime anxiety disorder, respectively. Currently recovered patients with versus without lifetime anxiety disorder had significantly higher rates of lifetime anxiety disorder (by definition), lifetime eating disorder, BD II subtype, episode accumulation, and current anhedonia, euphoria, irritability, and anxiety, and current antidepressant use as well as earlier BD onset age and worse current CGI-BP-OS score (Table 1). Also, current anxiety and euphoria (but not current anhedonia and irritability) dichotomized for occurrence versus lack of occurrence on at least four and seven of the prior ten days remained significantly associated with lifetime anxiety disorder in recovered BD patients. Recovered patients with versus without lifetime anxiety disorder did not differ significantly for any demographic parameter or other illness characteristic/current mood symptom in Table 1. Similarly, current anxiety symptoms were common among recovered patients, occurring in 36.2%, but even more common among depressed patients (occurring in 81.0%). Among recovered patients, there was a lower prevalence of BD patients with versus without current anxiety symptoms (36.2% vs. 63.8%), and current anxiety symptoms overlapped with (but were not identical to) lifetime anxiety disorder, which was seen in 46.7% of recovered patients overall and in 73.7% and 31.3% of recovered patients with and without current anxiety symptoms, respectively. Currently recovered patients with versus without current anxiety symptoms had significantly higher rates of female gender, being currently married, lifetime anxiety disorder, having at least one first degree relative with mood disorder, episode accumulation, and current sadness, anhedonia, euphoria, irritability, and anxiety (by definition), as well as worse current CGI-BP-OS (Table 1). Also, current sadness (but not anxiety, anhedonia, euphoria, and irritability) dichotomized for occurrence versus lack of occurrence on at least four (but not seven) of the prior ten days remained significantly associated with current anxiety symptoms dichotomized for occurrence versus lack of occurrence on at least four (but not seven) of the prior ten days. Dichotomizing current anxiety symptoms in recovered BD patients for occurrence versus lack of occurrence on at least four and seven of the prior ten days yielded a pattern of relationships with Table 1 parameters similar to that described above. Recovered patients with versus without current anxiety symptoms did not differ significantly for any other demographic parameter or illness characteristic/current mood symptom in Table 1. All of the above-mentioned significant relationships between current anxiety symptoms and other illness characteristics/current mood symptoms in recovered patients remained significant after covarying for gender and marital status.

3.3. Lifetime anxiety disorder associations with longitudinal course Lifetime anxiety disorder was significantly associated with hastened depressive recurrence in 49 versus 56 recovered patients with versus without lifetime anxiety disorder (Fig. 1), driven by earlier BD onset age, but not other parameters in Table 1. Lifetime anxiety disorder's association with hastened depressive recurrence was significant in the first year, but non-significant in the second year. The 2-year observed depressive recurrence rates in patients with versus without lifetime anxiety disorder were 49.0% versus 17.9%. Similarly, the 2-year Kaplan-Meier estimated depressive recurrence rate was significantly higher (more than two-fold, at 64.4% versus 29.0%) among patients with (95% CI (48.1–80.7%)) compared to without (95% CI (12.8–45.2%)) lifetime anxiety disorder. Importantly, lifetime anxiety disorder was only non-significantly related to delayed mood elevation recurrence and hastened any mood episode recurrence. The observed and Kaplan-Meier estimated overall (all patients, any episode) 2-year recurrence rates were 50.5% and 65.9% (95% CI 54.9–76.9%), respectively. As expected, the observed and Kaplan-Meier estimated overall (all patients, any episode) 1-year recurrence rates were lower at 41.0% and 50.6% (95% CI 39.8–61.4%), respectively. Lifetime anxiety disorder was only non-significantly related to delayed depressive recovery (in 118 versus 35 depressed patients with versus without lifetime anxiety disorder, not illustrated). As lifetime anxiety disorder was not significantly related to delayed depressive recovery, and not significantly related to depressive mixed features, we did not assess the impact of mixed features upon time to depressive recovery. Importantly, lifetime anxiety disorder was only non-significantly related to delayed mood elevation recovery (in 34 vs. 20 manic/

3.2. Demographics & illness characteristics/current mood symptoms/ current psychotropic use in depressed patients with versus without lifetime anxiety disorder/current anxiety symptoms With a prevalence of 77.1%, lifetime anxiety disorder among depressed patients was very common, significantly more common than lack of lifetime anxiety disorder, which was seen in only 22.9%, and significantly more common than lifetime anxiety disorder among recovered patients, which was seen in only 46.7% of recovered patients. 168

Journal of Affective Disorders 219 (2017) 165–171

S. Shah et al.

Fig. 1. Lifetime anxiety disorder associated with hastened depressive recurrence in bipolar disorder. Two-year survival analysis of time to depressive recurrence in recovered bipolar disorder patients indicated significantly hastened depressive recurrence in patients with (N=49, black line on bottom) versus without (N=56, gray line on top) lifetime anxiety disorder (Log-Rank p=0.002). Lifetime anxiety disorder was also significantly associated with hastened depressive recurrence using Cox Proportional Hazard analysis (Hazard Ratio (HR)=3.0, 95% Confidence Interval (CI)=1.4−6.3, p=0.004). Earlier bipolar disorder onset age (HR=1.1, 95% CI=1.0−1.1, p=0.039) drove hastened depressive recurrence in patients with versus without lifetime anxiety disorder.

Fig. 2. Current anxiety symptoms associated with hastened depressive recurrence in bipolar disorder. Two-year survival analysis of time to depressive recurrence in recovered bipolar disorder patients indicated significantly hastened depressive recurrence in patients with (N =38, black line on bottom) versus without (N=67, gray line on top) current anxiety symptoms (Log-Rank p=0.042). Current anxiety symptoms were also significantly associated with hastened depressive recurrence using Cox Proportional Hazard analysis (HR=2.0, 95% CI=1.0 – 3.9, p=0.046). Lifetime anxiety disorder (HR=2.9, 95% CI=1.2 – 7.2, p=0.022) drove hastened depressive recurrence in patients with versus without current anxiety symptoms.

165 versus 42 depressed/manic/hypomanic/mixed patients with versus without current anxiety symptoms (not illustrated), not driven by any assessed potential mediator. Upon examining current anxiety symptoms thresholded for having occurred on at least four or seven of the ten days prior to enrollment, relationships between current anxiety symptoms and times to depressive recurrence/recovery were all non-significant.

hypomanic/mixed patients with versus without lifetime anxiety disorder), and delayed any mood episode recovery (in 152 vs. 55 depressed/manic/hypomanic/mixed patients with versus without lifetime anxiety disorder). The observed and Kaplan-Meier estimated overall (all patients, any episode) 2-year recovery rates were 46.4% and 84.3% (95% CI 76.5–92.1%), respectively. As expected, the observed and Kaplan-Meier estimated overall (all patients, any episode) 1-year recovery rates were lower at 37.2% and 59.5% (95% CI 50.5–68.5%), respectively.

4. Discussion We found lifetime anxiety disorder was significantly associated with multiple baseline other unfavorable illness characteristics/current mood symptoms/current psychotropic use, hastened depressive recurrence (driven by earlier BD onset age), and a significantly (more than two-fold) higher Kaplan-Meier estimated depressive recurrence rate, whereas current anxiety symptoms were also significantly associated with multiple baseline other unfavorable illness characteristics/current mood symptoms/current psychotropic use and hastened depressive recurrence (driven by lifetime anxiety disorder), but only a nonsignificantly (slightly less than two-fold) higher Kaplan-Meier estimated depressive recurrence rate. Neither lifetime anxiety disorder nor current anxiety symptoms was significantly associated with delayed depressive recovery. Our finding of a robust association between lifetime anxiety disorder and hastened depressive recurrence is novel. While current comorbid anxiety disorder has previously been associated with shorter time euthymic (Simon et al., 2004) and has been a mediator of depressive recurrence (Otto et al., 2006; Perlis et al., 2006), lifetime anxiety disorder has not previously been significantly associated with hastened depressive recurrence (Coryell et al., 2012; Otto et al., 2006; Perlis et al., 2006), although it has been related to shorter time euthymic (Zutshi et al., 2006). Furthermore, we found that early BD onset, which had previously been associated with lifetime anxiety disorder (Holtzman et al., 2015; Simon et al., 2004), earlier and greater risk of mood episode recurrence, fewer days euthymic, and more mood symptom chronicity, and functional impairment (Perlis et al., 2009), was a driver of hastened depressive recurrence in patients with lifetime anxiety disorder. This phenomenological overlap between BD onset and

3.4. Current anxiety symptom associations with longitudinal course Current anxiety symptoms were also significantly associated with hastened depressive recurrence in 38 versus 67 recovered patients with versus without current anxiety symptoms (Fig. 2), driven by lifetime anxiety disorder. However, current anxiety symptoms’ association with hastened depressive recurrence was non-significant in both the first year and the second year. The 2-year observed depressive recurrence rates in patients with versus without current anxiety symptoms were 47.4% versus 23.9%. However, the 2-year Kaplan-Meier estimated depressive recurrence rate was only non-significantly (slightly less than two-fold at 64.2% versus 32.9%) higher among patients with (95% CI 44.4–84.0%) compared to without (95% CI 17.8–48.0%) current anxiety symptoms. Importantly, current anxiety symptoms were only non-significantly related to hastened mood elevation recurrence and hastened any mood episode recurrence. Current anxiety symptoms were only non-significantly related to delayed depressive recovery in 124 versus 29 depressed patients with versus without current anxiety symptoms, being significant in the first year, driven by current irritability, yet only non-significant in the second year. As current anxiety symptoms were not significantly related to delayed depressive recovery, and not significantly related to depressive mixed features, we did not assess the impact of mixed features upon time to depressive recovery. Importantly, current anxiety symptoms were also only non-significantly associated with delayed mood elevation recovery in 41 versus 13 manic/hypomanic/mixed patients with versus without current anxiety symptoms, but were significantly associated with delayed any mood episode recovery in 169

Journal of Affective Disorders 219 (2017) 165–171

S. Shah et al.

0.59), but numerically higher 2-year (0.24 versus 0.13) and first-year (0.051 versus 0.038) Cox Proportional HR p-values for delayed depressive recovery. Our study has noteworthy strengths and limitations. Strengths include being one of the first studies to assess relationships between illness characteristics/current mood symptoms and not only hastened depressive recurrence, but also delayed depressive recovery, using validated instruments to assess diagnosis and clinical course, and having a substantial number of not only recovered (N=105), but also depressed (N=153) well-characterized BD patients. Furthermore, our overall (all patients, any episode) recurrence/recovery rates were in broad agreement with prior studies (Otto et al., 2006; Perlis et al., 2006). However, these strengths are accompanied by limitations that include using a sample of patients referred to a suburban Northern California BD specialty clinic, limiting the generalizability of our findings in our relatively affluent, well-educated but relatively underemployed, measurement-based care oriented, and predominantly female sample of BD patients with medical insurance, rather than a more heterogeneous mixture of BD inpatients and outpatients being treated in non-specialty clinical settings. Moreover, our sample size, although substantial overall (N=503), was more limited for assessing longitudinal impact of lifetime anxiety disorder/current anxiety symptoms (153 depressed,105 recovered,and 54 elevated BD patients), and yielded limited statistical power when assessing some smaller subgroups (e.g., the 49 and 38 recovered BD patients with lifetime anxiety disorder and current anxiety symptoms, respectively). Hence, limited statistical power could have undermined our ability to detect relationships between DSM-5 depressive mixed features and lifetime anxiety disorder, current anxiety symptoms, and time to recovery from depression. Moreover, the number of observed recurrences for mood elevation was only 19, versus 42 for depression. Thus, limited statistical power likely contributed importantly to our inability to detect significant relationships between lifetime anxiety disorder/current anxiety symptoms and mood elevation recurrence/recovery. Furthermore, recovered status/episode duration prior to enrollment was not included in our analyses. Another important limitation is the open naturalistic treatment design, which entailed patients receiving diverse uncontrolled (albeit measurement-, evidence-, and guideline-based) therapies. Consequently, raters on follow up were not blind to baseline information – this is clearly a limitation (permitting potential rater bias) that may also be viewed as a strength (enhancing generalizability to “real-world” clinical settings). Finally, we did not correct for multiple comparisons, which particularly limited interpretation of findings with p-values between 0.05 and 0.01. However, this liberal statistical approach enhanced assay sensitivity with respect to our ability to detect relationships between illness characteristics/current mood symptoms and depressive burden. Nevertheless, we contend that our observations of associations between lifetime anxiety disorder/current anxiety symptoms and hastened depressive recurrence suggest that lifetime anxiety disorder/ current anxiety symptoms are important indicators of longitudinal bipolar depressive burden. Additional research is needed to assess potential differential relationships between lifetime anxiety disorder and current anxiety symptoms and hastened/delayed depressive recurrence/recovery, respectively – specifically, whether lifetime anxiety disorder versus current anxiety symptoms is more robustly associated with hastened depressive recurrence, and whether both are more robustly associated with hastened depressive recurrence versus recovery.

prognosis and lifetime anxiety disorder challenges the contention that BD and anxiety disorders are neurobiologically distinct disorders (McIntyre et al., 2006). Moreover, lifetime anxiety disorder appears to be an important driver of hastened depressive recurrence in BD patients with not only current anxiety symptoms (as observed in this report), but also with current irritability (Yuen et al., 2016b), episode accumulation (Ketter et al., 2016), and BD II versus BD I (Dell'Osso et al., 2017). In contrast, lifetime anxiety disorder was only non-significantly associated with delayed depressive recovery, suggesting the previously observed shorter time euthymic in patients with lifetime anxiety disorder might be more due to hastened depressive recurrence than due to delayed depressive recovery. However, other studies have found comorbid anxiety disorder in BD associated with longer time to remission (Feske et al., 2000), and lower likelihood of timely recovery from depression (Otto et al., 2006). Further studies are needed regarding the clinical correlates and psychopathological mechanisms subserving anxiety disorders in BD (McIntyre et al., 2006). Our finding of a significant association between current anxiety symptoms and hastened depressive recurrence is consistent with prior observations in BD of robust and continuous relationships with time spent depressed and number/severity of anxiety symptoms (Coryell et al., 2012) and increased percentage of days with anxiety during the prior year being associated with hastened depressive recurrence (Perlis et al., 2006). Lifetime anxiety disorder's substantial association with current anxiety symptoms in recovered BD patients and it's role in mediating hastened depressive recurrence in patients with current anxiety symptoms, indicates that lifetime anxiety comorbidity could be importantly related to current anxiety symptoms and current mood symptoms, inadequate treatment response, and poor functional outcome (Keller, 2006), once again challenging the contention that BD and anxiety disorders are neurobiologically distinct disorders (McIntyre et al., 2006). While in our study, current anxiety symptoms’ association with delayed depressive recovery was non-significant during the full two-year study interval, it was significant during the first year, consistent with prior reports of significant associations between delayed first-year depressive recovery and current comorbid anxiety disorder (Otto et al., 2006) and current anxiety symptoms (Feske et al., 2000), suggesting a time-limited (one- rather than two-year) impact of current anxiety symptoms upon bipolar depressive recurrence. Other research suggests current anxiety symptoms have robust relationships with current irritability (Yuen et al., 2016a), which in turn has been hypothesized to involve central noradrenergic system hyperactivity and amplification of stress responses, thereby increasing arousal/ emotional reactions to stress like irritability and anxiety (Yamamoto et al., 2014), and possibly even delayed BD depressive recovery observed in patients with current irritability (Yuen et al., 2016b), which in the current study drove the association between current anxiety symptoms and delayed first year BD depressive recovery. This robustness of irritability-anxiety relationships and associations of both irritability and anxiety with unfavorable BD illness characteristics and depressive burden underscore the need to further our understanding of such interactions (Yuen et al., 2016a). Potential differential relationships between lifetime anxiety disorder and current anxiety symptoms and hastened/delayed depressive recurrence/recovery, respectively, remain to be examined. Our observations raise the possibility that lifetime anxiety disorder compared to current anxiety symptoms has a more robust association with hastened depressive recurrence, as we saw numerically lower 2-year (0.004 versus 0.046) and first-year (0.004 versus 0.091), but higher secondyear (0.50 versus 0.26) Cox Proportional HR p-values for hastened depressive recurrence, and a significant versus non-significant increase in 2-year Kaplan-Meier estimated depressive recurrence rate. In contrast, lifetime anxiety disorder compared to current anxiety symptoms might not have more robust associations with delayed depressive recovery, as we saw numerically lower second-year (0.22 versus

Disclosure of financial relationships (past 36 months) Saloni Shah and Laura Yuen as well as Drs. Park, Jane Kim, Hyun Kim, Hooshmand and Wang report no financial relationships with commercial interests. Dr. Miller has received grant/research support from Merck and Company and Sunovion, Inc. Dr. Dell’Osso has received 170

Journal of Affective Disorders 219 (2017) 165–171

S. Shah et al.

significant clinical correlates. J. Affect. Disord. 179, 114–120. Judd, L.L., Akiskal, H.S., Schettler, P.J., Endicott, J., Leon, A.C., Solomon, D.A., et al., 2005. Psychosocial disability in the course of bipolar I and II disorders: a prospective, comparative, longitudinal study. Arch. Gen. Psychiatry 62, 1322–1330. Kauer-Sant'Anna, M., Frey, B.N., Andreazza, A.C., Cereser, K.M., Gazalle, F.K., Tramontina, J., et al., 2007. Anxiety comorbidity and quality of life in bipolar disorder patients. Can. J. Psychiatry 52, 175–181. Keller, M.B., 2006. Prevalence and impact of comorbid anxiety and bipolar disorder. J. Clin. Psychiatry 67, 5–7. Ketter, T.A., Park, D.Y., Shah, S., Yuen, L.D., Hooshmand, F., Dell’Osso, B., Miller, S., Wang, P.W., 2016. Faster depressive recurrence related to episode accumulation in bipolar disorder. Poster presented at the Annual Meeting of the American Psychiatric Association, 14–18 May, Atlanta, GA. Lee, J.H., Dunner, D.L., 2008. The effect of anxiety disorder comorbidity on treatment resistant bipolar disorders. Depression Anxiety 25, 91–97. MacQueen, G.M., Marriott, M., Begin, H., Robb, J., Joffe, R.T., Young, L.T., 2003. Subsyndromal symptoms assessed in longitudinal, prospective follow-up of a cohort of patients with bipolar disorder. Bipolar Disord. 5, 349–355. McIntyre, R.S., Soczynska, J.K., Bottas, A., Bordbar, K., Konarski, J.Z., Kennedy, S.H., 2006. Anxiety disorders and bipolar disorder: a review. Bipolar Disord. 8, 665–676. Merikangas, K.R., Akiskal, H.S., Angst, J., Greenberg, P.E., Hirschfeld, R.M., Petukhova, M., et al., 2007. Lifetime and 12-month prevalence of bipolar spectrum disorder in the National Comorbidity Survey replication. Arch. Gen. Psychiatry 64, 543–552. Merikangas, K.R., Jin, R., He, J.P., Kessler, R.C., Lee, S., Sampson, N.A., et al., 2011. Prevalence and correlates of bipolar spectrum disorder in the world mental health survey initiative. Arch. Gen. Psychiatry 68, 241–251. Miller, S., Dell'Osso, B., Ketter, T.A., 2014. The prevalence and burden of bipolar depression. J. Affect. Disord. 169, S3–S11. Otto, M.W., Simon, N.M., Wisniewski, S.R., Miklowitz, D.J., Kogan, J.N., ReillyHarrington, N.A., et al., 2006. Prospective 12-month course of bipolar disorder in out-patients with and without comorbid anxiety disorders. Br. J. Psychiatry 189, 20–25. Pavlova, B., Perlis, R.H., Alda, M., Uher, R., 2015. Lifetime prevalence of anxiety disorders in people with bipolar disorder: a systematic review and meta-analysis. Lancet Psychiatry 2, 710–717. Perlis, R.H., Ostacher, M.J., Patel, J.K., Marangell, L.B., Zhang, H., Wisniewski, S.R., et al., 2006. Predictors of recurrence in bipolar disorder: primary outcomes from the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD). Am. J. Psychiatry 163, 217–224. Perlis, R.H., Dennehy, E.B., Miklowitz, D.J., Delbello, M.P., Ostacher, M., Calabrese, J.R., et al., 2009. Retrospective age at onset of bipolar disorder and outcome during twoyear follow-up: results from the STEP-BD study. Bipolar Disord. 11, 391–400. Sachs, G.S., Guille, C., McMurrich, S.L., 2002. A clinical monitoring form for mood disorders. Bipolar Disord. 4, 323–327. Sachs, G.S., Thase, M.E., Otto, M.W., Bauer, M., Miklowitz, D., Wisniewski, S.R., et al., 2003. Rationale, design, and methods of the systematic treatment enhancement program for bipolar disorder (STEP-BD). Biol. Psychiatry 53, 1028–1042. Sheehan, D.V., Lecrubier, Y., Sheehan, K.H., Amorim, P., Janavs, J., Weiller, E., et al., 1998. The mini-International neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD10. J. Clin. Psychiatry 59, 22–33. Simon, N.M., Otto, M.W., Wisniewski, S.R., Fossey, M., Sagduyu, K., Frank, E., et al., 2004. Anxiety disorder comorbidity in bipolar disorder patients: data from the first 500 participants in the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD). Am. J. Psychiatry 161, 2222–2229. Spearing, M.K., Post, R.M., Leverich, G.S., Brandt, D., Nolen, W., 1997. Modification of the Clinical Global Impressions (CGI) Scale for use in bipolar illness (BP): the CGI-BP. Psychiatry Res. 73, 159–171. Tamam, L., Ozpoyraz, N., 2002. Comorbidity of anxiety disorder among patients with bipolar I disorder in remission. Psychopathology 35, 203–209. Thompson, P.M., Gonzalez, J.M., Singh, V., Schoolfield, J.D., Katz, M.M., Bowden, C.L., 2010. Principal domains of behavioral psychopathology identified by the bipolar Inventory of signs and symptoms scale (BISS). Psychiatry Res. 175, 221–226. Tohen, M., Waternaux, C.M., Tsuang, M.T., 1990. Outcome in Mania. A 4-year prospective follow-up of 75 patients utilizing survival analysis. Arch. Gen. Psychiatry 47, 1106–1111. Tohen, M., Calabrese, J., Vieta, E., Bowden, C., Gonzalez-Pinto, A., Lin, D., et al., 2007. Effect of comorbid anxiety on treatment response in bipolar depression. J. Affect. Disord. 104, 137–146. Yamamoto, K., Shinba, T., Yoshii, M., 2014. Psychiatric symptoms of noradrenergic dysfunction: a pathophysiological view. Psychiatry Clin. Neurosci. 68, 1–20. Yuen, L.D., Miller, S., Wang, P.W., Hooshmand, F., Holtzman, J.N., Goffin, K.C., et al., 2016a. Current irritability robustly related to current and prior anxiety in bipolar disorder. J. Psychiatr. Res. 79, 101–107. Yuen, L.D., Shah, S., Do, D., Miller, S., Wang, P.W., Hooshmand, F., et al., 2016b. Current irritability associated with hastened depressive recurrence and delayed depressive recovery in bipolar disorder. Int. J. Bipolar Disord. 4, 15. Zutshi, A., Reddy, Y.C., Thennarasu, K., Chandrashekhar, C.R., 2006. Comorbidity of anxiety disorders in patients with remitted bipolar disorder. Eur. Arch. Psychiatry Clin. Neurosci. 256, 428–436.

grant/research support from Cyberonics, Inc. and AstraZeneca and Lundbeck and lecture honoraria from AstraZeneca and Lundbeck. Dr. Ketter has received grant/research support from the Agency for Healthcare Research and Quality, AstraZeneca Pharmaceuticals LP, Cephalon Inc., Eli Lilly and Company, National Institute of Mental Health, Pfizer Inc., and Sunovion Pharmaceuticals; consultant fees from Allergan, Inc., Avanir Pharmaceuticals, Bristol-Myers Squibb Company, Cephalon Inc., Forest Pharmaceuticals, Janssen Pharmaceutical Products, LP, Merck & Co., Inc., Sunovion Pharmaceuticals, and Teva Pharmaceuticals; lecture honoraria from Abbott Laboratories, Inc., AstraZeneca Pharmaceuticals LP, GlaxoSmithKline, and Otsuka Pharmaceuticals; and publication royalties from American Psychiatric Publishing, Inc. In addition, Dr. Ketter's spouse is an employee of and holds stock in Janssen Pharmaceuticals. Funding sources This research was supported by a Government of Korea Overseas Research Fellowship (2014-I-0040, Dr. Park), 2014 and 2015 Research Year Grants from Inje University (Dr. Hyun Kim), the Pearlstein Family Foundation, the Mitchell Foundation, and the Holland Foundation. No pharmaceutical company funding. Acknowledgement These data have been accepted for presentation at the 170th Annual Meeting of the American Psychiatric Association in San Diego, California, May 20–24, 2017. Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.jad.2017.05.007. References Bauer, M.S., Altshuler, L., Evans, D.R., Beresford, T., Williford, W.O., Hauger, R., 2005. Prevalence and distinct correlates of anxiety, substance, and combined comorbidity in a multi-site public sector sample with bipolar disorder. J. Affect. Disord. 85, 301–315. Boylan, K.R., Bieling, P.J., Marriott, M., Begin, H., Young, L.T., MacQueen, G.M., 2004. Impact of comorbid anxiety disorders on outcome in a cohort of patients with bipolar disorder. J. Clin. Psychiatry 65, 1106–1113. Coryell, W., Fiedorowicz, J.G., Solomon, D., Leon, A.C., Rice, J.P., Keller, M.B., 2012. Effects of anxiety on the long-term course of depressive disorders. Br. J. Psychiatry 200, 210–215. Dell'Osso, B., Shah, S., Do, D., Yuen, L.D., Miller, S., Wang, P.W., et al., 2017. American tertiary clinic-referred bipolar II disorder versus bipolar I disorder associated with hastened depressive recurrence. Int. J. Bipolar Disord. 5 (1), 2. Di Marzo, S., Giordano, A., Pacchiarotti, I., Colom, F., Sánchez-Moreno, J., Vieta, E., 2006. The impact of the number of episodes on the outcome of bipolar disorder. Eur. J. Psychiatry 20, 21–28. Dilsaver, S.C., 2011. An estimate of the minimum economic burden of bipolar I and II disorders in the United States: 2009. J. Affect. Disord. 129, 79–83. Feske, U., Frank, E., Mallinger, A.G., Houck, P.R., Fagiolini, A., Shear, M.K., et al., 2000. Anxiety as a correlate of response to the acute treatment of bipolar I disorder. Am. J. Psychiatry 157, 956–962. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., 1996. Structured Clinical Interview for DSM-IV Axis I Disorders, Patient Edition (SCID-I/P, Version 2.0). New York: Biometrics Research Department, New York State Psychiatric Institute, New York. Geddes, J.R., Miklowitz, D.J., 2013. Treatment of bipolar disorder. Lancet 381, 1672–1682. Gitlin, M.J., Swendsen, J., Heller, T.L., Hammen, C., 1995. Relapse and impairment in bipolar disorder. Am. J. Psychiatry 152, 1635–1640. Henry, C., Van den Bulke, D., Bellivier, F., Etain, B., Rouillon, F., Leboyer, M., 2003. Anxiety disorders in 318 bipolar patients: prevalence and impact on illness severity and response to mood stabilizer. J. Clin. Psychiatry 64, 331–335. Holtzman, J.N., Miller, S., Hooshmand, F., Wang, P.W., Chang, K.D., Hill, S.J., et al., 2015. Childhood-compared to adolescent-onset bipolar disorder has more statistically

171