Anxiety disorders are frequently comorbid with depression, but there is also evidence of a strong association with bipolar disorder. Indeed, symptoms of anxiety are often present before the onset of bipolar disorder (Faedda et al., 2014, Duffy et al., 2014). This particular comorbidity presents a challenge for diagnosis and treatment, as the first-line medications for anxiety disorders are often SSRIs, which may induce mood elevation and switching to mania in those with bipolar disorder. Additionally, comorbidity of anxiety disorders leads to worse outcomes and lower quality of life in those with bipolar disorder (Das, 2013). However, research into comorbidity with specific anxiety disorders are lacking.
A recent paper by Preti et al (2018) aimed to systematically review the literature for papers reporting on comorbid panic disorder in bipolar disorder and performed meta-analyses for prevalence of the comorbidity.
Preti et al conducted a systematic literature search using the PubMed/MEDLINE database for all studies with primary data on patients with bipolar disorder and a comorbid diagnosis of panic disorder up until 1st August 2017. Additionally, the reference lists of relevant books and reviews were scanned for additional papers.
The inclusion criteria were studies reporting on patients with a confirmed diagnosis of bipolar disorder and panic disorder, and studies published in the English language.
Both fixed and random effects models were reported for the meta-analyses and the study used appropriate statistical tests to identify outliers. The authors followed the recommendations of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.
A total of 37 studies met inclusion criteria for the review:
- 16 were cross-sectional studies reporting point prevalence
- 26 were longitudinal studies reporting lifetime prevalence of panic disorder in those with bipolar disorder.
15 studies were included in the meta-analysis for the point prevalence of panic disorder in bipolar disorder. The point prevalence estimate was 15.1% (95% Confidence Interval (CI): 7.9% to 24.0%) in the random effects model.
The heterogeneity was substantial (I2=95.9%, 95% CI: 94.5% to 97.0%). One likely outlier was identified with a particularly high prevalence of panic disorder, and after its exclusion the point prevalence was slightly reduced to 13.0% (95% CI: 7.0% to 20.3%), with a small reduction in the heterogeneity (I2=94.9%, 95%CI: 92.9% to 96.3%).
Funnel plots revealed no asymmetry and Egger’s test did not reveal any evidence of publication bias.
25 studies were included in the meta-analysis of longitudinal data on the lifetime prevalence of panic disorder in those with bipolar disorder. The estimated lifetime prevalence was 16.8% (95% CI: 12.2% to 22.0%) in the random effects model.
The heterogeneity for the lifetime prevalence meta-analysis was also very high (I2=95.6%, 95% CI: 94.5% to 96.5%). One potential outlier was identified and following the removal of that study the prevalence was slightly reduced to 15.5% (95% CI: 11.6% to 9.9%) with little change to the heterogeneity (I2=95.3%, 95% CI: 94.0% to 96.3%).
Funnel plots revealed some evidence of asymmetry, but Egger’s test did not reveal any statistically significant publication bias.
Prevalence directly compared with other diagnoses
Two studies directly compared point prevalence of panic disorder in bipolar disorder with panic disorder in major depressive disorder (MDD), and six studies compared lifetime prevalence. The test for subgroup differences showed a significantly higher point prevalence of panic disorder in bipolar disorder compared to MDD (p=0.0047), but no significant difference in lifetime prevalence (p=0.82).
Subgroup analysis and meta-regression
Subgroup analysis of those with bipolar disorder type 1 and type 2 did not reveal any differences in point or lifetime prevalence estimates and did not reduce the heterogeneity.
Meta-regression analysis showed that neither age, gender or diagnostic procedure affected either the point or lifetime prevalence estimates.
The authors conclude that the prevalence of comorbid panic disorder in patients with bipolar disorder is greater than the prevalence of panic disorder in the general population, although the reported prevalence estimates varied substantially across studies. The similarity in the point prevalence and lifetime prevalence may suggest that panic disorder runs a more chronic course when comorbid with bipolar disorder.
Strengths and limitations
The study is a good quality systematic review of comorbid panic disorder in bipolar disorder. The authors used broad inclusion criteria and few exclusion criteria, which allowed a wide range of relevant papers to be identified.
Nonetheless, there are limitations that need to be considered when interpreting the results:
As the authors discuss, a large proportion of the included studies were not identified using their search strategy, but from the scanning of reference lists of relevant books and reviews, and therefore there may be additional studies that were not found. The use of only the PubMed/MEDLINE database in the search strategy may be part of the reason for this, and searches of additional databases may have yielded further studies for inclusion.
A further limitation is the high heterogeneity of the meta-analyses, which was not related to differences between bipolar types 1 and 2. The reason for such variability in the primary studies is not clear, and may be related to methodology, geographical location or diagnostic instrument. A strength of this study is the use of statistical techniques to identify outliers, but unfortunately excluding outliers did not reduce the heterogeneity. Such variability means it is not clear what the correct estimate for the prevalence of comorbid panic disorder is, and whether the pooled prevalence calculated here is applicable to a specific local population.
Additionally, very few of the identified studies directly compared prevalence of panic disorder in bipolar disorder to panic disorder in healthy controls, and therefore it is difficult to compare the prevalence estimates calculated in this meta-analysis with estimates of panic disorder prevalence in the general population.
Implications for practice
This study provided a comprehensive review of the comorbidity of panic disorder in those with bipolar disorder, finding a high lifetime and point prevalence, and comparable rates of comorbidity to MDD. The similarity of lifetime and point prevalence of panic disorder may suggest it runs more chronic rather than episodic course in those with bipolar disorder.
The evidence for treatment of panic disorder in those with bipolar disorder is severely lacking and is a key area for future research. As the presence of comorbid anxiety disorders are associated with poorer outcomes in bipolar disorder, more effective treatment for panic disorder could be one way of improving prognosis in this patient group. Therefore, clinicians should be vigilant for symptoms of panic disorder in those with bipolar disorder.
Conflicts of interest
Preti A, Vrublevska J, Veroniki AA, et al. (2018) Prevalence and treatment of panic disorder in bipolar disorder: systematic review and meta-analysis. Evid Based Ment Health, 21, 53-60.
Das A. (2013) Anxiety disorders in bipolar I mania: prevalence, effect on illness severity, and treatment implications. Indian J Psychol Med, 35, 53-9.
Duffy A, Horrocks J, Doucette S, et al. (2014) The developmental trajectory of bipolar disorder. Br J Psychiatry, 204, 122-8. [PubMed abstract]
Faedda GL, Serra G, Marangoni C, et al. (2014) Clinical risk factors for bipolar disorders: a systematic review of prospective studies. J Affect Disord, 168, 314-21. [PubMed abstract]
Nabavi B, Mitchell AJ, Nutt D. (2015) A Lifetime Prevalence of Comorbidity Between Bipolar Affective Disorder and Anxiety Disorders: A Meta-analysis of 52 Interview-based Studies of Psychiatric Population. EBioMedicine, 2, 1405-19.