ANNOTATION (Christopher Celano)

The Finding:  The authors performed a secondary analysis of data from the Heart Failure Health-Related Quality of Life Collaborative Registry to determine the prospective relationships between depressive symptoms/diabetes and mortality/cardiac rehospitalizations in 663 patients with heart failure (HF). When looking at depressive symptoms and diabetes individually, only depressive symptoms were associated with mortality and rehospitalizations; diabetes was not. However, a combination of elevated depressive symptoms (PHQ-9 score ≥ 10) and diabetes was associated with higher rates of mortality and cardiac rehospitalizations than either one alone. These relationships remained significant even when controlling for sociodeomgraphic, medical, and medication-related variables.

Strengths and Weaknesses: The main strengths of the study were its moderate sample size, its prospective design, and its systematic approach to obtaining information about cardiac rehospitalizations and mortality. Its main limitation was the likely inclusion of too many covariates (14 in the fully adjusted model) in the Cox regression analyses. This especially was the case when looking at mortality as an outcome, as only 47 individuals died during the follow-up period. Ultimately, this raises the significant possibility of overfitting in the regression analyses. Therefore, the results of the study should be interpreted with caution.

Relevance: The results of these analyses suggest that depression is strongly associated with mortality and cardiac rehospitalization in patients with heart failure. This increases the focus on depression as a potential risk factor of poor outcomes in patients with cardiac disease. While depression’s negative effects on cardiac health have been established in the post-acute coronary syndrome (post-ACS) population (leading to the American Heart Association’s 2014 recommendation that it be considered a risk factor for mortality in the post-ACS population), this manuscript extends those findings to patients with heart failure. It also raises the question of whether screening for depression in heart failure is worthwhile, and it provides support for other studies to determine whether the treatment of depression may improve cardiac outcomes in this high risk cohort. Finally, it suggests that patients with both depression and diabetes are at particularly high risk for poor cardiac outcomes and likely should be identified and closely monitored by clinicians.

ANNOTATION (Jeff Huffman)

The Finding:  The authors found that about 1/3 of ICD patients had elevated anxiety scores on the BAI, and about 1/4 had elevated scores on a measure of generalized anxiety disorder. Not surprisingly, more ICD shocks, and more recent shocks, were associated with higher anxiety. In addition, in a very small randomized pilot trial for high-anxiety ICD patients, the authors found that a low-burden, three-session CBT intervention appeared to lead to sustained beneficial effects on anxiety, compared to almost no change in anxiety in patients who were randomized to usual care.

Strengths and Weaknesses:  The main strength of the observational component of the study was systematic inclusion of a substantial number of ICD patients, with use of validated scales to assess anxiety and ICD acceptance, as well as collection of numerous variables to assess their association with anxiety. The main strength of the trial was its novelty—utilizing an intervention to help high-anxiety ICD patients—as well as the practical, low-burden nature of the intervention. The pilot trial had several important weaknesses, especially the very small sample sizes (with only 8 CBT patients and 6 usual care patients in the final primary analysis, a loss of nearly 50% of the participants), seriously limiting any conclusions that can be drawn about the intervention. Furthermore, the majority of eligible patients for the study declined (41 out of 70 declined), so in the end, of 70 eligible patients, 14 patients enrolled and provided full data.

Relevance: Anxiety in ICD patients is a sticky and complex clinical problem. Patients with ICDs are often anxious because of very real fears of the distress and pain associated with sudden firing of the ICD, and this anxiety can substantially limit function. This study confirms that a meaningful proportion of ICD patients have elevated anxiety, and their simple, practical intervention has shown some promise in reducing anxiety in this important cohort. Clearly, further study is needed to determine whether this intervention remains efficacious in larger, carefully controlled studies before such an intervention could be clinically implemented.

ANNOTATION (Scott Beach)

The Finding:
The authors examined the association of cluster disorders (stress disorders and fear disorders), individual disorders (major depression, generalized anxiety disorder (GAD), panic disordcer) and symptoms (anhedonia, negative affectivity, and anxious arousal) with major adverse cardiovascular and cerebrovascular events (MACCE) in the five years following coronary artery bypass graft surgery. They found that baseline generalized anxiety disorder, but not other distress disorders such as major depression or fear disorders such as panic disorder, was associated with increased rates of MACCE. They also found that specific negative symptoms were not independently associated with MACCE.

Strengths and Weaknesses:  The main strength of the study is the delineation of disorders at the diagnostic and cluster level, as well as the measurement of symptom dimensions. Other strengths included the use of standardized screening tools, a relatively long follow-up period (mean 4.6 years), and inclusion of medical covariates in the analyses. Weaknesses of the study included a racially and ethnically homogenous sample, assessment of psychiatric disorders prior to surgery with no reassessment during the study period, relatively low total number of MACCE, and lack of information about psychiatric treatment received. Oddly, according to the structured interview, GAD could not be diagnosed in the presence of major depression (despite high comorbidity between these disorders), so patients having comorbid depression and anxiety were not included in the GAD analysis. Additionally, the prevalence of generalized anxiety disorder and panic disorder was nearly 5 times higher than in the general population, raising the question of whether these entities were overdiagnosed in the setting of significant pre-surgical anxiety.

Relevance: Negative emotions have been associated with MACCE, though it has remained unclear whether the association pertains most to specific disorders, cluster disorders, or specific symptoms. The findings here suggest that generalized anxiety disorder is specifically associated with MACCE, whereas other diagnoses and symptoms are not. Psychiatric interventions targeting GAD may therefore be indicated for patients undergoing CABG, though further study is needed to determine whether such treatments reduce the risk of MACCE in this population.