Preventive Medicine 63 (2014) 36–42

Contents lists available at ScienceDirect

Preventive Medicine
journal homepage: www.elsevier.com/locate/ypmed

Review

Depression after heart failure and risk of cardiovascular and all-cause

mortality: A meta-analysis
Hongjie Fan a, Weidong Yu b, Qiang Zhang c, Hui Cao c, Jun Li c, Junpeng Wang c, Yang Shao c, Xinhua Hu c,⁎
a
Department of Neurology, Shengjing Hospital of China Medical University, Shenyang 110004, China
b
Department of Geriatrics, the First Affiliated Hospital of China Medical University, Shenyang 110001, China
c
Department of Surgery, the First Affiliated Hospital of China Medical University, Shenyang 110001, China

a r t i c l e i n f o a b s t r a c t

Available online 13 March 2014 Objectives. The aim of this study is to investigate whether depression after heart failure (HF) was a predictor
for subsequent cardiovascular and all-cause mortality in prospective observational studies.
Keywords: Methods. Pubmed, Embase, and PsycInfo databases were searched for prospective studies reported depression
Depression after HF and subsequent risk of cardiovascular or all-cause mortality (prior to May 2013). Pooled adjust hazard
Heart failure
ratio (HR) and corresponding 95% confidence intervals (CI) were calculated separately for categorical risk estimates.
Cardiovascular mortality
All-cause mortality
Results. Nine studies with 4012 HF patients were identified and analyzed. Pooled HR of all-cause mortality was
Meta-analysis 1.51 (95% CI 1.19–1.91) for depression compared with non-depressive patients. Subgroup analyses showed that
major depression significantly increased all-cause mortality (HR = 1.98, 95% CI 1.23–3.19), but not mild depression
(HR = 1.04, 95% CI 0.75–1.45). Pooled HR of cardiovascular mortality was 2.19 (95% CI 1.46–3.29) for depression
compared with non-depressive patients.
Conclusion. Major depression after HF was a predictor for subsequent all-cause mortality, but not mild depres-
sion. More well-designed studies are needed to explore the influence of depression and antidepressant medication
use on cardiovascular and all-cause mortality in HF patients.
© 2014 Elsevier Inc. All rights reserved.

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Patients and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Literature search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Study selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Data extraction and quality assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Statistical analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Literature search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Study characteristics and quality assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
All-cause mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Cardiovascular mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Subgroup analyses and sensitivity analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

⁎ Corresponding author at: Department of Surgery, the First Affiliated Hospital of China Medical University, 155 Nanjing Street, Heping District, Shenyang 110001, China.
Fax: + 86 24 83283288.
E-mail address: xinhuahu@126.com (X. Hu).

http://dx.doi.org/10.1016/j.ypmed.2014.03.007
0091-7435/© 2014 Elsevier Inc. All rights reserved.
 H. Fan et al. / Preventive Medicine 63 (2014) 36–42
Introduction
Heart failure (HF) is a clinical syndrome that is increasing in preva-
lence and incidence worldwide. Despite substantial advances in its
treatment, morbidity and mortality remain high (Dickstein et al.,
2008; Veien et al., 2011). Therefore, early detection of high-risk patients
would facilitate preventing HF and would further lower HF mortality
(Nair et al., 2012). Depressive symptoms are common among patients
with HF, and the prevalence of depressive symptoms in patients with
HF ranges from 9% to 60% (Rutledge et al., 2006). Both major depression
and the presence of depressive symptoms have been reported to
increase the risks of mortality and other adverse outcomes in patients
with coexistence of HF (Faris et al., 2002; Jiang et al., 2004; Junger
et al., 2005; Murberg and Furze, 2004; Rumsfeld et al., 2005).
Based on the previously published clinical evidence, a well designed
meta-analysis (Rutledge et al., 2006) showed that depressive symptoms
or a depressive disorder increased 2-fold risk of combined endpoints
(death and secondary events) in patients with HF, and the prevalence
of depression among patients with HF was 21.5%. However, there
were high heterogeneity with respect to the composition of the
samples, study design and methods used to assess depression. After
that, more prospective studies (Adams et al., 2012; Faller et al., 2007;
Kato et al., 2009; Lesman-Leegte et al., 2009; Moraska et al., 2013;
O'Connor et al., 2008; Rollman et al., 2012; van den Broek et al., 2011;
Zuluaga et al., 2010) have been published addressing the association
between depression status after heart failure and subsequent risk of
mortality. To the best of our knowledge, no meta-analysis of such
studies has been conducted on the association between depression
after HF and subsequent risk of cardiovascular and all-cause mortality.
Conflicting results whether depressive symptoms are an independent
risk factor for cardiovascular and all-cause mortality remained (Faller
et al., 2007; Moraska et al., 2013; Zuluaga et al., 2010).
Given above reasons, a meta-analysis may help clarify this issue. The
aim of the current meta-analysis was to evaluate findings from the
available prospective studies on depression after HF and subsequent
risk of mortality, and determine whether depression after HF was a
predictor of subsequent cardiovascular and all-cause mortality.
Patients and methods
Literature search
We conducted a literature research through Pubmed, Embase, and PsycInfo
databases (prior to May 2013) for studies reporting the association between
depression status after heart failure and subsequent risk of cardiovascular and
all-cause mortality. Only papers published in English language were considered.
Potentially relevant studies included the word ‘depression’, ‘depressive’ plus at
least one of the following terms: mortality/heart failure and death/heart failure;
prospective and follow-up. In addition, we also manually searched the reference
lists of all identified relevant publications to detect additional eligible
studies.
Study selection
Criteria for papers to be included in the current meta-analysis consisted of
1) describing prospective relationships between depression status after HF
diagnosis and cardiovascular and all-cause mortality; 2) providing adjusted
hazard ratio (HR) and the 95% confidence interval (CI) dealing with the risk of
cardiovascular and all-cause mortality with depressive patients compared
with non-depressive individuals; and 3) follow-up duration of at least 1 year.
For multiple publications in the same research group, only the most recent
publication was included. The definition of major and mild depression was
defined by individual studies by the different depression scales (Table 1).
Studies were excluded if 1) a case–control study or retrospective study;
2) unadjusted HR was reported; and 3) describing results as continuous or
quantitative scores without any dichotomization around a standardized cut-off
value for depression.
37
Data extraction and quality assessment
Two reviewers (WD Yu and HJ Fan) independently extracted the data from
each study. The most fully adjusted HR and 95% CI were extracted. We also
extracted the following items from individual study: author; year of publica-
tion; region; depression definitions and measures; antidepressants treatment;
time of follow-up; the sample size, gender, and age of patients; death events;
follow-up duration, and statistical adjustments for confounding factors. Quality
assessment was performed with the following checklists based on the Meta-
analysis of Observational Studies in Epidemiology guidelines (Stroup et al., 2000).
Statistical analyses
Data analyses used the most fully adjusted HR and 95% CI. For papers provid-
ing both major and minor depression data, we pooled the separate data based
on the major depression and minor depression category. Homogeneity of HR
across studies was assessed using the Cochrane Q statistic (p b 0.10 was indicat-
ed significant heterogeneity) and I2 statistic (values of more than 50% was
considered significant heterogeneity) (Higgins et al., 2003). As there was sub-
stantial heterogeneity in the types of depression and diagnosis between the
different studies, a random effects model was used to calculate the pooled HR.
The possibility of publication bias was assessed by Begg's rank correlation test
(Begg and Mazumdar, 1994) and Egger linear regression test at p b 0.10 (Egger
et al., 1997). Finally, sensitivity analyses were carried out by sequentially
omitting one study at each turn. All analyses were conducted using STATA
version 12.0 (Stata Corp LP, College Station). P b 0.05 was considered as statistical-
ly significant.
Results
Literature search
Following the application of the predefined search strategy, a total of
593 relevant papers were identified in our literature search. After
screening the abstracts or titles, 541 studies were excluded because
they were reviews, retrospective studies, or not relevant to our review.
After reading the full texts, nine studies (Adams et al., 2012; Faller et al.,
2007; Junger et al., 2005; Kato et al., 2009; Lesman-Leegte et al., 2009;
Moraska et al., 2013; Rollman et al., 2012; van den Broek et al., 2011;
Zuluaga et al., 2010) were satisfied the inclusion/exclusion. Fig. 1
presented a flow chart of the study selection.
Study characteristics and quality assessment
Nine studies with 4012 HF patients (1652 with and 2360 without
depressive subjects) were identified and analyzed. The year of publica-
tion ranged from 2005 to 2013. All articles were in English. The follow-
up duration ranged between 12 months and 11 years. Among these 9
articles, the assessment of depression varied across studies, with Patient
Health Questionnaire (PHQ) (Faller et al., 2007; Moraska et al., 2013;
Rollman et al., 2012), Center for Epidemiologic Studies Depression
Scale (CES-D) (Kato et al., 2009; Lesman-Leegte et al., 2009; van den
Broek et al., 2011), Beck Depression Inventory (BDI) (Adams et al.,
2012; O'Connor et al., 2008), Geriatric Depression Scale (GDS) (Zuluaga
et al., 2010),and Hospital Anxiety and Depression Scale (HADS) (Junger
et al., 2005). The characteristics of the included studies were listed in
Table 1. Supplement Table S1 presented the qualities of the included
studies. All the included studies stated clear inclusion and exclusion
criteria, clear definition of outcome, adjusted important confounders,
and performed the appropriate statistics. However, most of studies
were followed less than 5 years.
All-cause mortality
Seven studies (Adams et al., 2012; Junger et al., 2005; Kato et al.,
2009; Moraska et al., 2013; Rollman et al., 2012; van den Broek et al.,
2011; Zuluaga et al., 2010) reported on all-cause mortality for depres-
sion. The total number of participants included in this meta-analysis
 38
Table 1
Summary of clinical studies included in the meta-analysis.

Study/year Country Design Etiology/number Subjects Mean age, Depression Follow-up Outcome/events number/HR Adjustment for covariates
(%men) year (SD) measure/rate (95% CI)

Junger Germany Prospective CAD 55 209 (13.9) 54 (10) HADS ≥8 Mean Total death 45 NYHA class, LVEF and peak VO2
et al, 2005 study DC 144 Total: 63 (30.1) 2 years 1.08 (1.01–1.15)
Faller Germany Prospective Ischemic 99 231 (29.4) 64 (13) German version of PHQ-9 Median Total death 59 Age, sex, HF etiology, degree and
et al, 2007 cohort study Others 132 with 2 to 4 symptoms, and 2.7 years 2.40 (1.3 –4.6) Ma type of left ventricular dysfunction,
≥5 symptoms indicating Mi and NYHA.
(38) and Ma depression (31).
Total: 69 (29.9)
Lesman-Leegte The Netherlands Prospective NP 958 (37) 71 (11) 20-item CES-D with score 1.5 years Total death 259 Age, sex, and BNP.
et al, 2009 study 16–23 and ≥24 indicating 1.43 (1.02–2.02) Ma

H. Fan et al. / Preventive Medicine 63 (2014) 36–42

Mi (177) and Ma depression
(200). Total: 377 (39.4)
Kato Japan Prospective Ischemic 32 115 (26.1) 64.7 (15.7) 20-item CES-D ≥16 Median Total death 13 Age, BNP, and other potential
et al, 2009 cohort study Total: 27 (23.5%) 756 days 5.52 (1.65– 18.46) prognostic factors.
Zuluaga Spain Prospective Ischemic154 433 (56.4) 77.1 (6.9) N 10-items GDS with 3 to 4 Mean Total death 305 Age, sex, COPD, Charlson index,
et al, 2010 study Hypertensive 233 78.5 (6.5) Mi symptoms, and ≥5 5.7 years 0.93 (0.69–1.26) Mi creatinine, NYHA class, LVEF, smoking,
Heart valve 102 77.5 (6.6) Ma symptoms indicating 1.10 (0.82–1.49) Ma alcohol, physical activity,
Others 151 Mi (107) and Ma pharmacologic
depression (103). treatment, hospitalization for HF in
Total: 210 (48.5) last year, and possible etiology of HF.
Van den Broek The Netherlands Prospective, CHD 130 208 (51) 75.2 (6.1) 10-item CES-D ≥8 Median 1.49 (1.05–2.11) Age, gender, and race, SBP, cholesterol,
et al, 2011 community- based Others 78 Total: 75 (36.1%) 11 years Total death 168 DB, BMI, smoking, reduced physical
study 2.07 (1.31–3.27) activity, CHD at baseline, LVEF, and left
CVD death 97 ventricular hypertrophy
Rollman Pittsburgh, PA Prospective NP 471 (35.5) 69.6 (11.4) N Two-item PHQ-2 1 year 3.1 (1.4–6.7) Age, sex, LVEF, NYHA class, renal
et al, 2012 study 65.0 (13.4) Depression Total: 371 (78.8) Total death 83 insufficiency, anemia, DB, hyponatremia,
2.7 (1.1–6.6) SBP, DBP, use ACE-I, antidepressants,
CVD death 55 angiotensin receptor blocker
Adams USA Prospective Ischemic 597 985 (37.78) 69.07 (10.60) BDI ≥10 Median 1.40 (1.16–1.68) Age, NYHA score, HF etiology, history
et al, 2012 study Others 377 Total: 294 (29.8) 4.38 years Total death 731 of CABG, DB, EF, and marital status.
Moraska USA Prospective NP 402 (42.3) 73.3 (13.2) 9-items PHQ-9 scores 1.6 years Total death 74 Age, sex, Charlson comorbidity
et al, 2013 study with score 5–9 and ≥10 1.59 (0.89–2.83) Mi index, and incident vs. prevalent
indicating Mi (104) 4.06 (2.35–7.01) Ma heart failure status
and Ma depression (62).
Total: 166 (41.3)

Abbreviations: HF, Heart failure; CHD, coronary heart disease; BMI, body mass index; HR, hazard risk; OR, odds ratio; SBP, systolic blood pressure; DBP, diastolic blood pressure; BDI, Beck Depression Inventory; CABG, coronary artery bypass graft;
NYHA, New York Heart Association functional class; PHQ, Patient Health Questionnaire; HADS, Hospital Anxiety and Depression Scale; DC, dilated cardiomyopathy; LVEF, left ventricular ejection fraction; GDS, geriatric depression scale, COPD, chronic
obstructive pulmonary disease; BNP, brain natriuretic peptide; DB, diabetes mellitus; CES-D, Center for Epidemiologic Studies Depression Scale; PRIME-MD, Primary Care Evaluation of Mental Disorders; ACEI, angiotensin-converting enzyme
inhibitor.
N, normal; Mi, mild depression; Ma, major depression.
 H. Fan et al. / Preventive Medicine 63 (2014) 36–42 39

Fig. 1. Flow chart of study selection process for meta-analysis.

Fig. 2. HR and 95% CI from the included studies of depression with all-cause mortality compared with non-depressed patients in a random effects model.
40 H. Fan et al. / Preventive Medicine 63 (2014) 36–42
was 2823, with 1334 depressive patients. As shown in Fig. 2, depressive
symptom was associated with an increase in all-cause mortality in
a random effects model compared with non-depressive patients
(HR = 1.51, 95% CI 1.19–1.91). Substantial heterogeneity was obvious
(I2 = 82.7%; P = 0.000). Evidence of publication bias for studies
reporting HR of all-cause mortality was noted in Egger's linear regression
test (P = 0.021), and in the Begg's rank correlation test (P = 0.076).
Four studies (Faller et al., 2007; Lesman-Leegte et al., 2009; Moraska
et al., 2013; Zuluaga et al., 2010) reported on all-cause mortality for mild
depression and major depression. Another early study (Jiang et al.,
2001) from the same population of Van den Broek's publication (van
den Broek et al., 2011) provided a risk estimate of mortality for mild
depression and major depression. As shown in Fig. 3A, major depression
was associated with an increase in all-cause mortality compared with
non-depressive patients (HR = 1.98, 95% CI 1.23–3.19). Substantial
heterogeneity was observed (I2 = 80.5%; P = 0.000). As shown in
Fig. 3B, mild depression was not associated with an increase in all-
cause mortality compared with non-depressive patients (HR = 1.04,
95% CI 0.75–1.45). Substantial heterogeneity was not obvious (I2 =
31.1%; P = 0.234).
Cardiovascular mortality
Two studies (Rollman et al., 2012; van den Broek et al., 2011) reported
on cardiovascular mortality for depression. The total number of
participants included in this meta-analysis was 679, with 446 depressive
patients. As shown in Fig. 4, depression was associated with an increase
in cardiovascular mortality in a random effects model compared with
non-depressive patients (HR = 2.19, 95% CI 1.46–3.29). Substantial
heterogeneity was not observed (I2 = 0%; P = 0.605).
Subgroup analyses and sensitivity analyses
Depression was associated with an increased risk of mortality in
most subgroups except for mean age N70, follow-up duration N 4 years,
and sample sizes b400 subgroup. The increased risk was more evident
in several strata of study characteristics. Detail results stratified by
Fig. 3. HR and 95% CI from the included studies of major (A) and mild (B) depression of all-cause mortality compared with non-depressed patients in a random effect model.
characteristics of study design were listed in the Table 2. Sensitivity
analyses were performed based on all-cause mortality. In the sensitivity
analysis, there was little influence in the quantitative pooled measure of
HR or 95% CI when omission of anyone studies as shown in Supplement
Fig. S1–9. The pooled HR varied from 1.31 to 1.68 and low 95% CI varied
from 1.08 to 1.26. These results indicated that our conclusion was reliable.
Discussion
Findings of the current meta-analysis suggested that depressive
mood after HF diagnosis increased the risk of future cardiovascular
and all-cause mortality. Subjects with major depression increased 51%
risk of all-cause mortality after adjustment for potential confounding
factors. Subgroup analyses based on the severity of depression showed
that this association was increased 98% risk of all-cause mortality,
but not mild depression. Our results were in accordance with a well-
designed reviews' conclusion that major (or severe) depression is a
stronger predictor of mortality than is minor (or mild) depression
(Freedland et al., 2011).
Depression is more common in younger patients than in older
patients. In the current meta-analysis, subgroup analyses based on age
(N70 vs. b70) indicated that depressive patients with age b70 year
were associated with increased mortality, while in those depressive
patients with age N 70 years showed some trend. However, the age
and gender of patients in the included studies varied in these studies,
and it might influence the findings of the meta-analysis. Women have
more severe depressive episodes and are more likely to develop chronic
depression compared with men in the general population (Ernst and
Angst, 1992; Kornstein et al., 1995). Depressive symptoms were associ-
ated with 68%–96% increase in female gender among HF patients
(Lesman-Leegte et al., 2006; Williams et al., 2002). In another study
(Faller et al., 2007) addressing the gender differences on the prognostic
impact of depression, men with major depression was not predictive in
the multivariable analyses (HR = 2.1, 95% CI 0.9–4.6), while increasing
the prognostic relevance in women (HR = 4.5, 95% CI 1.3–15.8). More-
over, due to the limited number of studies, we were unable to conduct
subgroup analyses based on gender. Therefore, whether men or women
 H. Fan et al. / Preventive Medicine 63 (2014) 36–42 41

Fig. 4. HR and 95% CI from the included studies of depression with cardiovascular mortality compared with non-depressed patients in a random effects model.

with depression had more risk of mortality is still unclear. The relation- that HR of mortality was 1.04 for each 1-unit increase BDI scores of
ship between depression and mortality seems to be reduced to the dura- 291 patients with chronic heart failure. Therefore, the robustness of
tion of observation. Subgroup analysis of short term studies (b 4 years) depression as a risk factor was further reinforced by the clear relationship
showed a statistically significant association between depression and of depression as a continuous variable (per 1 unit score increase) with all
mortality, whereas subgroup analysis of long term studies (N 4 years) cause mortality. Change in symptoms of depression, as indicated by BDI
showed some trend. One possible explanation for the current find- scores change 1-point over 1-year interval was associated with increasing
ings is that medical management for HF has improved prognosis. death or cardiovascular hospitalization (HR = 1.07, 95% CI 1.02–1.12)
Given above reason, the generalizability of these findings is limited, (Sherwood et al., 2011). Our results were also consistent with a large ret-
as for the population studied varied with respect to age, gender, rospective study (Macchia et al., 2008) of 48,117 patients in a community
and follow-up duration. setting; the investigators found that depression significantly increased
The previous meta-analysis (Rutledge et al., 2006) was the first all-cause mortality (HR = 1.20, 95% CI 1.08–1.33). In addition, depression
meta-analysis concerning depression prevalence, associations with was also associated with increased risk of hospitalization (Johnson et al.,
clinical outcomes, and changes resulting from treatment interventions 2012; Sherwood et al., 2011).
in HF patients. However, this well-designed meta-analysis only summa- Increasing evidences suggested that screening and treatment of
rized the combined endpoints including mortality in combination with depressive symptoms were very important in the management of HF
cardiac events, not particularly focused on mortality. Our meta-analysis (Lichtman et al., 2008). However, the efficacy and safety of antidepres-
mainly addressed the impact of depression on the subsequent mortality sant therapy in patients with HF are still controversial. Conversely,
in subjects after HF diagnosis. Many studies that did not meet the evidence from observational studies suggested a 49% increased mortality
inclusion criteria for the meta-analysis also found a positive association risk in relation to treatment with antidepressants in systolic HF patients
between depression and mortality in HF patients. Using a continuous (Veien et al., 2011). It is also possible that the side effects of certain
variable of the Zung Self-rating Depression Scale (SDS) score, Murberg antidepressants have contributed to the unexpected increased risk
et al (Murberg and Furze, 2004) found that HR of mortality for per for mortality. Therefore, the role of antidepressant medication use in
1-point increase in SDS score was 1.05 based on the multivariate model modulating subsequent risk of mortality needs to be further studied.
among 119 patients with HF. Another study (Jiang et al., 2004) showed Exact mechanisms linking depression with cardiovascular and all-
cause mortality are not fully elucidated. Possible explanatory mecha-
nisms to link depression and mortality in CHF may include health
behavior and biological factors (Mann and Bristow, 2005; Pelle et al.,
Table 2 2008). Several possible explanations are as follows: depression is corre-
Subgroup analyses of depression and risk of all cause mortality. lated with other major comorbidities, such as diabetes (Park et al.,
2013) and hypertension (Patten et al., 2009), both of which are major
Group Number Death/depression/heart Pooled HR 95% CI
of studies failure number risk factors for mortality. Neurohormonal imbalance, immune and
inflammatory activation, and distorted autonomic functioning play a
Region
Asia 1 13/27/115 5.52 1.65–18.46
role in the development of both depression and progression of HF
No Asia 6 1406/1487/2708 1.43 1.14–1.80 (Parissis et al., 2005). Furthermore, depression is associated with
Mean age increased circulating Inflammatory factors levels (Miller et al., 2002).
N70 years 3 547/670/1043 1.51 0.99–2.31 Inflammatory processes play an important role in the pathophysiologi-
≤70 years 4 872/755/1780 1.55 1.09–2.21
cal pathways accounting for increased risk of HF progression among
Durations
N4 years 3 1204/579/1626 1.22 0.99–1.50 depressed individuals (Pasic et al., 2003).
≤4 years 4 215/935/1197 2.39 1.19–4.79 There are several potential limitations in this study. First, a major
Sample size limitation was the possibility of uncontrolled confounding, and the indi-
N400 4 1193/1349/2291 1.60 1.12–2.29 vidual studies did not adjust for potential risk factors in a consistent
≤400 3 226/165/532 1.49 0.92–2.41
way. The lack of adjustment for these confounding factors might have
42 H. Fan et al. / Preventive Medicine 63 (2014) 36–42
resulted in a slight overestimation of the HR. Second, another limitation
of the current study is that measuring depression at the initial diagnosis
did not reflect the severity of the syndrome at the end because the
course of depression is highly variable, and therefore, have not consid-
ered the effect of changes over time. Third, extreme heterogeneity in
the definition of depression and the neuropsychological tests used
was also an important limitation. Furthermore, depression assessments
were mainly based on self-reported questionnaires; therefore, selection
bias was inevitable due to some misdiagnosis of depression. Fourth,
there were only two studies (Rollman et al., 2012; van den Broek
et al., 2011) in the cardiovascular mortality meta-analysis, so the reli-
ability and generalizability should be cautioned. Subgroup analyses
will further explain the source of heterogeneity. However, we could
not conduct further deepen our analyses of the data, such as by gender
and age. In addition, the length of follow-up in studies (ranges from 1 to
11 years) is an additional limitation. It is difficult to determine beyond
the duration of the follow-up studies in the meta-analysis with respect
to long-term impact on cardiovascular and all-cause mortality. Finally,
we did not contact the authors of papers reporting on continuous
depression scores to recalculate based on dichotomous depression;
thus, we only included adjusted risk estimates in our analysis, and
some important studies might be overlooked.
Conclusions
This meta-analysis provides evidence that major depression after
HF was a predictor for subsequent all-cause mortality, but not in mild
depression. In addition, depressive mood after HF appeared to be a
predictor for cardiovascular mortality. Clinicians should regularly
evaluate symptoms of depression in patients with diagnosed HF. Routine
evaluation depressive mood after heart failure will be helpful in identify-
ing mortality risk in patients with HF. However, more well-designed
studies are needed to explore the influence of antidepressant medication
use on cardiovascular and all-cause mortality in HF patients.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.ypmed.2014.03.007.
Conflict of interest
None declared.
Acknowledgments
We gratefully thank Shijun Wang for his kind help in proofreading.
This work was supported by the Nature Science Foundation of China
(81070254, 30872527).
References
Adams, J., Kuchibhatla, M., Christopher, E.J., et al., 2012. Association of depression and
survival in patients with chronic heart failure over 12 years. Psychosomatics 53,
339–346.
Begg, C.B., Mazumdar, M., 1994. Operating characteristics of a rank correlation test for
publication bias. Biometrics 50, 1088–1101.
Dickstein, K., Cohen-Solal, A., Filippatos, G., et al., 2008. ESC guidelines for the diagnosis
and treatment of acute and chronic heart failure 2008: the Task Force for the diagno-
sis and treatment of acute and chronic heart failure 2008 of the European Society of
Cardiology. Developed in collaboration with the Heart Failure Association of the ESC
(HFA) and endorsed by the European Society of Intensive Care Medicine (ESICM).
Eur. J. Heart Fail. 10, 933–989.
Egger, M., Davey Smith, G., Schneider, M., Minder, C., 1997. Bias in meta-analysis detected
by a simple, graphical test. BMJ 315, 629–634.
Ernst, C., Angst, J., 1992. The Zurich Study. XII. Sex differences in depression. Evidence
from longitudinal epidemiological data. Eur. Arch. Psychiatry Clin. Neurosci. 241,
222–230.
Faller, H., Stork, S., Schowalter, M., et al., 2007. Depression and survival in chronic heart
failure: does gender play a role? Eur. J. Heart Fail. 9, 1018–1023.
Faris, R., Purcell, H., Henein, M.Y., Coats, A.J., 2002. Clinical depression is common and
significantly associated with reduced survival in patients with non-ischaemic heart
failure. Eur. J. Heart Fail. 4, 541–551.
Freedland, K.E., Carney, R.M., Rich, M.W., 2011. Effect of depression on prognosis in heart
failure. Heart Fail. Clin. 7, 11–21.
Higgins, J.P., Thompson, S.G., Deeks, J.J., Altman, D.G., 2003. Measuring inconsistency in
meta-analyses. BMJ 327, 557–560.
Jiang, W., Alexander, J., Christopher, E., et al., 2001. Relationship of depression to increased
risk of mortality and rehospitalization in patients with congestive heart failure. Arch.
Intern. Med. 161, 1849–1856.
Jiang, W., Kuchibhatla, M., Cuffe, M.S., et al., 2004. Prognostic value of anxiety and
depression in patients with chronic heart failure. Circulation 110, 3452–3456.
Johnson, T.J., Basu, S., Pisani, B.A., et al., 2012. Depression predicts repeated heart failure
hospitalizations. J. Card. Fail. 18, 246–252.
Junger, J., Schellberg, D., Muller-Tasch, T., et al., 2005. Depression increasingly
predicts mortality in the course of congestive heart failure. Eur. J. Heart Fail. 7,
261–267.
Kato, N., Kinugawa, K., Yao, A., Hatano, M., Shiga, T., Kazuma, K., 2009. Relationship of
depressive symptoms with hospitalization and death in Japanese patients with
heart failure. J. Card. Fail. 15, 912–919.
Kornstein, S.G., Schatzberg, A.F., Yonkers, K.A., et al., 1995. Gender differences in
presentation of chronic major depression. Psychopharmacol. Bull. 31, 711–718.
Lesman-Leegte, I., Jaarsma, T., Sanderman, R., Linssen, G., van Veldhuisen, D.J., 2006.
Depressive symptoms are prominent among elderly hospitalised heart failure
patients. Eur. J. Heart Fail. 8, 634–640.
Lesman-Leegte, I., van Veldhuisen, D.J., Hillege, H.L., Moser, D., Sanderman, R., Jaarsma, T.,
2009. Depressive symptoms and outcomes in patients with heart failure: data from
the COACH study. Eur. J. Heart Fail. 11, 1202–1207.
Lichtman, J.H., Bigger Jr., J.T., Blumenthal, J.A., et al., 2008. Depression and coronary
heart disease: recommendations for screening, referral, and treatment: a science
advisory from the American Heart Association Prevention Committee of the
Council on Cardiovascular Nursing, Council on Clinical Cardiology, Council on
Epidemiology and Prevention, and Interdisciplinary Council on Quality of Care
and Outcomes Research: endorsed by the American Psychiatric Association.
Circulation 118, 1768–1775.
Macchia, A., Monte, S., Pellegrini, F., et al., 2008. Depression worsens outcomes in elderly
patients with heart failure: an analysis of 48,117 patients in a community setting. Eur.
J. Heart Fail. 10, 714–721.
Mann, D.L., Bristow, M.R., 2005. Mechanisms and models in heart failure: the biomechanical
model and beyond. Circulation 111, 2837–2849.
Miller, G.E., Stetler, C.A., Carney, R.M., Freedland, K.E., Banks, W.A., 2002. Clinical
depression and inflammatory risk markers for coronary heart disease. Am. J.
Cardiol. 90, 1279–1283.
Moraska, A.R., Chamberlain, A.M., Shah, N.D., et al., 2013. Depression, healthcare utilization,
and death in heart failure: a community study. Circ. Heart Fail. 6, 387–394.
Murberg, T.A., Furze, G., 2004. Depressive symptoms and mortality in patients with con-
gestive heart failure: a six-year follow-up study. Med. Sci. Monit. 10, CR643–CR648.
Nair, N., Farmer, C., Gongora, E., Dehmer, G.J., 2012. Commonality between depression
and heart failure. Am. J. Cardiol. 109, 768–772.
O'Connor, C.M., Jiang, W., Kuchibhatla, M., et al., 2008. Antidepressant use, depression,
and survival in patients with heart failure. Arch. Intern. Med. 168, 2232–2237.
Parissis, J.T., Fountoulaki, K., Paraskevaidis, I., Kremastinos, D., 2005. Depression in chronic
heart failure: novel pathophysiological mechanisms and therapeutic approaches.
Expert Opin. Invest. Drugs 14, 567–577.
Park, M., Katon, W.J., Wolf, F.M., 2013. Depression and risk of mortality in individuals
with diabetes: a meta-analysis and systematic review. Gen. Hosp. Psychiatry 35,
217–225.
Pasic, J., Levy, W.C., Sullivan, M.D., 2003. Cytokines in depression and heart failure.
Psychosom. Med. 65, 181–193.
Patten, S.B., Williams, J.V., Lavorato, D.H., Campbell, N.R., Eliasziw, M., Campbell, T.S., 2009.
Major depression as a risk factor for high blood pressure: epidemiologic evidence
from a national longitudinal study. Psychosom. Med. 71, 273–279.
Pelle, A.J., Gidron, Y.Y., Szabo, B.M., Denollet, J., 2008. Psychological predictors of prognosis
in chronic heart failure. J. Card. Fail. 14, 341–350.
Rollman, B.L., Herbeck Belnap, B., Mazumdar, S., et al., 2012. A positive 2-item Patient
Health Questionnaire depression screen among hospitalized heart failure patients is
associated with elevated 12-month mortality. J. Card. Fail. 18, 238–245.
Rumsfeld, J.S., Jones, P.G., Whooley, M.A., et al., 2005. Depression predicts mortality and
hospitalization in patients with myocardial infarction complicated by heart failure.
Am. Heart J. 150, 961–967.
Rutledge, T., Reis, V.A., Linke, S.E., Greenberg, B.H., Mills, P.J., 2006. Depression in heart
failure a meta-analytic review of prevalence, intervention effects, and associations
with clinical outcomes. J. Am. Coll. Cardiol. 48, 1527–1537.
Sherwood, A., Blumenthal, J.A., Hinderliter, A.L., et al., 2011. Worsening depressive
symptoms are associated with adverse clinical outcomes in patients with heart
failure. J. Am. Coll. Cardiol. 57, 418–423.
Stroup, D.F., Berlin, J.A., Morton, S.C., et al., 2000. Meta-analysis of observational studies
in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies
in Epidemiology (MOOSE) group. JAMA 283, 2008–2012.
van den Broek, K.C., Defilippi, C.R., Christenson, R.H., Seliger, S.L., Gottdiener, J.S., Kop, W.J.,
2011. Predictive value of depressive symptoms and B-type natriuretic peptide for
new-onset heart failure and mortality. Am. J. Cardiol. 107, 723–729.
Veien, K.T., Videbaek, L., Schou, M., Gustafsson, F., Hald-Steffensen, F., Hildebrandt, P.R.,
2011. High mortality among heart failure patients treated with antidepressants. Int.
J. Cardiol. 146, 64–67.
Williams, S.A., Kasl, S.V., Heiat, A., Abramson, J.L., Krumholz, H.M., Vaccarino, V., 2002.
Depression and risk of heart failure among the elderly: a prospective community-
based study. Psychosom. Med. 64, 6–12.
Zuluaga, M.C., Guallar-Castillon, P., Rodriguez-Pascual, C., Conde-Herrera, M., Conthe, P.,
Rodriguez-Artalejo, F., 2010. Mechanisms of the association between depressive
symptoms and long-term mortality in heart failure. Am. Heart J. 159, 231–237.