Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Sep 1.
Published in final edited form as: Health Psychol. 2017 Jul 20;36(9):839–847. doi: 10.1037/hea0000529

The Impact of Medication Non-Adherence on the Relationship Between Mortality Risk and Depression in Heart Failure

Emily C Gathright 1, Mary A Dolansky 2, John Gunstad 1, Joseph D Redle 3, Richard Josephson 4,5, Shirley M Moore 2, Joel W Hughes 1
PMCID: PMC5573609  NIHMSID: NIHMS893495  PMID: 28726471

Abstract

Objective

Heart failure affects over 5 million U.S. adults and approximately 20% of individuals with heart failure experience depressive symptoms. Depression is detrimental to prognosis in heart failure, conferring approximately a 2-fold increase in mortality risk. Medication non-adherence may help explain this relationship, as depressed patients are less likely to adhere to the medication regimen.

Methods

Depression, electronically-monitored medication adherence, and mortality were measured in a sample of 308 patients with heart failure participating in a study of self-management behavior. Cardiovascular and all-cause mortality data were obtained from the Centers for Disease Control and Prevention’s National Death Index (median 2.9 year follow-up). Cox proportional hazards regression was used to assess the relationship between depression and mortality, with and without adjustment for age, gender, disease severity, and medication non-adherence.

Results

In adjusted analyses, depression was associated with increased all-cause mortality risk, (HR: 1.87; 95% CI: 1.04 – 3.37). Depression was not related to cardiovascular mortality, potentially due to a low number of cardiac-related deaths. When medication non-adherence was added to the model, non-adherence (HR: 1.01; 95% CI: 1.004 – 1.02), but not depression, predicted all-cause mortality risk.

Conclusions

Depressive symptoms confer increased all-cause mortality risk in heart failure, and medication non-adherence contributes to this relationship. Depression and non-adherence represent potentially modifiable risk factors for poor prognosis. Future research is needed to understand whether interventions that concomitantly target these factors can improve outcomes.

Keywords: heart failure, depression, mortality, adherence

In the United States, 5.7 million persons ≥ 20 years of age have heart failure (HF) and 870,000 new HF cases are diagnosed annually (AHA, 2015). The prevalence of HF will increase 46% by 2030, so that > 8 million U.S. adults ≥ 18 years of age will have HF (AHA, 2015; Heidenreich et al, 2013). Despite efforts to improve outcomes, mortality in HF remains high. The 1-year mortality rate for Medicare beneficiaries with HF from 1998 to 2008 was 29.6% (Chen, Normand, Wang, & Krumholz, 2011). Demographic and medical predictors of mortality in HF include older age, diabetes, lower ejection fraction, higher New York Heart Association (NYHA) class, cardiomegaly, prior HF hospitalization, lower body mass index, and lower diastolic blood pressure (Pocock et al., 2006). After accounting for these risk factors, depression also confers increased risk of mortality in HF (Sokoreli, de Vries, Pauws, & Steyerberg, 2016).

Approximately one-fifth of HF patients experience clinically significant symptoms of depression symptoms during the course of their illness (Rutledge et al., 2006). Freedland and colleagues (1991) first reported a trend towards increased length of hospitalization and 1-year mortality in depressed hospitalized HF patients. Although there was no difference in discharge vital sign status or 3-month readmission between patients with and without major depression, 50% of patients who met criteria for major depression at the time of index hospitalization died by 1-year follow-up compared to 29% of non-depressed patients (Freedland et al., 1991). Following Freedland et al.’s (1991) investigation and additional studies, meta-analyses have confirmed the pattern of reduced survival in depressed HF patients (Rutledge et al., 2006; Sokoreli, de Vries, Pauws, & Steyerberg, 2016). Given the increased mortality risk conferred by co-morbid depressive symptoms, greater understanding of factors that contribute to worse prognosis in depressed HF patients is needed to guide intervention efforts.

One potential mechanism of increased risk is poor disease self-management. Depression has long been recognized as a barrier to treatment adherence. As such, depression may lead to increased mortality through its negative effects on self-management behaviors such as medication adherence. In the case of HF, managing medication regimens encompassing varied dosing, timing, and frequency can be difficult. Medication adherence estimates in HF range from approximately 40–60% (Wu, Moser, Chung, & Lennie, 2008), although at least 88% adherence may be necessary to achieve improved outcomes (Wu et al., 2009). Proper medication adherence is critical for managing risk related to cardiac events and worsening cardiovascular disease (CVD), such as reducing high blood glucose levels and improving lipid profiles. Better adherence is associated with lower rates of emergency room visits, fewer cardiac events and readmissions, and shorter hospital stays (Esposito, Bagchi, Verdier, Bencio, & Kim, 2009; Gehi, Ali, Na, & Whooley, 2007). Better medication adherence is also linked to reduced mortality risk in HF (Fitzgerald et al., 2011).

Medication adherence appears to be more difficult for patients with comorbid depressive symptomology. Depression has been linked to non-adherence in CVD (Bane, Hughes, & McElnay, 2006), as well as patient-reported difficulty taking medications in HF (Morgan et al., 2006). A cross-sectional analysis from the Heart and Soul study revealed that depressed patients with CVD were more likely to report skipping and incorrectly taking their medications than non-depressed patients, although no difference emerged in reports of forgetting to take medications (Gehi et al., 2005). Thus, it is reasonable to expect that poor medication adherence would reduce treatment effectiveness and increase mortality risk in patients with comorbid HF and depression. Consistent with this reasoning, Wu, Lennie, Dekker, Biddle, & Moser (2013) reported that the combined depression and medication non-adherence were associated with 5 times greater risk of cardiac events during 3.5-year follow-up in patients with HF (Wu et al., 2013). Given these findings, it is important to determine whether medication non-adherence contributes to the relationship between depression and mortality in HF.

Method

Participants

The present study is a secondary analysis from the Heart Adherence, Behavior, and Cognition (HeartABC) study (Trial Identifier: NCT01461629), which was a longitudinal, observational study of the impact of psychosocial and cognitive factors on self-care behaviors in older adults with HF (Dolansky et al., in press; Dolansky et al., 2016). The HeartABC study enrolled 372 older adults with HF, of which 308 had complete data for the variables of interest in this report. Participants were recruited from inpatient (n = 50) and outpatient (n = 322) cardiology services at two hospital systems in northeast Ohio. Individuals with documented systolic HF (i.e., HF diagnosis was verified within 36 months prior to study enrollment) were eligible for inclusion if they were aged 50–85 years at enrollment and classified as NYHA class II or III by their physician. Individuals were ineligible if they had cardiac surgery within 3 months prior to enrollment, had history of neurological disorder or injury (e.g., Alzheimer’s disease, dementia, stroke, seizures), moderate or severe head injury, past or current significant psychiatric disorders (e.g., psychotic disorders, bipolar disorder, learning disorder, developmental disability), renal failure requiring dialysis, untreated sleep apnea, current substance abuse or within the past 5 years, or were currently using a home telehealth HF monitoring program.

Measures

Depressive Symptoms

The Patient Health Questionnaire—9 (PHQ-9), a 9-item questionnaire, was used to assess depressive symptoms (Kroenke, Spitzer, & Williams, 2001). Scores range from 0 to 27, with higher scores representing greater depressive symptomology. Total scores are categorized as representing none to minimal (0–4), mild (5–9), moderate (10–14), moderately severe (15–19), and severe depression (20–27) (Kroenke, Spitzer, & Williams, 2001). A trained research assistant administered the PHQ-9 at the baseline and third study visits. The two scores were averaged to create a more reliable measure of each participant’s level of depression. A score of ≥ 5 indicated presence of clinically significant depression symptoms.

Medication Non-adherence

Participants completed 28 days of medication monitoring using an electronic Medsignals pillbox (VitalSignals, LLC, Lexington, KY). The Medsignals pillbox transferred data to an electronic server through the participants’ home telephone lines. The pillbox contained four individual bins so that up to four common HF medications (e.g., beta-blockers, angiotensin-converting-enzyme inhibitors, angiotensin II receptor blockers, diuretics, aspirin, aldosterone antagonists, Plavix, and statins, selected in ascending order) were tracked. If necessary, such as in the event of a prescribed change in the medication regimen, adjustments were made in the pills selected for monitoring. Participants were instructed to take their medications as prescribed.

Medication non-adherence was defined as the percent of days a participant did not follow their prescribed regimen. This was calculated as the number of days that the regimen was not followed (i.e., missed doses or overdoses) divided by the number of days monitored, expressed as a percentage). The percent non-adherence for each of the monitored medications was averaged to create a total score with possible scores ranging from 0 – 100%. Days the pillbox was not used for a specific reason, such as travel or hospitalization, were excluded. Medication adherence was treated as a continuous variable in analysis. The first 7 days of monitoring occurred between the second and third study visits and was considered to be a behavioral run-in to allow the participants the opportunity to learn to use the pillbox and incorporate the pillbox into their routine. Only the final 21 days of monitoring were included in analyses.

Covariates

Demographic and medical covariates assessed included age, gender, and disease severity. Age and gender were determined through self-report questionnaires. HF severity was assessed using physician-documented New York Heart Association (NYHA) class. A trained research assistant assessed self-reported HF symptoms and limitations at baseline to confirm NYHA classification at the time of study participation.

Mortality

All-cause and cardiovascular mortality were determined using data obtained from the Centers for Disease Control and Prevention’s National Death Index (NDI; http://www.cdc.gov/nchs/ndi.htm). Deaths recorded through December 2014 were obtained which provided a median 2.9 year follow-up. Cardiovascular deaths were defined by the following ICD-10 codes: acute rheumatic fever/chronic rheumatic heart diseases (I00–I09), hypertensive heart disease (I11), hypertensive heart and renal disease (I13), coronary heart disease (I20–I25), other heart diseases (I26–I51), essential hypertension and hypertensive renal disease (I10, I12, I15), or cerebrovascular diseases (I60–169) (WHO, 2004).

Procedure

The study protocol was approved by the institutional review boards of Kent State University, Summa Health System, and University Hospitals of Cleveland. Participants were recruited from August 2010 through October 2013. Trained research assistants conducted medical chart review of patients admitted under the inpatient cardiology service or scheduled for an outpatient cardiology visit to determine whether individuals potentially met eligibility criteria based on screening of relevant factors (e.g., chart diagnosis of systolic HF with ejection fraction < 40% for at least 3 months and verifiable in the medical record within 3 years of study enrollment). Following written informed consent, participants completed four study visits over the course of 1–2 months, as well as 28 days of medication monitoring. At the baseline visit, demographic, psychosocial, and medical history was collected using self-report questionnaires and interviews. Approximately 2 weeks later, a research assistant conducted the second visit at the participant’s home. This visit included installation of the pillbox. The third visit occurred approximately 1 week later and included a check of the electronic pillbox. The final visit, which included collection of the pillbox, occurred 3 weeks later.

Analytic Plan

Missing Data

Of the enrolled sample, 332 individuals completed the 4 study visits and 309 had complete medication adherence data. Equipment failure and refusal and/or inability to use the pillbox were the most common reasons for missing medication monitoring data. Of those with complete medication adherence data (n = 309), 1 individual was missing PHQ-9 data. We used listwise deletion of individuals missing data on variables included in multivariable analyses (i.e., age, gender, PHQ-9 scores), yielding a final sample of 308.

Preliminary Analyses

Means, standard deviations, frequencies, and percentages were used to summarize sample characteristics. Pearson chi-square tests and t-tests were used to examine differences between depressed and non-depressed participants.

Prior to testing hypotheses, data were examined for violations of the assumptions of regression and Cox proportional hazards modeling. To examine potential violations of univariate normality, skewness and kurtosis statistics were calculated. Bivariate correlations were examined to assess for the presence of multicollinearity. The Schonfield residuals were used to test the proportionality of hazards assumption for continuous data. Examination of plots was used to test the assumption for categorical variables. No violations of assumptions were evident. The criterion for statistical significance was set at p < .05 for correlation and linear regression analyses. For Cox proportional hazards ratios, 95% CI were used. In addition, a reduction of ≥ 3% in the age-adjusted log HR for depression following adjustment for a potential mediator was deemed to indicate significant mediation.

Primary and Secondary Analyses

Hierarchical linear regression was used to test the relationship between PHQ-9 scores and medication non-adherence. In Block 1, covariates (i.e., age, gender, and disease severity) were entered. In Block 2, categorical PHQ-9 scores were added to the model. Medication non-adherence was the dependent variable. Cox proportional hazards regression (Cox, 1972; Cox & Oakes, 1984) was then used to assess the relationship between depressive symptoms and mortality, with and without adjustment confounding variables. For both outcomes (i.e., all-cause and cardiovascular mortality), three Cox proportional hazards regression models were performed. Model 1 included only covariates. Next, Model 2 included covariates included in Model 1 as well as depressive symptoms (PHQ-9 < 5 = 0, PHQ-9 ≥ 5 = 1). Finally, Model 3 included medication non-adherence to assesse whether medication adherence attenuated the hypothesized relationship between depressive symptoms and mortality. All analyses were repeated treating PHQ-9 scores continuously.

Results

Sample Characteristics

The sample included 308 patients with HF with an average age of 68.5 ± 9.64 years. Complete participant characteristics stratified by depressive symptoms are presented in Table 1.

Table 1.

Demographic and Clinical Characteristics of Participants at Baseline (n = 308).

Total Sample Depressive Symptoms
PHQ-9 ≥ 5 PHQ-9 < 5 P value
Sample size (n) 308 107 201
Age 68.48 (9.60) 66.62 (10.12) 69.47 (9.18) .016
Male 186 (60.4) 53 (49.5) 133 (66.2) .003
Caucasian 226 (73.4) 69 (64.5) 157 (78.1) .008
Highest Education Level
 8th grade or less 5 (1.6) 5 (4.7) 0 (0) <.001
 9th – 11th grade 29 (9.4) 14 (13.1) 15 (7.5)
 High School 88 (28.6) 36 (33.6) 52 (25.9)
 Technical/trade school 30 (9.7) 10 (9.3) 20 (10.0)
 Some college 85 (27.6) 30 (28.0) 55 (27.4)
 Bachelor’s degree 42 (13.6) 3 (2.8) 39 (19.4)
 Master’s degree 29 (9.4) 9 (8.4) 20 (10.0)
Married 184 (59.7) 54 (50.5) 130 (64.7) .013
Charlson 3.30 (1.73) 3.31 (1.69) 3.29 (1.75) .924
NYHA
 Class I/II 106 (34.4) 16 (15.0) 90 (44.8) <.001
 Class III/IV 202 (65.6) 91 (85.0) 111 (55.2)
Medication Adherence 73.08 (25.30) 65.32 (27.47) 77. 21 (23.09) <.001
Open in a new tab

Note. Means and standard deviations presented for continuous variables. Frequencies and percentages presented for categorical variables. Pearson χ2 tests (categorical variables) or t-tests (continuous variables) were used to calculate p values.

Abbreviations: Charlson = Charlson Comorbidity Index; NYHA = New York Heart Association; PHQ-9 = Patient Health Questionnaire—9.

Participants averaged 996.08 ± 334.03 days of follow-up (median = 1048). During follow-up, 51 deaths (16.8%) occurred. Forty-three percent (n = 22) of deaths were classified as cardiovascular. Of the sample, 104 patients (34.3%) were classified as at least mildly depressed (PHQ-9 ≥ 5). Of the depressed individuals, 22 (21.2%) experienced the primary outcome (i.e., mortality) by the end of follow-up. Eight participants classified as depressed died from a cardiovascular cause.

Primary Analyses

Depressive symptoms and medication non-adherence

Hierarchical multiple linear regression analysis was performed to examine the relationship between depressive symptoms (PHQ-9 ≥ 5) and medication non-adherence. In the first block, the linear combination of demographic and clinical covariates explained 2.7% of the variability in medication non-adherence, F(3, 304) = 2.81, p < .05. In the second block, the addition of depression significantly improved model fit, ΔR2 = .03, ΔF(1, 303) = 10.56, p < .01. Depressive symptoms were positively related to medication non-adherence, (β = .19, p < .01).

All-cause mortality

A series of Cox proportional hazards regression analyses were conducted to assess the relationships between depression, medication non-adherence, and all-cause mortality risk before and after adjustment for covariates. Univariate analysis indicated no relationship between depression and all-cause mortality (HR = 1.52; 95% CI: .88 – 2.65). Following adjustment for covariates, an association emerged between depression and all-cause mortality risk (HR = 1.87; 95% CI: 1.04 – 3.37). See Figure 1.

Figure 1.

Figure 1

Open in a new tab

Cumulative survival curve for presence or absence of depressive symptoms and all-cause mortality in 303 patients with heart failure.

Next, medication non-adherence was added to the model. In the fully adjusted model, higher non-adherence remained associated with all-cause mortality (HR = 1.01; 95% CI: 1.004 – 1.02). Following adjustment for non-adherence, depression was no longer a significant predictor. Also, adjustment for nonadherence led to a 32.49% decrease in the age-adjusted log HR estimating the effect of depressive symptoms.

Cardiovascular mortality

A second series of Cox proportional hazards regression analyses were performed to examine the relationships between depressive symptoms, medication adherence, and cardiovascular mortality risk. Depression was not associated with cardiovascular mortality risk in univariate nor multivariate analysis. In the final model, medication non-adherence was added to assess the relationship between non-adherence and cardiovascular mortality. Medication non-adherence was not associated with cardiovascular mortality. See Table 2 for a summary of univariate Cox proportional hazards models. Given the low number of cardiac deaths, this series of analyses were likely underpowered to detect an effect after adjusting for covariates. Multivariate analyses are presented in Tables 3 and 4.

Table 2.

Univariate Cox proportional hazard ratios for mortality.

All-Cause Mortality Cardiovascular Mortality
HR (95% CI) P value HR (95% CI) P value
Age 1.04 (1.01 – 1.07) .008 1.08 (1.02 – 1.13) .004
Sex .46 (.24 – .88) .019 .33 (.11 – .98) .046
NYHA 1.29 (.71 – 2.35) .399 .93 (.39 – 2.19) .868
PHQ-9 (categorical) ≥ 5 1.56 (.90 – 2.70) .111 1.62 (.70 – 3.76) .258
PHQ-9 (continuous) 1.04 (.99 – 1.09) .086 1.02 (.94 – 1.11) .664
Medication Non-adherence 1.01 (1.004 – 1.02) .004 1.02 (1.00 – 1.03) .044
Open in a new tab

Note. PHQ-9 (categorical): 0 = not depressed; 1 = depressed.

Abbreviations: NYHA = New York Heart Association; PHQ-9 = Patient Health Questionnaire—9.

Table 3.

Multivariate Cox Proportional Hazard Ratios of relationship among depression and mortality.

All-Cause Mortality Cardiovascular Mortalityǂ
HR (95% CI) P value HR (95% CI) P value
Block 1
 Age 1.04 (1.01 – 1.07) .012 1.07 (1.02 – 1.13) .006
 Sex .47 (.25 – .91) .024 .37 (.12 – 1.08) .068
 NYHA 1.36(.90 – 2.04) .142 1.16 (.63 – 2.13) .626
Block 2
 Age 1.05 (1.01 – 1.08) .004 1.09 (1.03 – 1.15) .002
 Sex .43 (.24 – .87) .018 .35 (.12 – 1.05) .061
 NYHA 1.18 (.77 – 1.79) .453 .93 (.50 – 1.76) .830
 PHQ-9 ≥ 5 1.87 (1.04 – 3.37) .036 2.36 (.96 – 5.80) .060
Block 3
 Age 1.04 (1.02 – 1.08) .003 1.09 (1.03 – 1.15) .002
 Sex .44 (.23 – .86) .015 .35 (.12 – 1.05) .061
 NYHA 1.10 (.73 – 1.68) .642 .87 (.46 – 1.64) . 674
 PHQ-9 ≥ 5 1.59 (.87 – 2.91) .135 2.07 (.82 – 5.18) .122
 Medication Non-adherence 1.01 (1.004 – 1.02) .010 1.01 (.99 – 1.03) .085
Open in a new tab

Abbreviations: NYHA = New York Heart Association; PHQ-9 = Patient Health Questionnaire—9.

ǂ

Given the low number of cardiac deaths, this series of analyses were likely underpowered to detect an effect after adjusting for covariates.

Table 4.

Multivariate Cox Proportional Hazard Ratios of depression severity and mortality.

All-Cause Mortality Cardiovascular Mortalityǂ
HR (95% CI) P value HR (95% CI) P value
Block 1
 Age 1.04 (1.01 – 1.07) .012 1.07 (1.02 – 1.13) .006
 Sex .47 (.25 – .91) .024 .37 (.12 – 1.08) .068
 NYHA 1.36(.90 – 2.04) .142 1.16 (.63 – 2.13) .626
Block 2
 Age 1.04 (1.01 – 1.08) .004 1.09 (1.03 – 1.15) .002
 Sex .45 (.24 – .87) .017 .37 (.12 – 1.09) .071
 NYHA 1.17 (.76 – 1.78) .480 .99 (.52 – 1.89) .985
 PHQ-9 (continuous) 1.06 (1.01 – 1.12) .025 1.06 (.96 – 1.15) .249
Block 3
 Age 1.05 (1.02 – 1.08) .003 1.08 (1.03 – 1.15) .002
 Sex .44 (.23 – .86) .015 .37 (.12 – 1.09) .072
 NYHA 1.08 (.72 – 1.65) .668 .92 (.49 – 1.75) .807
 PHQ-9 (continuous) 1.05 (.99 – 1.11) .081 1.05 (.95 – 1.15) .372
 Medication Adherence 1.01 (1.003 – 1.02) .013 1.02 (.99 – 1.03) .055
Open in a new tab

Abbreviations: NYHA = New York Heart Association; PHQ-9 = Patient Health Questionnaire—9.

ǂ

Given the low number of cardiac deaths, this series of analyses were likely underpowered to detect an effect after adjusting for covariates.

Secondary Analyses

A series of analyses were also conducted treating PHQ-9 scores as a continuous variable to assess the dose-response relationship between depressive symptoms and mortality.

Depressive symptoms and medication non-adherence

Hierarchical multiple linear regression tested the relationship between depressive symptoms and medication non-adherence. In the first block, the linear combination of covariates explained 2.7% of the variability in medication non-adherence, F(3, 304) = 2.81, p < .05. In the second block, adding PHQ-9 scores accounted for an additional 3.0% of the variance in medication non-adherence and improved model fit, ΔF(1, 303) = 9.63, p < .01. Higher levels of depressive symptoms were associated with higher non-adherence, (β = .19, p < .01).

All-cause mortality

Univariate analysis indicated no relationship between depression and all-cause mortality. Following covariate adjustment, higher depressive symptoms were related to increased all-cause mortality (HR = 1.06; 95% CI: 1.01 – 1.12).

Multivariate analyses examining relationships among medication non-adherence and mortality were also performed. Medication non-adherence was associated with increased mortality risk after adjustment for covariates and depressive symptoms (HR = 1.01; 95% CI: 1.003 – 1.02). When non-adherence was added to the model, depressive symptoms were no longer significantly related to mortality risk. This represents a 22.61% in the age-adjusted log HR estimating the effect of depressive symptoms when non-adherence was added to the model.

Depressive symptoms, medication non-adherence, and cardiovascular mortality

Finally, Cox proportional hazards models were conducted to examine relationships among depressive symptoms, medication adherence, and cardiovascular mortality. No univariate nor multivariate relationship emerged between depressive symptoms and cardiovascular mortality. A fully adjusted model including medication non-adherence showed no relationship between medication non-adherence and mortality risk. However, given the low number of cardiac deaths, this series of analyses were likely underpowered to detect an effect.

Discussion

Here we report that the presence of at least mild depressive symptoms was associated with lower adherence and increased all-cause and cardiovascular mortality risk in HF. Medication non-adherence was also related to mortality risk; as medication nonadherence increased 1%, mortality risk increased by 1%, indicating that a 10% increase in medication nonadherence reflected a 12% increase in 2.7 year mortality risk. Medication non-adherence attenuated the relationship between depressive symptoms and all-cause mortality. When treated as a continuous variable, depressive symptoms were again positively associated with all-cause, but not cardiovascular, mortality. A graded association between depression severity and all-cause mortality risk emerged. Each 1-point increase in depressive symptoms corresponded to nearly a 10% increase in all-cause mortality risk, which appeared to be partially accounted for by medication non-adherence. We believe that this is the first report on the contribution of objectively-measured medication adherence to the relationship between depression and mortality in patients with HF. This finding adds to a large body of research demonstrating the association between increasing depression severity and poor prognosis in HF.

In line with hypotheses, depressive symptoms were positively associated with medication non-adherence. There are several potential explanations for this relationship. Motivation and fatigue associated with depression may impede adherence to treatment recommendations. An exploration of relationships among depression, anxiety, and fatigue in HF revealed that depression and anxiety were associated with different aspects of fatigue (Falk, Patel, Swedburg, & Ekman, 2009). Individuals with anxiety displayed more mental fatigue (Falk et al., 2008). Depression was linked to fatigue-related reductions in activity, motivation, and functioning (Falk et al., 2008). Falk et al. (2008) suggested that depression may interfere with activity because of decreased energy necessary to follow through with self-care activities. The directionality of these co-morbid factors needs clarification.

Furthermore, negative thinking and poor coping prevalent in depression may contribute to non-adherence. HF patients with depression appear more likely to employ coping methods such as denial and disengagement as opposed to active coping, acceptance, and planning-related coping strategies (Allman, Berry, & Nasir, 2009; Klein, Turvey, & Pies, 2007). This is unfortunate, as problem-focused and emotion-focused (such as acceptance) coping strategies appear to be related to better self-care in HF (Li & Shun, 2016). It is likely that individuals able to demonstrate flexibility when employing emotion- or problem-focused strategies experience advantages regarding managing a chronic illness; for example, individuals may use problem-focused strategies to manage concrete barriers to disease management or emotion-focused strategies, such as acceptance or emotional-support seeking, when negative mood may otherwise impact motivation for self-care. Depression may dampen such flexibility.

In addition to regimen-related factors, patient beliefs about their disease and the efficacy of treatment may also impact medication-taking behavior. For example, patients who view their disease as fatal with little hope for longevity or improvement may be less prone to adhere to medication protocols (Rashid, Edwards, Walter, & Mant, 2014). In addition, beliefs more specifically related to the necessity or concerns related to side effects of medications likely impact adherence (Horne et al., 2013), and depressive symptoms may also impact the balance of weighing potential necessity and advantages versus side effects and costs. As depression is associated with negative thinking styles, including hopelessness and negative views of the future, depressed patients may also hold negative beliefs about the efficacy of their treatment regimens.

Both depression and poor medication adherence represent modifiable, prognostic risk factors. The relationship between depression and risk for poor prognosis in patients with HF is likely impacted by a number of physiological changes that occur in both CVD and depression, as well as behavioral factors (Joynt, Whellan, O’Connor, 2004; Kop, Synowski, & Gottlieb, 2011; Whooley et al., 2008; Zuluaga et al., 2010). A number of behavioral factors, in addition to medication adherence, may also play a role. Although behaviors such as physical activity, diet, smoking, and medication taking represent different aspects of health and disease management, many of these factors covary. For example, medication non-adherence is often considered a marker of several unhealthy lifestyle behaviors associated with increased risk of cardiac events and poor outcomes (Bonnet et al., 2005; Whooley, 2006; Simpson et al., 2006). As such, it remains possible that medication non-adherence in the current study may serve as a proxy of poor engagement in other heath behaviors. Prior research exploring these mechanisms of action has suggested physical inactivity to be an important behavioral factor (Whooley et al., 2008; Zuluaga et al., 2010). Whooley et al. (2008) demonstrated that physical inactivity was the largest behavioral factor impacting the relationship between depressive symptoms and risk of experiencing cardiac events, reducing the age-adjusted log HR by 31%. Self-reported medication non-adherence reduced the age-adjusted log HR by 5.3%. The current investigation, which incorporated objectively-measured medication adherence, found a significant impact on the association between depressive symptoms and all-cause mortality.

Limitations

Limitations of the current study warrant mention. The mortality rate in the current study was lower than expected. Similar studies of HF with follow-up periods of 18 months (Coyne et al., 2011) and 32 months (Faller et al., 2007) reported mortality rates of 27.4%, and 26% respectively. The low mortality rate may be attributed to recruitment of motivated, well-managed patients who are closely followed by their cardiologists. Nonetheless, the presence of at least mild symptoms of depression conferred a significant increase in mortality risk, with individuals reporting elevated depressive symptoms experiencing mortality at over twice the rate of individuals who did not report elevated symptoms. Similarly, Rutledge and colleagues (2006) reported approximately a 2-fold all-cause mortality risk for depressed individuals with HF.

Furthermore, many of the relationships that emerged when examining all-cause mortality were not evident for cardiovascular mortality. This may be a result of the low rate of cardiovascular deaths, leading to limited power to detect an effect. A prior meta-analysis suggested that increased cardiovascular mortality was associated with depressive symptomology in HF (Fan et al., 2014). In an analysis of 48,117 HF patients identified through hospital records as receiving psychotropic treatment for depression prior to HF diagnosis, Macchia and colleages (2008) suggested that increased rates of 1-year all-cause mortality among patients previously treated for depression were largely due to increased rates of vascular events (i.e., MI, stroke, transischemic attack). Future studies may benefit from larger sample sizes and/or longer follow-up periods to detect relationships with cardiovascular mortality. The possibility remains that CVD contributed to some deaths which were recorded as non-cardiac (e.g., falls) given that NDI recorded causes of death were not adjudicated by obtaining individual death certificates or examining hospital records.

In addition, the direction of the relationship between depressive symptoms and medication adherence could not be established given the current study design. Finally, history of depression was not assessed. It will be important for future research to explore whether recurrent depressive episodes or newly onset depression following development of CVD are differentially related to prognosis. Changes in depressive symptoms over time could not be investigated in the current study, but the persistence of depression may be differentially associated with mortality risk relative to single-episode or remitting symptoms of depression. Similarly, changes in medication adherence over time may contribute to changes in mortality risk.

Conclusions and Future Directions

Medication non-adherence contributes to the relationship between depression and all-cause mortality risk in individuals with HF. Future research is needed to explore how depression and medication adherence relate to other outcomes (e.g., cardiovascular mortality, cardiac events, hospitalization) in larger samples or in samples with longer follow-up. Healthcare providers should continue to screen patients for depressive symptoms and monitor changes in mood, as even mild symptoms of depression can indicate risk for poor medication adherence and prognosis. However, the evidence-base for treatment recommendations is sparse at this time. Cognitive behavioral therapy reduces depression in HF and appeared to reduce hospitalizations, but it did not improve self-care assessed via self-report (Freedland, Carney, Rich, Steinmeyer, & Rubin, 2015). With respect to antidepressants, neither sertraline nor escitalopram were effective for depression in HF (Angermann et al., 2016; O’Connor et al., 2010). Perhaps interventions should address medication adherence directly, although a pilot clinical trial of medication reminding found no improvement in adherence from pillbox alarms or smartphone reminders (Goldstein et al., 2014). Medication adherence interventions for patients with HF need further refinement and evidence of effectiveness. Of course, whether improvements in medication adherence would reduce the risks associated with depression symptoms is not known. Cardiac rehabilitation (CR) programs, which have become more available to HF patients in recent years, have been offer a promising setting to target mood, adherence, and other health behaviors such as physical activity. Mood improvement is commonly observed following CR, and many comprehensive CR programs offer opportunities for engagement with a psychologist and/or pharmacist, in addition to monitored exercise. Nonetheless, increased awareness of the impact of depression on the quality and quantity of life in patients with HF will likely benefit patients, families, and providers by encouraging the identification of targets for intervention and need for support.

Acknowledgments

Sources of Funding: This research was supported by the National Heart, Lung, and Blood Institute R01 HL096710-01A1 awarded to Drs. Dolansky and Hughes.

Footnotes

Conflict of Interest: The authors declare no conflict of interest.

References

  1. Allman E, Berry D, Nasir L. Depression and coping in heart failure patients: a review of the literature. Journal of Cardiovascular Nursing. 2009;24(2):106–117. doi: 10.1097/JCN.0b013e318197a985. [DOI] [PubMed] [Google Scholar]
  2. Albert NM, Fonarow GC, Abraham WT, Gheorghiade M, Greenberg BH, Nunez E, … Young JB. Depression and clinical outcomes in heart failure: an OPTIMIZE-HF analysis. The American Journal of Medicine. 2009;122(4):366–373. doi: 10.1016/j.amjmed.2008.09.046. [DOI] [PubMed] [Google Scholar]
  3. Angermann CE, Gelbrich G, Störk S, Gunold H, Edelmann F, Wachter R, … Blankenberg S. Effect of escitalopram on all-cause mortality and hospitalization in patients with heart failure and depression: the MOOD-HF randomized clinical trial. JAMA. 2016;315(24):2683–2693. doi: 10.1001/jama.2016.7635. [DOI] [PubMed] [Google Scholar]
  4. Bane C, Hughes CM, McElnay JC. The impact of depressive symptoms and psychosocial factors on medication adherence in cardiovascular disease. Patient Education and Counseling. 2006;60(2):187–193. doi: 10.1016/j.pec.2005.01.003. [DOI] [PubMed] [Google Scholar]
  5. Bonnet F, Irving K, Terra JL, Nony P, Berthezène F, Moulin P. Anxiety and depression are associated with unhealthy lifestyle in patients at risk of cardiovascular disease. Atherosclerosis. 2005;178(2):339–344. doi: 10.1016/j.atherosclerosis.2004.08.035. [DOI] [PubMed] [Google Scholar]
  6. Caldeira D, Vaz-Carneiro A, Costa J. The impact of dosing frequency on medication adherence in chronic cardiovascular disease: Systematic review and meta-analysis. Revista Portuguesa de Cardiologia (English Edition) 2014;33(7):431–437. doi: 10.1016/j.repc.2014.01.013. [DOI] [PubMed] [Google Scholar]
  7. Chen J, Normand SLT, Wang Y, Krumholz HM. National and regional trends in heart failure hospitalization and mortality rates for Medicare beneficiaries, 1998–2008. The Journal of the American Medical Association. 2011;306(15):1669–1678. doi: 10.1001/jama.2011.1474. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cox D. Breakthroughs in Statistics. Springer; New York: 1972. Regression models and life-tables (with Discussion) pp. 527–541. [Google Scholar]
  9. Cox DR, Oakes D. Analysis of survival data. Vol. 21. CRC Press; 1984. [Google Scholar]
  10. Dolansky MA, Hawkins MAW, Schaefer JT, Sattar A, Gunstad J, Redle J, Josephson RA, Moore SM, Hughes JW. The Association between Poorer Cognitive Function and Reduced Objectively-Monitored Medication Adherence in Patients with Heart Failure. Circulation: Heart Failure. doi: 10.1161/CIRCHEARTFAILURE.116.002475. in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Dolansky MA, Hawkins MA, Schaefer JT, Gunstad J, Sattar A, Redle JD, … Hughes JW. Cognitive Function Predicts Risk for Clinically Significant Weight Gain in Adults With Heart Failure. Journal of Cardiovascular Nursing. doi: 10.1097/JCN.0000000000000376. in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dolansky MA, Schaefer JT, Hawkins MA, Gunstad J, Basuray A, Redle JD, … Hughes JW. The association between cognitive function and objective adherence to dietary sodium guidelines in patients with heart failure. Patient Preference and Adherence. 2016;10:233–241. doi: 10.2147/PPA.S95528. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Esposito D, Bagchi AD, Verdier JM, Bencio DS, Kim MS. Medicaid beneficiaries with congestive heart failure: association of medication adherence with healthcare use and costs. The American Journal of Managed Care. 2009;15(7):437–445. [PubMed] [Google Scholar]
  14. Falk K, Patel H, Swedberg K, Ekman I. Fatigue in patients with chronic heart failure—a burden associated with emotional and symptom distress. European Journal of Cardiovascular Nursing. 2009;8(2):91–96. doi: 10.1016/j.ejcnurse.2008.07.002. [DOI] [PubMed] [Google Scholar]
  15. Faller H, Stork S, Schowalter M, Steinbuchel T, Wollner V, Ertl G, Angermann CE. Depression and survival in chronic heart failure: Does gender play a role? The European Journal of Heart Failure. 2007;9:1018–1023. doi: 10.1016/j.ejheart.2007.06.011. [DOI] [PubMed] [Google Scholar]
  16. Fan H, Yu W, Zhang Q, Cao H, Li J, Wang J, … Hu X. Depression after heart failure and risk of cardiovascular and all-cause mortality: A meta-analysis. Preventive Medicine. 2014;63:36–42. doi: 10.1016/j.ypmed.2014.03.007. [DOI] [PubMed] [Google Scholar]
  17. Fitzgerald AA, Powers JD, Ho PM, Maddox TM, Peterson PN, Allen LA, … Havranek EP. Impact of medication nonadherence on hospitalizations and mortality in heart failure. Journal of Cardiac Failure. 2011;17(8):664–669. doi: 10.1016/j.cardfail.2011.04.011. [DOI] [PubMed] [Google Scholar]
  18. Freedland KE, Carney RM, Rich MW, Caracciolo A. Depression in elderly patients with congestive heart failure. Journal of Geriatric Psychiatry. 1991;24(1):59–71. [Google Scholar]
  19. Freedland KE, Carney RM, Rich MW, Steinmeyer BC, Rubin EH. Cognitive behavior therapy for depression and self-care in heart failure patients: a randomized clinical trial. JAMA internal medicine. 2015;175(11):1773–1782. doi: 10.1001/jamainternmed.2015.5220. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Garcia S, Spitznagel MB, Cohen R, Raz N, Sweet L, Colbert L, … Gunstad J. Depression is associated with cognitive dysfunction in older adults with heart failure. Cardiovascular Psychiatry and Neurology. 2011;2011 doi: 10.1155/2011/368324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gehi AK, Ali S, Na B, Whooley MA. Self-reported medication adherence and cardiovascular events in patients with stable coronary heart disease: the heart and soul study. Archives of Internal Medicine. 2007;167(16):1798–1803. doi: 10.1001/archinte.167.16.1798. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gehi A, Haas D, Pipkin S, Whooley MA. Depression and medication adherence in outpatients with coronary heart disease: findings from the Heart and Soul Study. Archives of Internal Medicine. 2005;165(21):2508–2513. doi: 10.1001/archinte.165.21.2508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Gilberg K, Laouri M, Wade S, Isonaka S. Analysis of medication use patterns: Apparent overuse of antibiotics and underuse of prescription drugs for asthma, depression, and CHF. Journal of Managed Care Pharmacy. 2003;9(3):232–237. doi: 10.18553/jmcp.2003.9.3.232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Goldstein CM, Gathright EC, Dolansky MA, Gunstad J, Sterns A, Redle JD, … Hughes JW. Randomized controlled feasibility trial of two telemedicine medication reminder systems for older adults with heart failure. Journal of telemedicine and telecare. 2014;20(6):293–299. doi: 10.1177/1357633X14541039. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Heidenreich PA, Trogdon JG, Khavjou OA, Butler J, Dracup K, Ezekowitza MD, … Woo YJ. Forecasting the future of cardiovascular disease in the United States: a policy statement from the American Heart Association. Circulation. 2011;123:933–944. doi: 10.1161/CIR.0b013e31820a55f5. [DOI] [PubMed] [Google Scholar]
  26. Heidenreich PA, Albert NM, Allen LA, Bluemke DA, Butler J, Fonarow GC, Ikonomidis JS, Khavjou O, Konstam MA, Maddox TM, Nichol G, Pham M, Piña IL, Trogdon JG on behalf of the American Heart Association Advocacy Coordinating Committee; Council on Arteriosclerosis, Thrombosis and Vascular Biology; Council on Cardiovascular Radiology and Intervention; Council on Clinical Cardiology; Council on Epidemiology and Prevention; Stroke Council. Forecasting the impact of heart failure in the United States: a policy statement from the American Heart Association. Circulation: Heart Failure. 2013;6:606–619. doi: 10.1161/HHF.0b013e318291329a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Horne R, Chapman SC, Parham R, Freemantle N, Forbes A, Cooper V. Understanding patients’ adherence-related beliefs about medicines prescribed for long-term conditions: a meta-analytic review of the Necessity-Concerns Framework. PloS one. 2013;8(12):e80633. doi: 10.1371/journal.pone.0080633. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hunt SA, Baker DW, Chin MH, Cinquegrani MP, Feldman AM, Francis GS, … Smith SC. ACC/AHA guidelines for the evaluation and management of chronic heart failure in the adult: executive summary: a report of the American College of Cardiology/American Heart Association task force on practice guidelines (committee to revise the 1995 guidelines for the evaluation and management of heart failure) developed in collaboration with the International Society for Heart and Lung Transplantation endorsed by the Heart Failure Society of America. Journal of the American College of Cardiology. 2001;38(7):2101–2113. doi: 10.1016/s0735-1097(01)01683-7. [DOI] [PubMed] [Google Scholar]
  29. Insel K, Morrow D, Brewer B, Figueredo A. Executive function, working memory, and medication adherence among older adults. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences. 2006;61(2):P102–P107. doi: 10.1093/geronb/61.2.p102. [DOI] [PubMed] [Google Scholar]
  30. Joynt KE, Whellan DJ, O’Connor CM. Why is depression bad for the failing heart? A review of the mechanistic relationship between depression and heart failure. Journal of Cardiac Failure. 2004;10(3):258–271. doi: 10.1016/j.cardfail.2003.09.008. [DOI] [PubMed] [Google Scholar]
  31. Klein DM, Turvey CL, Pies CJ. Relationship of coping styles with quality of life and depressive symptoms in older heart failure patients. Journal of Aging and Health. 2007;19(1):22–38. doi: 10.1177/0898264306296398. [DOI] [PubMed] [Google Scholar]
  32. Kop WJ, Synowski SJ, Gottlieb SS. Depression in heart failure: Biobehavioral mechanisms. Heart Failure Clinics. 2011;7(1):23–38. doi: 10.1016/j.hfc.2010.08.011. [DOI] [PubMed] [Google Scholar]
  33. Kronish IM, Rieckmann N, Halm EA, Shimbo D, Vorchheimer D, Haas DC, Davidson KW. Persistent depression affects adherence to secondary prevention behaviors after acute coronary syndromes. Journal of General Internal Medicine. 2006;21(11):1178–1183. doi: 10.1111/j.1525-1497.2006.00586.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Li CC, Shun SC. Understanding self-care coping styles in patients with chronic heart failure: A systematic review. European Journal of Cardiovascular Nursing. 2016;15(1):12–19. doi: 10.1177/1474515115572046. [DOI] [PubMed] [Google Scholar]
  35. Macchia A, Monte S, Pellegrini F, Romero M, D’Ettorre A, Tavazzi L, … Maggioni AP. Depression worsens outcomes in elderly patients with heart failure: an analysis of 48,117 patients in a community setting. European Journal of Heart Failure. 2008;10(7):714–721. doi: 10.1016/j.ejheart.2008.05.011. [DOI] [PubMed] [Google Scholar]
  36. McDermott LM, Ebmeier KP. A meta-analysis of depression severity and cognitive function. Journal of Affective Disorders. 2009;119(1):1–8. doi: 10.1016/j.jad.2009.04.022. [DOI] [PubMed] [Google Scholar]
  37. Morgan AL, Masoudi FA, Havranek EP, Jones PG, Peterson PN, Krumholz HM, … Rumsfeld JS. Difficulty taking medications, depression, and health status in heart failure patients. Journal of Cardiac Failure. 2006;12(1):54–60. doi: 10.1016/j.cardfail.2005.08.004. [DOI] [PubMed] [Google Scholar]
  38. Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, … Stroke SS. Heart disease and stroke statistics-2015 update: a report from the American Heart Association. Circulation. 2015;131(4):e29. doi: 10.1161/CIR.0000000000000152. [DOI] [PubMed] [Google Scholar]
  39. Murphy SL, Xu J, Kochanek KD. Deaths: final data for 2010. National Vital statistics reports. 2013;61(4):1–118. [PubMed] [Google Scholar]
  40. O’Connor CM, Jiang W, Kuchibhatla M, Silva SG, Cuffe MS, Callwood DD, … Krishnan R. Safety and efficacy of sertraline for depression in patients with heart failure: results of the SADHART-CHF (Sertraline Against Depression and Heart Disease in Chronic Heart Failure) trial. Journal of the American College of Cardiology. 2010;56(9):692–699. doi: 10.1016/j.jacc.2010.03.068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Pilkonis PA, Choi SW, Reise SP, Stover AM, Riley WT, Cella D PROMIS Cooperative Group. Item banks for measuring emotional distress from the Patient-Reported Outcomes Measurement Information System (PROMIS®): depression, anxiety, and anger. Assessment. 2011 Sep;18(3):263–83. doi: 10.1177/1073191111411667. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Pocock SJ, Wang D, Pfeffer MA, Yusuf S, McMurray JJ, Swedberg KB, … Granger CB. Predictors of mortality and morbidity in patients with chronic heart failure. European Heart Journal. 2006;27(1):65–75. doi: 10.1093/eurheartj/ehi555. [DOI] [PubMed] [Google Scholar]
  43. Rashid MA, Edwards D, Walter FM, Mant J. Medication taking in coronary artery disease: a systematic review and qualitative synthesis. The Annals of Family Medicine. 2014;12(3):224–232. doi: 10.1370/afm.1620. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Rutledge T, Reis VA, Linke SE, Greenberg BH, Mills P. Depression in Heart Failure: A Meta-Analytic Review of Prevalence, Intervention Effects, and Associations With Clinical Outcomes. Journal of the American College of Cardiology. 2006;48:1527–1537. doi: 10.1016/j.jacc.2006.06.055. [DOI] [PubMed] [Google Scholar]
  45. Simpson SH, Eurich DT, Majumdar SR, Padwal RS, Tsuyuki RT, Varney J, Johnson JA. A meta-analysis of the association between adherence to drug therapy and mortality. BMJ. 2006;333:15. doi: 10.1136/bmj.38875.675486.55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Sokoreli I, de Vries JJG, Pauws SC, Steyerberg EW. Depression and anxiety as predictors of mortality among heart failure patients: systematic review and meta- analysis. Heart failure reviews. 2016;21(1):49–63. doi: 10.1007/s10741-015-9517-4. [DOI] [PubMed] [Google Scholar]
  47. Snyder HR. Major depressive disorder is associated with broad impairments on neuropsychological measures of executive function: a meta-analysis and review. Psychological Bulletin. 2013;139(1):81. doi: 10.1037/a0028727. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Whooley MA. Depression and cardiovascular disease: healing the broken-hearted. Journal of the American Medical Association. 2006;295(24):2874–2881. doi: 10.1001/jama.295.24.2874. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Whooley MA, de Jonge P, Vittinghoff E, Otte C, Moos R, Carney RM, … Browner WS. Depressive symptoms, health behaviors, and risk of cardiovascular events in patients with coronary heart disease. Journal of the American Medical Association. 2008;300(20):2379–2388. doi: 10.1001/jama.2008.711. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. World Health Organization. Global status report on noncommunicable diseases 2010. World Health Organization; 2011. [Google Scholar]
  51. World Health Organization. International statistical classification of diseases and related health problems. Vol. 1. World Health Organization; 2004. [Google Scholar]
  52. Wu JR, Lennie TA, Dekker RL, Biddle MJ, Moser DK. Medication Adherence, Depressive Symptoms, and Cardiac Event–Free Survival in Patients With Heart Failure. Journal of Cardiac Failure. 2013;19(5):317–324. doi: 10.1016/j.cardfail.2013.03.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Wu JR, Moser DK, De Jong MJ, Rayens MK, Chung ML, Riegel B, Lennie TA. Defining an evidence-based cutpoint for medication adherence in heart failure. American Heart Journal. 2009;157(2):285–291. doi: 10.1016/j.ahj.2008.10.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Wu JR, Moser DK, Chung ML, Lennie TA. Objectively measured, but not self-reported, medication adherence independently predicts event-free survival in patients with heart failure. Journal of Cardiac Failure. 2008a;14(3):203–210. doi: 10.1016/j.cardfail.2007.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Wu JR, Moser DK, Chung ML, Lennie TA. Predictors of medication adherence using a multidimensional adherence model in patients with heart failure. Journal of cardiac failure. 2008b;14(7):603–614. doi: 10.1016/j.cardfail.2008.02.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zuluaga MC, Guallar-Castillón P, Rodríguez-Pascual C, Conde-Herrera M, Conthe P, Rodríguez-Artalejo F. Mechanisms of the association between depressive symptoms and long-term mortality in heart failure. American Heart Journal. 2010;159(2):231–237. doi: 10.1016/j.ahj.2009.11.011. [DOI] [PubMed] [Google Scholar]

RESOURCES