Objective: Depression is common among heart failure (HF) patients and can influence cardiovascular outcomes, including both increased symptom reporting and objective health endpoints. We investigated the extent to which increased reporting of cardiovascular symptoms associated with depression in heart failure occurs in relation to functional status and hospitalizations.

Methods: Sixty-three patients (mean = 54.4 ± 11.5 years) with HF (ejection fraction ≤ 40) were examined at baseline and again 3 months later. The Beck Depression Inventory (BDI), the Kansas City Cardiomyopathy Questionnaire (KCCQ) as a measure of subjective symptom reporting, and a standardized Six Minute Walk Test (6MWT) measure of functional status, were administered at baseline and 3 months. Hospitalizations were noted for up to 12 months after initial assessment and plasma levels of b-type natriuretic peptide (BNP), a HF biomarker, were determined at both time points.

Results: Baseline BDI scores correlated negatively with 8 of the 9 subscales of the KCCQ (Physical Limitation: r = -0.40, p =.002; Symptom Frequency: r = -.54, p < .001; Symptom Burden: r = -0.56, p <.001; Total Symptoms: r = -0.57, p < .001; Self-Efficacy: r = -0.27, p =.047; Quality of Life: r = -0.67, p <.001; Clinical Summary: r = -0.55, p < .001; Overall Summary: r = -0.63, p <.001); BDI was not related to the Symptom Stability subscale. Furthermore, changes in BDI scores and the KCCQ from baseline to 3 months were also negatively related to symptom status for all KCCQ subscales (r range: -0.50 to -0.34, p < .05) except for Physical Limitation (r = -0.29, p = .067). BDI did not relate to 6MWT performance nor correlate with BNP at either time point. With regard to objective outcomes, hospitalizations in the year following baseline occurred in 19 of 63 patients (28.8%). Baseline KCCQ scores did not predict subsequent hospitalizations, but BNP was a significant predictor of hospitalizations (OR = 12.85, 95% CI: 1.63 to 101.56, p = .015) after controlling for age, gender, BMI, smoking status, atrial fibrillation, diabetic status, serum creatinine, and prior myocardial infarction (MI). After accounting for the contribution of effect of BNP, BDI significantly predicted hospitalizations (OR = 1.12, 95% CI: 1.01 to 1.23, p = .032) and BDI scores improved the reclassification index in correctly identifying hospitalized patients by 11.9% over a model using baseline BNP.

Conclusions: In HF patients studied prospectively, subjective depressive symptoms related well to symptom reporting, but were unrelated to objective functional status as measured by the 6MWT and to physiological outcomes from plasma BNP. Depression somewhat improved identification of future hospitalizations along with baseline BNP.