Association of physical versus affective depressive symptoms with cardiac event-free survival in patients with heart failure.

OBJECTIVE

To determine whether physical depressive symptoms inflate the association between depressive symptoms as measured with the nine-item Patient Health Questionnaire (PHQ-9) and cardiac event-free survival in patients with heart failure (HF).

METHODS

A total of 210 patients with HF were recruited from HF clinics affiliated with two academic medical centers. The PHQ-9 was used to assess levels of depressive symptoms. Cardiac event-free survival data (cardiac death, cardiac hospitalization, or cardiac emergency department visit) were collected for a median follow-up of 360 days. Cox proportional hazards regression analyses were performed separately for physical and affective depressive symptom dimensions of the PHQ-9 to examine predictive ability for time to the first cardiac event.

RESULTS

Scores of both physical and affective depressive symptom dimensions of the PHQ-9 predicted time to the first cardiac event in separate unadjusted models. However, scores of the physical depressive symptom dimension did not predict time to the first cardiac events, whereas scores of the affective depressive symptom dimension remained as an independent predictor (hazard ratio = 1.12, 95% confidence interval = 1.03-1.22) after controlling for health status (comorbidities and the New York Heart Association functional class) and clinical and sociodemographic factors.

CONCLUSIONS

Affective depressive symptoms, not physical depressive symptoms, predicted time to the first cardiac event independent of health status and clinical and sociodemographic characteristics. The use of the full PHQ-9 does not inflate the relationship of depressive symptoms to cardiac event-free survival. Thus, clinicians can use the PHQ-9 to assess depressive symptoms in their patients with HF without concern that the instrument overestimates the relationship between depressive symptoms and outcomes.