Abstract
BACKGROUND: Diabetes mellitus (DM) and depression are bidirectionally interrelated. We recently identified long-term trajectories of depression symptom severity in individuals with coronary heart disease (CHD), which were associated with the risk for subsequent cardiovascular events (CVE). We now investigated the prognostic value of these trajectories of symptoms of depression with the risk of incident DM in patients with stable coronary heart disease. METHODS: The KAROLA cohort included CHD patients participating in an in-patient rehabilitation program (years 1999/2000) and followed for up to 15 years. We included 1048 patients (mean age 59.4 years, 15% female) with information on prevalent DM at baseline and follow-up data. Cox proportional hazards models were used to model the risk for incident DM during follow-up by depression trajectory class adjusted for age, sex, education, smoking status, body mass index, and physical activity. In addition, we modeled the excess risk for subsequent CVE due to incident DM during follow-up for each of the depression trajectories. RESULTS: DM was prevalent in 20.7% of patients at baseline. Over follow-up, 296 (28.2%) of patients had a subsequent CVE. During follow-up, 157 (15.0%) patients developed incident DM before experiencing a subsequent CVE. Patients following a high-stable depression symptom trajectory were at substantially higher risk of developing incident DM than patients following a low-stable depression symptom trajectory (hazard ratio (HR) = 2.50; 95% confidence interval (CI) (1.35, 4.65)). A moderate-stable and an increasing depression trajectory were associated with HRs of 1.48 (95%-CI (1.10, 1.98)) and 1.77 (95%-CI (1.00, 3.15)) for incident DM. In addition, patients in the high-stable depression trajectory class who developed incident DM during follow-up were at 6.5-fold risk (HR = 6.51; 95%-CI (2.77, 15.3)) of experiencing a subsequent cardiovascular event. CONCLUSIONS: In patients with CHD, following a trajectory of high stable symptoms of depression was associated with an increased risk of incident DM. Furthermore, incident DM in these patients was associated with a substantially increased risk of subsequent CVE. Identifying depressive symptoms and pertinent treatment offers might be an important and promising approach to enhance outcomes in patients with CHD, which should be followed up in further research and practice.