Somatic and sociodemographic predictors of depression outcome among depressed patients with coronary artery disease - a secondary analysis of the SPIRR-CAD study.

BACKGROUND: Depressive symptoms are common in patients with coronary artery disease (CAD) and are associated with an unfavourable outcome. Establishing prognostic patient profiles prior to the beginning of mental health care may facilitate higher efficacy of targeted interventions. The aim of the current study was to identify sociodemographic and somatic predictors of depression outcome among depressed patients with CAD. METHODS: Based on the dataset of the multicentre SPIRR-CAD randomised controlled trial (n = 570 patients with CAD and ≥ 8 points on the Hospital Anxiety and Depression Scale (HADS)), 141 potential sociodemographic and somatic predictors of the change in the HADS-D depression score from baseline to 18-month-follow-up were derived in two different ways. We screened for univariable association with response, using either analysis of (co)variance or logistic regression, respectively, both adjusted for baseline HADS-D value and treatment group. To guard against overfitting, multivariable association was evaluated by a linear or binomial (generalised) linear model with lasso regularisation, a machine learning approach. Outcome measures were the change in continuous HADS-D depression scores, as well as three established binary criteria. The Charlson Comorbidity Index (CCI) was calculated to assess possible influences of comorbidities on our results and was also entered in our machine learning approach. RESULTS: Higher age (p = 0.002), unknown previous myocardial infarction (p = 0.013), and a higher heart rate variability during numeracy tests (p = .020) were univariably associated with a favourable depression outcome, whereas hyperuricemia (p ≤ 0.003), higher triglycerides (p = 0.014), NYHA class III (p ≤ 0.028), state after resuscitation (p ≤ 0.042), intake of thyroid hormones (p = 0.007), antidiabetic drugs (p = 0.015), analgesic drugs (p = 0.027), beta blockers (p = 0.035), uric acid drugs (p ≤ 0.039), and anticholinergic drugs (p = 0.045) were associated with an adverse effect on the HADS-D depression score. In all analyses, no significant differences between study arms could be found and physical comorbidities also had no significant influence on our results. CONCLUSION: Our findings may contribute to identification of somatic and sociodemographic predictors of depression outcome in patients with CAD. The unexpected effects of specific medication require further clarification and further research is needed to establish a causal association between depression outcome and our predictors. TRIAL REGISTRATION: www.clinicaltrials.gov NCT00705965 (registered 27th of June, 2008). www.isrctn.com ISRCTN76240576 (registered 27th of March, 2008).

Prevalence of mental disorders among depressed coronary patients with and without Type D personality. Results of the multi-center SPIRR-CAD trial.

OBJECTIVE: Type D personality, as with formal mental disorders, is linked to increased mortality in coronary heart disease (CHD). Our aim was to determine the prevalence of mental disorders among depressed CHD patients with and without Type D personality. METHODS: Depressive symptoms (HADS, HAM-D), Type D personality (DS-14) and mental disorders based on DSM-IV (SCID I and II) were assessed. Results were calculated by Kruskal-Wallis tests, Fisher's exact tests and logistic regression analyses. RESULTS: 570 CHD patients were included (age 59.2±9.5years; male 78.9%, HADS-D depression 10.4±2.5; HAM-D 11.3±6.6; Type D 60.1%). 84.8% of patients with Type D personality and 79.3% of non-Type D patients suffered from at least one mental disorder (p=0.092), while 41.8% of Type D positives and 27.8% of Type D negatives had at least two mental disorders (p=0.001). Patients with Type D personality significantly more often had social phobia [odds ratio (95% confidence interval): 3.79 (1.1 to 13.12); p=0.035], dysthymia [1.78 (1.12 to 2.84); p=0.015], compulsive [2.25 (1.04 to 4.86); p=0.038] or avoidant [8.95 (2.08 to 38.49); p=0.003] personality disorder. CONCLUSIONS: Type D personality among depressed CHD patients is associated with more complex and enduring mental disorders. This implies higher treatment demands. TRIAL REGISTRATION: ISRCTN 76240576; NCT00705965.