Depressive symptoms and the risk of sudden cardiac death among the elderly

Aims Mental depression is associated with an increased risk of cardiovascular mortality, but the reasons for this association are not well known. We tested the hypothesis that depressive symptoms (DS) may specifically increase the risk of sudden cardiac death (SCD) among the elderly subjects. Methods All persons aged 70 years or over of a defined area in Northern Finland (n=1113) were the target population, and 915 (82%); 336 men and 579 women, participated in the questionnaire on depressive symptoms according to the Short Zung Depression Rating Scale. Potential risk factors of SCD were examined in the same connection, and the mode of death was examined via the official death certificates during the following eight years. SCD and non-SCD as well as overall mortality, cardiac mortality and non-fatal myocardial infarction (MI) were the end-points. Results Four hundred and seventy-six subjects out of 915 (52%) died during the follow-up of 8 years. Thirty-eight of deaths were SCDs (8%) and non-SCD occurred in 106 subjects (22%). In univariate analysis, a high score of DS was a significant predictor of subsequent SCD; hazard ratio 2.67 (95% confidence interval 1.34-5.32), non-SCD; 1.67 (1.06-2.63), cardiac mortality; 1.90 (1.30-2.78) and total mortality; 1.88 (1.52-2.32), but not that of non-fatal MI; 1.74 (0.89-3.38). SCD was also predicted by gender (P<0.01), MI (P<0.01) and tablet- or insulin-treated diabetes mellitus (P<0.01). In multivariate Cox regression analysis by adjusting for the clinical variables, high score of DS remained as a significant predictor of SCD; hazard ratio 2.74 (1.37-5.50) and total mortality; 1.70 (1.37-2.10), but not of cardiac mortality 1.50 (0.998-2.27), non-SCD; 1.38 (0.85-2.24) and non-fatal MI; 1.37 (0.69-2.71). Conclusion Depressive symptoms increase the risk of SCD, but not that of non-SCD and non-fatal MI among the elderly subjects. (C) 2003 Published by Elsevier Ltd on behalf of The European Society of Cardiology.