BACKGROUND: The DANISH study (Danish Study to Assess the Efficacy of ICDs [Implantable Cardioverter Defibrillators] in Patients With Non-Ischemic Systolic Heart Failure on Mortality) did not demonstrate an overall effect on all-cause mortality with ICD implantation. However, the prespecified subgroup analysis suggested a possible age-dependent association between ICD implantation and mortality with survival benefit seen only in the youngest patients. The nature of this relationship between age and outcome of a primary prevention ICD in patients with nonischemic systolic heart failure warrants further investigation. METHODS: All 1116 patients from the DANISH study were included in this prespecified subgroup analysis. We assessed the relationship between ICD implantation and mortality by age, and an optimal age cutoff was estimated nonparametrically with selection impact curves. Modes of death were divided into sudden cardiac death and nonsudden death and compared between patients younger and older than this age cutoff with the use of χ2 analysis. RESULTS: Median age of the study population was 63 years (range, 21-84 years). There was a linearly decreasing relationship between ICD and mortality with age (hazard ratio [HR], 1.03; 95% confidence interval [CI], 1.003-1.06; P=0.03). An optimal age cutoff for ICD implantation was present at ≤70 years. There was an association between reduced all-cause mortality and ICD in patients ≤70 years of age (HR, 0.70; 95% CI, 0.51-0.96; P=0.03) but not in patients >70 years of age (HR, 1.05; 95% CI, 0.68-1.62; P=0.84). For patients ≤70 years old, the sudden cardiac death rate was 1.8 (95% CI, 1.3-2.5) and nonsudden death rate was 2.7 (95% CI, 2.1-3.5) events per 100 patient-years, whereas for patients >70 years old, the sudden cardiac death rate was 1.6 (95% CI, 0.8-3.2) and nonsudden death rate was 5.4 (95% CI, 3.7-7.8) events per 100 patient-years. This difference in modes of death between the 2 age groups was statistically significant (P=0.01). CONCLUSIONS: In patients with systolic heart failure not caused by ischemic heart disease, the association between the ICD and survival decreased linearly with increasing age. In this study population, an age cutoff for ICD implantation at ≤70 years yielded the highest survival for the population as a whole.