Does sympathetic hyperactivity adversely impact on the effect of implantable cardioverter-defibrillator in patients with diabetes and non-ischaemic systolic heart failure?

In their elegant study ‘The effect of implantable cardioverter-defibrillator in patients with diabetes and non-ischaemic systolic heart failure’, Rørth et al.1 reported that diabetic patients had a higher incidence of all-cause mortality which was mostly driven by increased cardiovascular (CV) mortality including sudden cardiac death (SCD). The treatment effect of implantable cardioverter-defibrillator therapy was not significantly modified by diabetes, which the authors suggest may be due to a lack of power. Throughout follow-up, more diabetic patients died compared to those without diabetes. Death from CV causes and SCD were also more frequent in patients with diabetes compared to their non-diabetic counterparts. In unadjusted analyses, diabetes was associated with an increased risk of all three outcomes with hazard ratios (HRs) of 1.77 (95% CI 1.34–2.33); P < 0.0001 for all-cause mortality, 2.31 (95% CI 1.68–3.18); P < 0.0001 for CV death, and 2.29 (95% CI 1.39–3.76); P = 0.001 for SCD. Nonetheless, the risks of all-cause mortality and SCD in patients with diabetes were no longer significant in adjusted Cox regression analyses, whereas the risk of CV death (HR 1.59, 95% CI 1.14–2.23; P = 0.007) remained significantly increased.1 In this context, it is relevant to note that at baseline, conditions characterized by sympathetic overactivity, such as long-standing heart failure (HF), obesity, hypertension, and reduced estimated glomerular filtration rate (eGFR) were significantly more common in diabetic patients. Furthermore, loop diuretics known particularly in higher doses to enhance neurohormonal activation in HF were more frequently prescribed in patients with diabetes.2 The primary mechanism of action of loop diuretics, that is, inhibiting sodium reabsorption in the ascending loop of Henle, triggers tubuloglomerular feedback and renin release thereby further eliciting neurohormonal activation and intensifying the hypersympathetic state.3 Each of these factors can trigger ventricular arrhythmias and contribute to SCD, particularly in combination. We would, therefore, expect that in diabetic patients with higher rates of chronic kidney disease (CKD), the rate of SCD should be higher than in non-diabetics.4 On the other hand, patients without diabetes were less frequently on statin therapy, which has been suggested as a predictor of both cardiac death and major CV adverse events in HF.5 Importantly, in their manuscript, Rørth et al.1 described that the Cox regression models were adjusted for age, sex, treatment arm, New York Heart Association class, HF duration, smoking status, left ventricular ejection fraction, body mass index, systolic blood pressure, heart rate, loge (NT-proBNP), eGFR, history of chronic obstructive pulmonary disease, stroke, atrial fibrillation, and hypertension. However, important variables with significant between-group differences, such as CKD (defined as eGFR <60 mL/min/1.73 m2), statin intake, loop diuretic use, and its dose not included in the models. Furthermore, other parameters that inhibit overall sympathetic nervous system activity, such as β-blockers and mineralocorticoid-receptor antagonists, should perhaps have been included. A repeat adjusted Cox regression analysis, including the aforementioned factors, might, therefore, reveal a different risk ratio between diabetic and non-diabetic patients for the outcomes under investigation. M.P.S. is supported by an NHMRC Research Fellowship and has received consulting fees, and/or travel and research support from Medtronic, Abbott, Novartis, Servier, Pfizer, and Boehringer-Ingelheim. Conflict of interest: none declared.