Association between device-detected subclinical atrial fibrillation and diabetes in patients with implantable cardioverter-defibrillators: A propensity score-matched analysis

Background and objective: Device-detected subclinical atrial fibrillation (DDAF) and diabetes mellitus (DM) are common in patients with cardiac implantable devices. Our objective was to compare DDAF incidence between diabetic and non-diabetic patients with implantable cardioverter-defibrillator (ICD) or cardiac resynchronization therapy defibrillator (CRT-D), using propensity score (PS) matching to adjust for confounders. Methods: Data from the Home Monitoring Expert Alliance dataset were analyzed for patients with ICD or CRT-D and no prior clinical AF. The primary endpoint was time to the first DDAF, categorized by four different burden thresholds: ≥15 min, ≥5 h, ≥24 h, and ≥7 days. The impact of new-onset DDAF on mortality was also assessed. Results: Among 1619 patients (median age 69 years, 79 % male), 363 (22.4 %) had DM. Over a median follow-up of 2.3 years, DM patients had higher incidence of DDAF lasting ≥15 min (3-year: 65.7 % vs. 57.7 %, P = 0.032) and ≥5 h (3-year: 62.0 % vs. 52.2 %, P = 0.010) compared to non-diabetics; however, after PS matching, these differences were nonsignificant (P ≥ 0.15). No differences were found for ≥24 h and ≥7 days DDAF in unmatched or matched analyses. Using a landmark approach, DDAF was associated with increased mortality risk in DM patients (hazard ratio: 4.20, 95 %CI: 1.89–9.33, P < 0.001), whereas no effect was observed in non-diabetics (hazard ratio: 1.76, 95 %CI: 0.89–3.46, P = 0.099). Conclusions: In ICD/CRT-D recipients without AF history, DM was not linked to higher DDAF incidence. However, DDAF in diabetic patients was associated with poorer outcomes, emphasizing the importance of continuous atrial rhythm monitoring in this group.