Discussion
Our findings indicate that in African American patients with cryptogenic stroke detection of AF occurs earlier and yet at similar rate that in other populations1,7-9. Brachmann et al, specifically found near identical detection rates in the ILR monitoring arm at 36 months (30% vs 29%), but detection occurred later after implantation compared to our population (median time to detection = 8.4 months vs 37.5 days. Furthermore, the detection rate after 1 month was 3.7% in this study compared to around 14% in our population.9 In the CRYSTAL-AF trial at 1 month out from implantation the detection rate was also lower that in our population with only about 4%. 1
Given the retrospective nature and lack of control group we are not able to conclude with certainty that the detection of AF would occur earlier after implantation that in other ethnic groups. Our patients had more comorbidities as indicated by a higher CHADS2VASc than those in the CRYSTAL-AF trial (4.4 vs 3.0) and Brachmann et al. yet similar rates of AF detection suggesting a “protective effect”.1,9 African Americans appear to have an increased risk of developing stroke in the setting of AF independent of the relative higher prevalence of comorbidities in this population.10
Those with detected AF had a numerical older age, greater prevalence of prior stroke, and incidence of CAD but none of these results were statistically significant. This is likely related to insufficient power to detect a difference as multiple other studies have demonstrated a correlation between these variables and an increased detection rate of AF. 8,11 No electrocardiographic or echocardiographic parameters were associated with an increased detection rate in this group despite some parameters like PR interval prolongation, LA size and frequent premature atrial beats have been reported as predictors of AF detection in other studies. 8, 11, 12
When a multivariate regression analysis was performed including variables that approached significance the presence of CAD, hyperlipidemia, age, and nonuse of calcium channel blockers demonstrated predictive capacity. It is uncertain though what mechanism may be involved but CCB are known to decrease aldosterone synthesis and that in turn might decrease the odds of developing AF. 13 The combination of these factors demonstrating significance might have been driven mainly by the combination of age and CAD which are well documented predictors of development of atrial fibrillation.14
In conclusion, it appears that extended electrocardiographic monitoring with an implantable loop recorder it as effective for the detection of AF in African Americans as in the general population, time to detection is shorter and baseline characteristics were unable to predict the detection of AF.