Our results found that the CHA2DS2-VASc score is a feasible tool used to predict NOAF during a long-term follow-up in patients with a history of MI. The hazard of developing NOAF within one year after MI was 4.6-fold higher in those with the highest CHA2DS2-VASc score (6 points) than in those with the lowest score (1 point). Importantly, the occurrence of NOAF after MI was associated with an increased risk of IS. The IS subdistribution hazard was 2.1-fold higher in MI survivors who developed NOAF during the study follow-up. Despite the potential influence of cofounding factors, our results suggest PCI to be the more favorable method for revascularization in relation to NOAF occurrence, as CABG was associated with a higher NOAF incidence at one year (1.9-fold) and, interestingly, at ten years (1.3-fold) as well.

A higher NOAF rate after MI compared to the general population has been well documented, with incidences ranging from 6% to 21% at five years post-MI [13]. Our findings are similar, as the NOAF rates were 4.1% within one year and 19.8% within ten years after MI. MI can cause irreversible structural and hemodynamic changes that manifest as clinical arrhythmias, most commonly within the first year after MI [3]. This temporal trend was also observed in our study. However, myocardial scarring may also act as a substrate for arrhythmias after index MI, resulting in the high incidence of NOAF observed in our long-term follow-up [14].Several independent risk factors for AF have been identified in the general population, but fewer have been identified for MI survivors [5]. Most studies address NOAF occurrence during MI hospitalization or compare patients with a history of AF and those who develop AF during follow-up [4]. A recent study found that a history of angina, worse patient-reported quality of life, European origin, and a bleeding event prior to index MI all predicted AF incidence over two years of follow-up [4]. The focus of our study was the investigation of the applicability of the commonly used CHA2DS2-VASc score used to predict NOAF in patients with a history of either STEMI (ST-elevation myocardial infarction) or non-STEMI. Our results show that the CHA2DS2-VASc score can indeed be used to identify at-risk patients to yield more targeted follow-ups and rhythm monitoring to detect AF. In 2014, Lau et al. studied the usefulness of the CHA2DS2-VASc score in predicting NOAF and IS in post-STEMI patients [15]. Conversely, our study results apply to all MI survivors, covering a much larger and clinically important population. Considering the high incidence of AF and IS risk in MI survivors, our findings emphasize the significance of MI as the only CHA2DS2-VASc point when considering whether to initiate anticoagulation in those with paroxysmal AF. AF develops in 20–40% of patients within the first week after CABG. Revascularization by CABG has been found to be associated with higher NOAF rates compared to patients who undergo PCI [16]. Furthermore, patients suffering from postoperative AF are at an increased risk of IS [17]. Despite the fact that most studies report AF episodes occurring during operation-related hospitalization, we found a similar and long-lasting pattern of NOAF occurrence after the index hospital visit. Our results show a 1.9-fold increase in cumulative NOAF incidence after CABG compared to PCI at one year and a 1.3-fold increase at ten years, whereas the highest cumulative NOAF incidence was recorded in patients assigned to noninterventional treatment (optimal medical therapy without PCI or CABG). It is reasonable to argue that patients undergoing CABG most likely suffer from multivessel disease more often than PCI patients and may have more comorbidities (beyond our hazard ratio adjustments), and thus healthier patients are selected for PCI. In addition, secondary prevention after MI is important when considering AF risk factor management (e.g., blood pressure, lipid levels, and diabetes) [18]. Consequently, there may also be differences in secondary preventive treatments prescribed by cardiologists vs. surgeons, which could have an influence on the observed differences in NOAF rates.Patients with a history of MI are also at an increased risk of IS during the first three months after MI [19]. Previous data have identified several cardiogenic mechanisms behind this risk [20], most of which appear to be attenuated after the first month following MI, whereas the importance of AF increases thereafter [19]. According to our results, the hazard of IS is two times higher in patients who develop NOAF after MI compared to those who remain in a sinus rhythm. Previous studies have found IS to occur most commonly within the first year after AF diagnosis, while in up to 20% of IS survivors, the IS is the first manifestation of AF [21,22]. NOAF manifested as IS in 4% of our study patients. Of the patients who developed NOAF after MI, increasing age and prior cerebral ischemia were identified as long-lasting predictors of IS as they also remained significant at 10 years. Female sex was a risk factor within a year following MI, but not within ten years. This is probably because some individual IS risk factors are more significant IS predictors in women than in men, but the effect gets attenuated as competing IS risk factors in men, such as atherosclerosis, develop over time [23,24]. Therefore, early efforts to identify AF after MI are especially important in women. The need for careful patient follow-up is emphasized by the finding that IS and NOAF often occurred simultaneously, that is, without warning.The incidence of IS was lowest in patients who underwent revascularization by PCI compared to those treated with CABG or medical therapy. This has also been suggested previously [25]. In addition, recent findings of Head et al. reported a significantly lower stroke rate five years after PCI vs. CABG in a pooled analysis of randomized studies comparing PCI and CABG, regardless of rhythm status [26]. Considering that NOAF occurs less frequently after PCI and patients usually have fewer comorbidities, the lower IS rate during long-term follow-up after PCI makes sense [27]. The antiplatelet regimens used after PCI [28] would not be expected to lower the risk of (cardio)embolic strokes, but they might lower the risk of atherosclerotic IS, which is more common in men and might also contribute to the observed sex difference within the first year following MI.Our study has strengths and limitations. The major strength is the population-based design, which included nearly all patients with MI in Finland during a 14-year period. The major limitations are the retrospective design and use of registry data. We did not have access to more detailed clinical data and lacked blood pressure, laboratory, imaging, and angiographical findings. Although we used an extended version of a previously validated method to detect CHA2DS2-VASc components [9], it is possible that the true prevalence of some components, especially hypertension, could be underestimated in our data. Additionally, we did not have data on the evolvement of the CHA2DS2-VASc score during follow-up. Atrial fibrillation was detected using registries of specialist health care and death certificates, [11] but we did not have access to primary health care registries. A recent Finnish study showed that 29]. The proportion of undetectable NOAF patients among MI survivors is likely to be even lower. In addition, information on OAC usage was not available in the current study. An inherent limitation of administrative registries is related to coding errors. However, the large number of patients makes it unlikely that these errors would significantly influence our main findings. We did not have data on the ethnic backgrounds of patients, but because the Finnish population is predominantly white, the generalizability of our results to more diverse populations may be limited.