Objective: To assess the risk of liver injury hospitalisation in patients with atrial fibrillation (AF) after initiation of direct oral anticoagulants (DOACs) or warfarin and to determine predictors of liver injury hospitalisation in this population. Methods: We studied 113 717 patients (mean age 70, 39% women) with AF included in the MarketScan Commercial and Medicare Supplemental databases with a first prescription for oral anticoagulation after 4 November 2011, followed through 31 December 2014. Of these, 56 879 initiated warfarin, 17 286 initiated dabigatran, 30 347 initiated rivaroxaban and 9205 initiated apixaban. Liver injury hospitalisation and comorbidities were identified from healthcare claims. Results: During a median follow-up of 12 months, 960 hospitalisations with liver injury were identified. Rates of liver injury hospitalisation ( per 1000 person-years) by oral anticoagulant were 9.0 (warfarin), 4.0 (dabigatran), 6.6 (rivaroxaban) and 5.6 (apixaban). After multivariable adjustment, liver injury hospitalisation rates were lower in initiators of DOACs compared with warfarin: HR (95% CI) of 0.57 (0.46 to 0.71), 0.88 (0.75 to 1.03) and 0.70 (0.50 to 0.97) for initiators of dabigatran, rivaroxaban, and apixaban, respectively (vs. warfarin). Compared with dabigatran initiators, rivaroxaban initiators had a 56% increased risk of liver injury hospitalisation (HR 1.56, 95% CI 1.22 to 1.99). In addition to type of anticoagulant, prior liver, gallbladder and kidney disease, cancer, anaemia, heart failure and alcoholism significantly predicted liver injury hospitalisation. A predictive model including these variables had adequate discriminative ability (C-statistic 0.67, 95% CI 0.64 to 0.70). Conclusions: Among patients with non-valvular AF, DOACs were associated with lower risk of liver injury hospitalisation compared with warfarin, with dabigatran showing the lowest risk.

Background Patients with atrial fibrillation (AF) experience an increased risk of heart failure (HF). However, data are lacking on current trends in the risk of HF after AF. Objective The purpose of this study was to describe the temporal trends in HF occurrence after AF in a community cohort of patients with incident AF from 2000 to 2013. Methods Cox regression was used to examine the association of year of AF diagnosis with HF and the predictors of developing HF after AF. Results Among 3491 AF patients without prior HF, 750 (21%) developed incident HF over mean follow-up of 3.7 years. Among those with an echocardiogram, 422 (61%) had HF with preserved ejection fraction (HFpEF), and 270 (39%) had HF with reduced ejection fraction (HFrEF). After adjusting for demographics and comorbidities, the risk of developing HF did not change over time (hazard ratio [HR] (95% confidence interval [CI]) per year of AF diagnosis: 1.01 (0.98–1.03) overall; 1.00 (0.98–1.03) for HFpEF; 1.00 (0.96–1.03) for HFrEF). Increasing age, obesity, smoking, diabetes, chronic pulmonary disease, and renal disease were predictors of developing HF. Compared to the Olmsted County, Minnesota, population, a substantial excess risk of developing HF was observed after AF diagnosis [standardized morbidity ratio (95% CI): 9.60 (7.44–12.19), 2.13 (1.56–2.84), and 1.70 (1.34–2.14) at 90 days, 1 year, and 3 years after diagnosis]. Conclusion In the community, HF is a frequent adverse outcome among patients with AF, and HFpEF is more common than HFrEF. The rates of HF after AF have not declined, thus highlighting the importance of continued efforts to improve outcomes in AF.

Background Prolongation of the QT interval has been associated with an increased risk of developing atrial fibrillation (AF), but the responsible mechanism remains unknown. Objectives The aims of this study were to subdivide the QT interval into its components and identify the resultant electrocardiographic interval(s) responsible for the association with AF. Methods Predefined QT-interval components were assessed for association with incident AF in the Atherosclerosis Risk in Communities study using Cox proportional hazards models. Hazard ratios (HRs) were calculated per 1-SD increase in each component. Among QT-interval components exhibiting significant associations, additional analyses evaluating long extremes, defined as greater than the 95th percentile, were performed. Results Of the 14,625 individuals, 1505 (10.3%) were diagnosed with incident AF during a mean follow-up period of 17.6 years. After multivariable adjustment, QT-interval components involved in repolarization, but not depolarization, exhibited significant associations with incident AF, including a longer ST segment (HR 1.27; 95% confidence interval [CI] 1.14–1.41; P <.001) and a prolonged T-wave onset to T-wave peak (T-onset to T-peak) (HR 1.13; 95% CI 1.07–1.20; P <.001). Marked prolongation of the ST segment (HR 1.31; 95% CI 1.04–1.64; P =.022) and T-onset to T-peak (HR 1.36; 95% CI 1.09–1.69; P =.006) was also associated with an increased risk of incident AF. Conclusion The association between a prolonged QT interval and incident AF is primarily explained by components involved in ventricular repolarization: prolongation of the ST segment and T-onset to T-peak. These observations suggest that prolongation of phases 2 and 3 of the cardiac action potential drives the association between the QT interval and AF risk.

Background Circulating NT-proBNP (N-terminal pro-B-type natriuretic peptide) levels, a well-known indicator of atrial wall stress and remodeling, inversely correlate with body mass index. Both are strongly predictive of atrial fibrillation (AF). Their potential interaction in relation to incident AF, however, has not been explored. Methods and Results In total, 9556 participants of the ARIC (Atherosclerosis Risk in Communities) study who had 2 measurements of NT-proBNP and no baseline AF or heart failure were followed from 1996 to 1998 through 2016 for the occurrence of incident AF. Participants were categorized as obese (body mass index ≥30) and nonobese (body mass index <30) and by NT-proBNP levels (using the median of 68.2 pg/mL as the cutoff). Over a median follow-up of 18.3 years, we identified 1806 incident cases of AF. Analysis using multivariable Cox regression models showed that obese participants with high NT-proBNP levels at visit 4 had a higher adjusted risk of incident AF (hazard ratio: 3.64; 95% CI, 3.15-4.22) compared with nonobese individuals with low NT-proBNP levels. The association of obesity with AF risk was not modified by NT-proBNP levels (P=0.46 for interaction). Increasing BNP among participants from 1990-1992 to 1996-1998 was associated with increased AF risk. After further adjustment for clinical risk factors and medications, results were similar. Conclusions Individuals who had both elevated body mass index and NT-proBNP and were free of clinically recognized heart failure were at higher risk of AF development. Those who experienced an increase in NT-proBNP levels between visits 2 and 4 were at higher risk of AF.

BACKGROUND: Atrial fibrillation (AF) has been associated with faster cognitive decline and increased dementia risk. Factors associated with dementia in patients with AF have been seldom studied. METHODS AND RESULTS: We studied 6432 individuals from the ARIC-NCS (Atherosclerosis Risk in Communities Neurocognitive Study). In 2011 to 2013, participants underwent a physical exam, echocardiography, detailed cognitive assessments, and a subset, brain magnetic resonance imaging. Dementia and mild cognitive impairment (MCI), as well as etiology of MCI/dementia, Alzheimer's disease-related or vascular, were adjudicated by an expert panel. AF was defined by study ECGs and past hospitalizations. We used logistic regression to estimate odds ratios and 95% CI of MCI/dementia by AF status and to assess cross-sectional correlates of MCI/dementia in patients with AF. Among 6432 participants, 611 (9.5%) had prevalent AF. AF was associated with increased odds of dementia and MCI (odds ratio, 95% CI, 2.25, 1.64-3.10, and 1.28, 1.04-1.56, respectively). Prevalence of Alzheimer's disease-related MCI/dementia and vascular MCI/dementia were higher in participants with AF than without AF (odds ratio, 95% CI, 1.29, 1.04-1.61, and 1.50, 0.99-2.25, respectively). In multivariable analyses, older age, lower body mass index, diabetes mellitus, stroke, and APOE genotype were associated with dementia prevalence in participants with AF. In models evaluating MCI/dementia subtypes, diabetes mellitus was associated with Alzheimer's disease-related MCI/dementia, whereas male sex and stroke were risk factors for vascular MCI/dementia. CONCLUSIONS: In a large, community-based study, AF was associated with higher prevalence of MCI and dementia. Controlling cardiometabolic risk factors is a potential target for prevention of adverse cognitive outcomes in AF patients.

BACKGROUND: Atrial fibrillation and heart failure are 2 of the most common diseases, yet ready means to identify individuals at risk are lacking. The 12-lead ECG is one of the most accessible tests in medicine. Our objective was to determine whether a premature atrial contraction observed on a standard 12-lead ECG would predict atrial fibrillation and mortality and whether a premature ventricular contraction would predict heart failure and mortality. METHODS AND RESULTS: We utilized the CHS (Cardiovascular Health) Study, which followed 5577 participants for a median of 12 years, as the primary cohort. The ARIC (Atherosclerosis Risk in Communities Study), the replication cohort, captured data from 15 792 participants over a median of 22 years. In the CHS, multivariable analyses revealed that a baseline 12-lead ECG premature atrial contraction predicted a 60% increased risk of atrial fibrillation (hazard ratio, 1.6; 95% CI, 1.3-2.0; P<0.001) and a premature ventricular contraction predicted a 30% increased risk of heart failure (hazard ratio, 1.3; 95% CI, 1.0-1.6; P=0.021). In the negative control analyses, neither predicted incident myocardial infarction. A premature atrial contraction was associated with a 30% increased risk of death (hazard ratio, 1.3; 95% CI, 1.1-1.5; P=0.008) and a premature ventricular contraction was associated with a 20% increased risk of death (hazard ratio, 1.2; 95% CI, 1.0-1.3; P=0.044). Similarly statistically significant results for each analysis were also observed in ARIC. CONCLUSIONS: Based on a single standard ECG, a premature atrial contraction predicted incident atrial fibrillation and death and a premature ventricular contraction predicted incident heart failure and death, suggesting that this commonly used test may predict future disease.

BACKGROUND: Rivaroxaban is an oral anticoagulant approved in the US for prevention of stroke and systemic embolism in patients with non-valvular atrial fibrillation (NVAF). We determined the effectiveness and associated risks of rivaroxaban versus other oral anticoagulants in a large real-world population. METHODS: We selected NVAF patients initiating oral anticoagulant use in 2010-2014 enrolled in MarketScan databases. Rivaroxaban users were matched with warfarin and dabigatran users by age, sex, enrolment date, anticoagulant initiation date, and high-dimensional propensity score. Study endpoints, including ischemic stroke, intracranial bleeding (ICB), myocardial infarction (MI), and gastrointestinal (GI) bleeding, were identified from inpatient diagnostic codes. Multivariable Cox models were used to assess associations between type of anticoagulant and outcomes. RESULTS: The analysis included 44,340 rivaroxaban users matched to 89,400 warfarin and 16,957 dabigatran users (38% female, mean age 70) with 12 months of mean follow-up. Anticoagulant-naïve rivaroxaban initiators, but not those switching from warfarin, had lower risk of ischemic stroke [hazard ratio (HR) (95% confidence interval (CI)): 0.75 (0.62, 0.91)] and ICB [HR (95%CI): 0.55, (0.39, 0.78)] than warfarin users. In contrast, anticoagulant-experienced rivaroxaban initiators had higher risk of GI bleeding than warfarin users [HR (95%CI): 1.55 (1.32, 1.83)]. Endpoint rates were similar when comparing anticoagulant-naïve rivaroxaban and dabigatran initiators, with the exception of higher GI bleeding risk in rivaroxaban users [HR (95%CI) 1.28 (1.06, 1.54)]. There were no significant differences in the risk of MI among the comparison groups. CONCLUSION: In this large real-world sample of NVAF patients, effectiveness and risks of rivaroxaban versus warfarin differed by prior anticoagulant status, while effectiveness of rivaroxaban versus dabigatran differed in GI bleeding risk.

Background: People living with HIV (PLWH) experience higher risk of myocardial infarction (MI) and heart failure (HF) compared with uninfected individuals. Risk of other cardiovascular diseases (CVDs) in PLWH has received less attention. Methods and Results: We studied 19 798 PLWH and 59 302 age- and sex-matched uninfected individuals identified from the MarketScan Commercial and Medicare databases in the period 2009 to 2015. Incidence of CVDs, including MI, HF, atrial fibrillation, peripheral artery disease, stroke and any CVD-related hospitalization, were identified using validated algorithms. We used adjusted Cox models to estimate hazard ratios and 95% CIs of CVD end points and performed probabilistic bias analysis to control for unmeasured confounding by race. After a mean follow-up of 20 months, patients experienced 154 MIs, 223 HF, 93 stroke, 397 atrial fibrillation, 98 peripheral artery disease, and 935 CVD hospitalizations (rates per 1000 person-years: 1.2, 1.7, 0.7, 3.0, 0.8, and 7.1, respectively). Hazard ratios (95% CI) comparing PLWH with uninfected controls were 1.3 (0.9–1.9) for MI, 3.2 (2.4–4.2) for HF, 2.7 (1.7–4.0) for stroke, 1.2 (1.0–1.5) for atrial fibrillation, 1.1 (0.7–1.7) for peripheral artery disease, and 1.7 (1.5–2.0) for any CVD hospitalization. Adjustment for unmeasured confounding led to similar associations (1.2 [0.8–1.8] for MI, 2.8 [2.0–3.8] for HF, 2.3 [1.5–3.6] for stroke, 1.3 [1.0–1.7] for atrial fibrillation, 0.9 [0.5–1.4] for peripheral artery disease, and 1.6 [1.3–1.9] for CVD hospitalization). Conclusions: In a large health insurance database, PLWH have an elevated risk of CVD, particularly HF and stroke. With the aging of the HIV population, developing interventions for cardiovascular health promotion and CVD prevention is imperative.

Background: Timing and trajectories of cardiovascular risk factor (CVRF) development in relation to atrial fibrillation (AF) have not been described previously. We assessed trajectories of CVRF and incidence of AF over 25 years in the ARIC study (Atherosclerosis Risk in Communities). Methods: We assessed trajectories of CVRF in 2456 individuals with incident AF and 6414 matched control subjects. Subsequently, we determined the association of CVRF trajectories with the incidence of AF among 10 559 AF-free individuals (mean age, 67 years; 52% men; 20% blacks). Risk factors were measured during 5 examinations between 1987 and 2013. Cardiovascular events, including incident AF, were ascertained continuously. We modeled the prevalence of risk factors and cardiovascular outcomes in the period before and after AF diagnosis and the corresponding index date for control subjects using generalized estimating equations. Trajectories in risk factors were identified with latent mixture modeling. The risk of incident AF by trajectory group was examined with Cox models. Results: The prevalence of stroke, myocardial infarction, and heart failure increased steeply during the time close to AF diagnosis. All CVRFs were elevated in AF cases compared with controls > 15 years before diagnosis. We identified distinct trajectories for all the assessed CVRFs. In general, individuals with trajectories denoting long-term exposure to CVRFs had increased AF risk even after adjustment for single measurements of the CVRFs. Conclusions: AF patients have increased prevalence of CVRF many years before disease diagnosis. This analysis identified diverse trajectories in the prevalence of these risk factors, highlighting their different roles in AF pathogenesis.

Background: Abnormal atrial repolarization is important in the development of atrial fibrillation (AF), but no direct measurement is available in clinical medicine. Objective: To determine whether the QT interval, a marker of ventricular repolarization, could be used to predict incident AF. Methods: We examined a prolonged QT interval corrected by using the Framingham formula (QTFram) as a predictor of incident AF in the Atherosclerosis Risk in Communities (ARIC) study. The Cardiovascular Health Study (CHS) and Health, Aging, and Body Composition (ABC) study were used for validation. Secondary predictors included QT duration as a continuous variable, a short QT interval, and QT intervals corrected by using other formulas. Results: Among 14,538 ARIC study participants, a prolonged QTFram predicted a roughly 2-fold increased risk of AF (hazard ratio [HR] 2.05; 95% confidence interval [CI] 1.42-2.96; P <.001). No substantive attenuation was observed after adjustment for age, race, sex, study center, body mass index, hypertension, diabetes, coronary disease, and heart failure. The findings were validated in Cardiovascular Health Study and Health, Aging, and Body Composition study and were similar across various QT correction methods. Also in the ARIC study, each 10-ms increase in QTFram was associated with an increased unadjusted (HR 1.14; 95% CI 1.10-1.17; P <.001) and adjusted (HR 1.11; 95% CI 1.07-1.14; P <.001) risk of AF. Findings regarding a short QT interval were inconsistent across cohorts. Conclusions: A prolonged QT interval is associated with an increased risk of incident AF.