Table 3.
Risk Factors for Diagnosed AF
| Condition | Study Type | Effect on Risk of AF | Summary Risk of Incident AF | Effect of LRFM |
|---|---|---|---|---|
| Risk Factors | ||||
| Advancing age | SR/MA | Age per 5 y: ↑ risk (HR, 1.43–1.66)2,3 | ↑ Risk | N/A |
| MR | Accelerated epigenetic age by MR: no association4 | |||
| Smoking | Single study | Current smoking: ↑ risk (9.8%)5 | ↑ Risk | N/A |
| SR/MA | Smoking: ↑ risk (HR, 1.21–1.43)2,6 | |||
| MR | Smoking initiation: ↑ risk (OR, 1.11)7 | |||
| Physical activity | SR/MA | Sedentary lifestyle: ↑ risk (OR, 2.47)8 Guideline-recommended physical activity: ↓ risk (HR, 0.94)9 Elite athletes vs nonathletes: ↑ risk (OR, 2.46)10 |
U curve: Sedentary lifestyle and elite/extreme exercise: ↑ risk | Exercise: ↓ AF burden, recurrence, symptoms; ↑ quality of life, functional capacity11–16 |
| Alcohol | Single studies | Risk of AF episode within 4 h of 1 drink: ↑ risk (OR, 2.02)17 Greater access to alcohol law: ↑ risk18 |
↑ Risk | Randomized abstinence: ↑ AF recurrence and burden19 N-of-1 studies of alcohol avoidance: ↓ near-term AF20 Alcohol avoidance or reduction as part of a comprehensive LRFM program: ↓ AF burden, symptoms, progression of AF21–24 |
| SR/MA | Dose response (#drinks/d): ↑ risk (RR) 1: 1.08; 2: 1.17; 3: 1.33; 4: 1.36; 5: 1.4725 | |||
| MR | Genetically predicted heavy alcohol consumption (>35 U/wk for women and >50 U/wk for men): ↑ risk (OR, 1.11)7 | |||
| Adiposity markers: weight, BMI, obesity | Single study | Obesity: population attributable fraction 12.7%–16.9%5,26 | ↑ Risk | Weight loss in overweight or obese patients with AF as part of a comprehensive LRFM program: ↓ AF symptoms, burden, recurrence, progression21–24 Bariatric surgery in class III obesity: associated with reversal of AF type, ↑ sinus rhythm postablation27–29 Weight loss in long-lasting persistent AF and obesity: ↔30 |
| SR/MA | BMI: RR, 1.28 per 5-unit ↑ in BMI31 Weight:2 HR, 1.12 per 15 kg ↑ |
|||
| MR | Obesity3 Birthweight: 1.26 per SD ↑32 Childhood BMI (OR, 1.18)32 BMI 1.31 per unit BMI33 |
|||
| Height | MA | Height per 10 cm: ↑ risk (HR, 1.28)2 | ↑ Risk | N/A |
| SR/MA | Increasing height: ↑ risk3 | |||
| MR | Increasing height: ↑ risk (OR per unit, 1.33)33 | |||
| Hypertension and BP | Single studies | Elevated BP: ↑ risk, population attributable fraction, 21.6%5 Presence of hypertension treatment: ↑ risk (HR, 1.35–1.68), incidence 9.8%–19.5%; both AF and SBP decreased over time26 |
Hypertension: ↑ risk SBP: ↑ risk DBP: ↑↓↔ risk |
Renal denervation: ↓ AF postablation34 Mineralocorticoid receptor antagonists: ↓ AF burden35 BP control postablation: ↔36 Intensive BP control to SBP <120 mm Hg in patients with hypertension at high risk for CVD: ↓ AF risk37 BP control as part of a comprehensive LRFM program: ↓ AF burden21–24,38 |
| MA | BP: SBP: ↑ risk (HR per 20 mm Hg, 1.22); DBP per 10 mm Hg ↓ risk (HR, 0.90); use of BP medications ↑ risk (HR, 1.42)2 |
|||
| SR/MA | Hypertension: ↑ risk3 | |||
| MR | SBP33,39 ↑ risk; DBP mixed results ↔↑ risk39,40; pulse pressure ↑ risk40 | |||
| Resting heart rate | SR/MA | Resting heart rate: J-shaped relationship with incident AF. Lowest risk at 68–80 bpm; <70 bpm (RR, 1.09 per 10 bpm ↓); >70 bpm (RR, per 10 bpm ↑ RR 1.06)41 | Slow heart rate: ↑↓ variable risk Higher heart rate: ↑↓ variable risk |
N/A |
| MR | Heart rate: <65 bpm slower (HR ↑ risk); heart rate per 5 bpm ↑, 0.8242 | |||
| Diabetes | Single study | Diabetes: ↑ risk, population attributable fraction 3.1%5 Diabetes: ↑ risk, population attributable fraction ↑ over time 3.2%–5.9%26 |
↑ Risk | Optimal glycemic control preablation may ↓ AF recurrence postablation43 |
| MA | Diabetes: ↑ risk (HR, 1.27 [95% CI, 1.10–1.46])2 | |||
| SR/MA | Diabetes: ↑ risk (RR, 1.28, excluding large outlying study)44 Pre-diabetes: ↑ risk (RR, 1.20)44 Blood glucose; ↑ risk (RR per 20 mg/dL ↑, 1.11)44 |
|||
| Cardiovascular disease | ||||
| HF or CAD | Single study | HF or CAD: population attributable fraction 5.4%5 | ↑ Risk | N/A |
| HF | Single studies | HF: ↑ risk but population attributable fraction ↑ d over time 7.8%–1.4%26 Bidirectional relation between AF and HF45 |
↑ Risk | N/A |
| MA | History of HF: ↑ risk (HR, 2.02)2 | |||
| MR | Genetically predicted HF: ↑ risk (OR, 1.86)46 | |||
| CAD | Single study | MI: Population attributable fraction 3.6%26 | ↑ Risk | N/A |
| MA | History of MI: HR, 1.642 | |||
| MR | Genetically predicted CAD: OR, 1.1833 | |||
| VHD | Single studies | Significant heart murmur: ↑ risk (HR, 2.38)47 Significant heart murmur (any diastolic and grade ≥3/6 systolic murmur): ↑ risk, population attributable fraction 21.9% ↓ d over time to 3.1%26 |
↑ Risk | N/A |
| MR | Genetically predicted risk of AF in individuals of European ancestry: associated with VHD with rheumatic fever (OR, 1.26) and nonrheumatic VHD (OR, 1.27)48 | |||
| Cardiac surgery | Single study | Multicenter validated risk prediction model: ↑ risk AF after CABG49 | ↑ Risk | Prophylactic amiodarone, beta blockers: ↓↔ postop AF50–54 Posterior left pericardiotomy during CABG, aortic valve, ascending aortic aneurysm surgery: ↓ postop AF55,56 |
| SR/MA | Postop AF incidence: 23.7%–25.5%56 of cardiac surgery patients57 | |||
| Other conditions | ||||
| CKD | SR/MA | CKD: ↑ risk (HR, 1.47)58 | ↑↔ Risk | N/A |
| MR | Bidirectional relation between CKD and AF59 AF causal for CKD; CKD not causal for AF60 |
|||
| Obstructive sleep apnea | SR/MA | OSA: ↑ risk (OR, 1.71), with potential dose response relation by severity61 | ↑ Risk | Observational studies of SDB treatment: ↓ AF burden62–67 Small RCTs of SDB treatment: ↔68–70 |
| MR | Genetically predicted OSA: ↑ risk (OR, 1.21)71 | |||
| Thyroid disease | SR/MA | Clinical hyperthyroidism: ↑ risk (RR, 2.35)72 | ↑ Risk | |
| MR | Hyperthyroidism: ↑ risk (OR, 1.31)73 | |||
| Sepsis | Single study | Severe sepsis: ↑ risk (OR, 6.82)74; Medicare population75 | ↑ Risk | N/A |
| SR/MA | Sepsis severity: ↑ risk76 | |||
| Markers on ECG | ||||
| PR interval | SR/MA | Prolonged PR: ↑ risk (RR, 1.45)77 | Prolonged PR: ↓ risk PR interval polygenic risk score: ↓ risk PR interval risk SNPs: variable ↑↓ risk |
N/A |
| MR | Polygenic risk score PR interval prolongation: ↓ AF risk (OR, 0.95; P=4.30×10−8) with some variants associated with ↑ and some with ↓ AF risk78 | |||
| LVH | Single study | ECG LVH: Population attributable fraction 10.4% ↓ d over time to 1.8%26 | ↑ Risk | N/A |
| SR/MA | LVH: ↑ risk (RR, 1.46)79 | |||
| Biomarkers | ||||
| Natriuretic peptides | MA | BNP: ↑ risk (HR per 1-SD ln-BNP 1.66)80 | ↑↔ Risk | N/A |
| MR | Natriuretic peptides not associated81 | |||
| Inflammatory markers | SR/MA | CRP: ↑ risk (SMD, 0.95)82 IL-6: ↑ risk (SMD, 0.89)82 TNF-α: ↑ risk (SMD, 2.20)82 |
CRP IL-6, TNF-α, DUSP13, FKBP7, Spondin-1: ↑ risk IL-6R, TNFS12: ↓ risk |
N/A |
| MR | DUSP13, FKBP7, Spondin-1 ↑ risk33 IL-6R, TNFS12 ↓ risk33 |
|||
| Lp(a) | SR/MA | Lp(a): HR, 1.03; only 39% of Lp(a) risk mediated via ASCVD83 | ↑ Risk | N/A |
| MR | Genetically predicted ↑ Lp(a): ↑ risk (HR per 23 mg/dL genetically predicted ↑ Lp(a), 1.04)83 | |||
| Imaging markers | ||||
| LA size or function | Single studies | LA anterior-posterior dimension: ↑ risk (HR per 5 mm ↑, 1.39)84 End diastolic LA volume (min): ↑ risk (HR, 1.12)85 LA emptying fraction: ↑ risk (HR, 1.03)85 |
↑ LA size, emptying fraction: ↑ risk | Surgical LA reduction in conjunction with cardiac surgery or surgical AF ablation in patients with persistent AF may ↑ rates of sinus rhythm86–89 |
| MR | Genetic susceptibility to AF (independent measure) is associated with ↑indexed LA size and ↓ LA ejection fraction (dependent measures)90 | |||
| LV wall thickness | Single study | LV posterior wall thickness: ↑ risk (HR per 4-mm ↑, 1.28)84 | ↑ Risk | N/A |
| SR/MA | LVH: ↑ risk (RR, 1.46)79 | |||
| Social determinants of health | ||||
| Education | Single studies | Higher education: ↑ lifetime risk of AF (US-based ARIC study)91 Higher education in young individuals: ↓ risk of AF diagnosis (Danish study)92 |
Variable ↑↓ risk | N/A |
| MR | AF risk related but largely mediated via BMI (57.5%), type 2 diabetes (9.8%), SBP (18.7%), and smoking (7.1%)93 | |||
| Income | Single studies | Higher income: ↑ lifetime risk of AF (US-based ARIC study)91 Higher income in young individuals: ↓ risk of AF diagnosis (Danish study)92 |
Variable ↑↓ risk | N/A |
| SES | Single studies | Cumulative socioeconomic disadvantage: ↑ risk (HR, 1.57)94 Individual’s poorest areas: 12% ↑ d risk95 |
Low SES: ↑↔ risk | N/A |
| SR/MA | Heterogeneous results96 | |||
| Genetics | ||||
| Family history/heritability | Single studies | Family history of AF: ↑ risk97–99 | ↑ Risk | N/A |
| MR | Proportion heritability explained by loci in European ancestry analysis, 42%100 | |||
| GWAS | MA | Number of AF risk loci ↑s with ↑ number of subjects studied. In 2018, 97–111 loci explained ~11%–42% of the heritability of AF in individuals of European ancestry100,101 | ↑ Risk | N/A |
Population attributable fraction: the proportional disease incidence in the population that is estimated to be due to the risk factor. Statistically significant associations reported, unless otherwise indicated.
↓ indicates decreased; ↑, increased; ↔ no significant change in risk; AF, atrial fibrillation; ASCVD, atherosclerotic cardiovascular disease; BMI, body mass index; BNP, brain naturiuretic peptide; BP, blood pressure; CABG, coronary artery bypass graft surgery; CAD, coronary artery disease; CI, confidence interval; CKD, chronic kidney disease; DBP, diastolic blood pressure; ECG, electrocardiogram; GWAS, genome-wide association study; HF, heart failure; HR, hazard ratio; LA, left atrial; LRFM, lifestyle and risk factor modification; LV, left ventricular; LVH, left ventricular hypertrophy; MA, meta-analysis; MR, Mendelian randomization; N/A, not available/applicable; OR odds ratio; RR, relative risk; OSA, obstructive sleep apnea; SMD, standardized mean difference; SBP, systolic blood pressure; SES, socioeconomic status; SR, systematic review; and VHD, valvular heart disease.