The global burden of atrial fibrillation and atrial flutter in older adults (≥ 65 years): trends, disparities, and drivers (1990–2021)

Abstract

Background

The global burden of atrial fibrillation (AF) and atrial flutter (AFL), major contributors to stroke and heart failure, has substantially increased since 1990 due to aging populations and improved chronic disease survival. This study analyzed global, regional, and national trends in AF/AFL prevalence, incidence, mortality, and disability-adjusted life years (DALYs) among those aged ≥ 65 years (1990–2021), stratified by age, sex, and socioeconomic development index (SDI).

Methods

The Global Burden of Disease (GBD) 2021 dataset provided data on AF/AFL. The analysis included age standardization, average annual percentage change (AAPC) calculations, frontier analysis, and decomposition analysis to quantify the contributions of demographic and epidemiological changes.

Results

Global AF/AFL cases in those aged ≥ 65 years increased significantly (1990–2021), with a 203.91% increase in deaths. However, age-standardized rates varied; high-income North America showed the highest prevalence rate in 2021 at 11,815.10 per 100,000. Decomposition analysis showed that the increase was attributable to population aging and growth. Frontier analysis revealed countries with disproportionately high DALYs relative to their SDI.

Conclusions

The escalating global burden of AF/AFL in older adults demands targeted interventions and resource allocation to address regional disparities. While demographic changes are primary drivers, further research is crucial to understand the contribution of epidemiological factors and develop effective prevention and management strategies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12872-025-05284-5.

Introduction

Atrial fibrillation (AF) and atrial flutter (AFL) are common intractable arrhythmias in clinical practice and serve as significant triggers for ischemic stroke and heart failure [1, 2]. Over the past two decades, the disease burden related to AF/AFL has notably increased due to population aging, lifestyle changes, prolonged survival of patients with chronic illnesses, and the accumulation of risk factors [3, 4]. This rise has led to a distinct clustering trend among the elderly. Recent research indicates an enhanced life expectancy for a growing population of older adults with AF/AFL, alongside advancements in the management of AF/AFL and its complications [5, 6]. However, guidelines tailored to the clinical practice and management of AF/AFL in older adults remain inadequate and require further refinement.

The diagnosis of paroxysmal AF relies heavily on capturing AF episodes through electrocardiogram (ECG), resulting in a low detection rate from a single ECG [7, 8]. Additionally, many AF patients exhibit non-specific or mild symptoms, or they may lack awareness of the disease, contributing to low disease awareness among this population [9–11]. Research shows a 33% increase in the prevalence of AF/AFL over the last two decades [12]. With the global population aging and improved survival rates from chronic illnesses, the incidence and prevalence of AF/AFL are projected to rise over the next 30 years. This presents an opportunity to potentially save millions of lives in the coming decade. Current evidence indicates that the prevalence of AF/AFL among older adults significantly increases with rising life expectancy [5]. Therefore, understanding both the epidemiology and clinical complexities, such as comorbidities and frailty, in older AF/AFL patients (aged ≥ 65 years) is crucial for effective management.

We analyzed the global burden of disease data from 1990 to 2021 to delineate the prevalence, morbidity, mortality, and disability-adjusted life-years (DALYs) status of AF/AFL among older adults aged ≥ 65 years on a global, regional, and national scale. This analysis was stratified by social development level, age, and sex to identify key demographic and epidemiological drivers of the changing burden and their relationship to national health and economic indicators. Specifically, we aimed to determine the trends in AF/AFL burden globally and regionally, identify disparities based on socioeconomic status, and quantify the contributions of demographic changes and epidemiological shifts to the observed trends.

Methods

Study population and data collection

The Global Burden of Disease, Injury, and Risk Factor Study (GBD) is one of the most comprehensive scientific investigations evaluating global health levels and trends. It provides evaluations of 371 diseases, injuries, and disabilities, along with 88 risk factors, across 204 nations and territories [13–15]. In this study, we extracted data on the prevalence, morbidity, mortality, and DALYs of AF/AFL, including their respective 95% uncertainty intervals (UIs), stratified by sex, age, region, and country, from the GBD 2021 dataset for individuals aged ≥ 65 years (https://vizhub.healthdata.org/gbd-results/). The study population consists of individuals aged over 65 with AF/AFL, and specific methodological details can be found in earlier studies.

We utilized DisMod-MR 2.1 and spatiotemporal Gaussian process regression modeling to generate initial estimates of prevalence, incidence, and exposure to risk factors, adjusting for variations in measurement methods and case definitions. We employed ICD-coded vital registration data or household mortality surveys to estimate mortality rates, applying suitable statistical methods to enhance comparability [16]. DALYs serve as a metric for the burden of disease, comprising two components: years of life lost (YLLs) and years lived with disability (YLDs) [17]. Comorbidity adjustments were implemented using microsimulation of 40,000 individuals across different age groups, sexes, countries, and years, based on independent probabilities of prevalence and exposure to conditions such as heart failure, stroke, and diabetes. This standard GBD approach allows consistency across causes and locations, although it does not fully capture the interactive effects of multiple comorbidities on AF/AFL outcomes. The potential implications of this limitation are further discussed in the Discussion.

For this study, AF/AFL data were gathered from 21 districts in geographically proximate and epidemiologically similar countries, encompassing seven age groups (65–69, 70–74, 75–79, 80–84, 85–89, 90–94, ≥ 95 years) for both males and females. Additionally, we computed the socio-demographic index (SDI) for each country, serving as a composite indicator of the social and economic circumstances reflecting health outcomes [15]. These indices can be classified into five categories: low, low-middle, middle, high-middle, and high. Age standardization was conducted using the direct method based on the global age structure in 2021, and the 95% UIs were calculated using the ‘ageadjust.direct’ function in the R package ‘epitools‘ [18].

Data processing and disease model

A descriptive analysis was performed to assess the burden of AF/AFL in individuals aged 65 years and above on a global scale. We compared the age-standardized prevalence, incidence, mortality, and DALYs for AF/AFL across various age groups, genders, regions, and countries, expressed per 100,000 population. Additionally, we estimated the average annual percentage changes (AAPCs) using link-point regression models to assess temporal trends [19].

We also employed Das Gupta’s decomposition method to analyze AF/AFL prevalence, incidence, mortality, and DALYs in the context of aging, population growth, and epidemiological shifts [20–22]. The formulas are as follows:

Prevalence _{ay, py, ey} =

Here, Ay, py, and ey represent prevalence rates corresponding to aging, population growth, and year-specific prevalence, respectively. Additionally, a.i, y denotes the proportion of the population in a specific age group for year y. py signifies the total population in year y, and ei,y signifies the prevalence rate within a particular age group in year y. The prevalence rate reflects the distribution across the seven age groups. The impact of each factor on the change in prevalence from 1990 to 2021 is assessed by altering one factor while holding the others constant, with gender further elucidated by breaking it down into subgroups.

We utilized the ‘geom_smooth’ function from the ‘ggplot2’ package to model the correlation between the SDI and disease burden among individuals aged 65 years and older across 204 countries and territories. Subsequently, we employed frontier analysis, a quantitative analytical approach, to establish the minimum achievable age-standardized prevalence, morbidity, DALYs, and mortality rates based on the SDI [23]. This analysis aimed to evaluate the association between the burden of AF/AFL and socio-demographic development.

In studying health inequalities, the Slope Index and the Concentration Index are commonly used methods to quantify and visualize health disparities between various groups [24]. The Slope Index quantifies the magnitude of health inequalities by categorizing populations based on socio-economic status, computing the average health status of each group, and determining the slope of the regression line according to the socio-economic hierarchy. In contrast, the Concentration Index captures the distribution and extent of health status inequality across the population by examining the area between the concentration curve and the line of equality. Both approaches not only uncover health disparities but also provide precise quantitative measures of health inequalities among diverse groups.

R version 4.4.0 was used for data analysis, with statistical significance set at a two-sided P-value below 0.05. These analytical approaches and tools were selected to uphold the reliability and scientific validity of the study findings.

Results

Global level

Globally, the prevalence of AF/AFL in individuals aged ≥ 65 years increased substantially between 1990 and 2021, rising from 15,797.76 thousand to 38,959.43 thousand (a 146.61% increase) (Table 1). However, the age-standardized prevalence rate (ASPR) decreased slightly by 0.59%, from 5,285.30 to 5,254.21 per 100,000, with an average annual decline of 0.02% (Table 1). This apparent discrepancy reflects the increasing proportion of the population aged ≥ 65 years; the proportion of elderly patients with AF/AFL among all AF/AFL cases rose from 71% in 1990 to 74% in 2021(Supplementary Figure S1).

Table 1.

Cases and age-standardized rates of AF/AFL among individuals aged ≥ 65 years at the global and regional levels in 2021, and the AAPCs from 1990 to 2021

Characteristics	Prevalence (95% UI)			Incidence (95% UI)			Deaths (95% UI)			DALYs (95% UI)
	Counts (no.×103) (2021)	ASR (2021)	AAPCs	Counts (no.×103) (2021)	ASR (2021)	AAPCs	Counts (no.×103) (2021)	ASR (2021)	AAPCs	Counts (no.×103) (2021)	ASR (2021)	AAPCs
Global	38959.43 (31166.85 to 47108.5)	5254.21 (5252.55 to 5255.86)	−0.02 (−0.04 to 0.00)	2885.69 (2121.35 to 3713.4)	427.09 (424.98 to 429.21)	−0.13 (−0.19 to −0.07)	322.25 (287.05 to 356.76)	47.03 (46.87 to 47.19)	0.12 (0.04 to 0.19)	6703.66 (5700.84 to 7872.55)	930.95 (930.25 to 931.66)	0.01 (−0.03 to 0.06)
Sex
Males	19688.97 (15694.94 to 23491.54)	6083.49 (6080.75 to 6086.24)	−0.04 (−0.05 to −0.02)	1355.76 (983.77 to 1742.3)	426.03 (422.21 to 429.88)	−0.31 (−0.35 to −0.26)	125.28 (114.83 to 137.55)	47.56 (47.29 to 47.83)	0.19 (0.11 to 0.27)	3048.27 (2558.67 to 3554.73)	1007.36 (1006.19 to 1008.52)	0.06 (0.01 to 0.1)
Females	19270.46 (15358.77 to 23393.85)	4605.5 (4603.44 to 4607.56)	−0.07 (−0.09 to −0.05)	1529.93 (1103.78 to 1953.28)	420.13 (417.61 to 422.67)	−0.08 (−0.13 to −0.03)	196.97 (170.38 to 227.97)	46.57 (46.36 to 46.77)	0.06 (−0.04 to 0.16)	3655.39 (3081.33 to 4257.62)	870.17 (869.28 to 871.07)	−0.05 (−0.14 to 0.04)
Socio-demographic index
Low	1286.1 (948.98 to 1621.82)	5282.08 (5212.09 to 5352.97)	0.16 (0.12 to 0.19)	111.33 (72.68 to 146.83)	457.54 (437.01 to 478.99)	0.06 (0.03 to 0.09)	9.45 (6.77 to 12.25)	39.91 (39.01 to 40.82)	0.74 (0.43 to 1.04)	227.84 (171.64 to 278.56)	779.47 (775.93 to 783.03)	0.47 (0.26 to 0.69)
Low-middle	4913 (3616.56 to 6293.32)	6715.8 (6682.51 to 6749.25)	0.18 (0.16 to 0.2)	424.27 (284.94 to 567.87)	550.79 (541.47 to 560.26)	0.04 (0.02 to 0.06)	35.88 (29.92 to 42.19)	43.65 (43.18 to 44.12)	0.94 (0.65 to 1.23)	834.31 (681.98 to 982.81)	871.79 (869.84 to 873.74)	0.61 (0.38 to 0.84)
Middle	10115.23 (7699.21 to 12673.6)	7118.34 (7095.28 to 7141.46)	0.27 (0.2 to 0.33)	802.28 (541.85 to 1084.64)	558.52 (552.11 to 565.01)	0.05 (0.02 to 0.08)	78.66 (67.89 to 90.52)	46 (45.67 to 46.33)	0.08 (−0.08 to 0.24)	1711.37 (1424.12 to 2019.05)	877.37 (876.02 to 878.72)	0.08 (−0.06 to 0.22)
High-middle	8464.6 (6604.19 to 10348.03)	6363.24 (6342.95 to 6383.58)	0.02 (−0.06 to 0.1)	588.7 (402.23 to 771.96)	378.73 (374.23 to 383.29)	−0.08 (−0.11 to −0.05)	76.04 (65.96 to 86.05)	46.56 (46.23 to 46.9)	0.05 (−0.18 to 0.28)	1518.31 (1279.89 to 1770.24)	883.66 (882.25 to 885.08)	−0.06 (−0.24 to 0.12)
High	14140.43 (12040.37 to 16218.04)	7971.38 (7957.69 to 7985.09)	−0.29 (−0.32 to −0.26)	956.31 (774.12 to 1143.76)	422.45 (420.08 to 424.84)	−0.27 (−0.32 to −0.22)	121.87 (104.24 to 139.96)	50.05 (49.76 to 50.33)	−0.04 (−0.09 to 0)	2404.64 (2018.94 to 2792.39)	1075.3 (1073.93 to 1076.67)	−0.04 (−0.1 to 0.03)
Region
Central Asia	251.19 (187.65 to 305.93)	4505.88 (4487.83 to 4523.99)	0.16 (0.14 to 0.18)	17.96 (12.27 to 24.42)	310.92 (289.15 to 334.71)	−0.08 (−0.1 to −0.06)	1.22 (1.1 to 1.34)	24.9 (23.5 to 26.38)	0.76 (0.24 to 1.28)	35.72 (28.4 to 42.46)	661.53 (654.52 to 668.61)	0.44 (0.08 to 0.81)
Central Europe	1296.34 (1020.01 to 1555.41)	6133.74 (6089.36 to 6178.48)	0.32 (0.25 to 0.38)	84.41 (63 to 105.96)	333.74 (325.96 to 341.82)	0.01 (−0.02 to 0.03)	10.66 (9.7 to 11.63)	48.53 (47.61 to 49.47)	−0.36 (−0.52 to −0.21)	228.72 (196.31 to 260.7)	1031.84 (1027.61 to 1036.08)	−0.09 (−0.17 to −0.01)
Eastern Europe	1680.4 (1261.52 to 2093.7)	5484.47 (5446.13 to 5523.11)	0.34 (0.33 to 0.36)	100.54 (67.2 to 132.27)	294.27 (286.5 to 302.32)	0.14 (0.1 to 0.17)	14.81 (13.42 to 16.26)	45.96 (45.21 to 46.72)	0.37 (−0.06 to 0.8)	308.68 (258.05 to 358.29)	944.47 (941.08 to 947.87)	0.4 (0.19 to 0.61)
Australasia	403.28 (308.61 to 504.01)	9379.15 (9277.12 to 9482.3)	0.14 (0.12 to 0.16)	26.52 (18.17 to 35.31)	473.22 (455.5 to 491.86)	−0.36 (−0.41 to −0.31)	4.24 (3.62 to 4.8)	71.58 (69.42 to 73.78)	−0.18 (−0.37 to 0)	75.8 (63.33 to 87.09)	1354.87 (1345.19 to 1364.61)	−0.12 (−0.2 to −0.04)
High-income Asia Pacific	1685.13 (1240.64 to 2126.75)	3338.72 (3333.51 to 3343.93)	−0.58 (−0.62 to −0.53)	107.47 (71.42 to 143.85)	261.3 (257.14 to 265.53)	−0.99 (−1.06 to −0.92)	16.66 (13.87 to 19.61)	26.03 (25.62 to 26.44)	−0.66 (−1 to −0.32)	310.72 (256.54 to 366.44)	566.1 (564.04 to 568.18)	−0.69 (−0.92 to −0.46)
High-income North America	5911.16 (5385.59 to 6451.09)	11815.1 (11781.87 to 11848.41)	0.39 (0.38 to 0.41)	434.46 (375.31 to 493.12)	557.52 (553.02 to 562.08)	0.04 (0.01 to 0.08)	37.36 (32.1 to 42.68)	54.63 (54.07 to 55.19)	0.86 (0.7 to 1.01)	865.45 (740.28 to 993.85)	1319 (1316.21 to 1321.79)	0.62 (0.58 to 0.66)
Western Europe	6985.4 (5637.84 to 8309.21)	8447.56 (8425.96 to 8469.22)	−0.08 (−0.09 to −0.06)	436.48 (333.42 to 559.63)	401.04 (397.98 to 404.14)	−0.42 (−0.45 to −0.38)	70.83 (60.98 to 81.26)	60.08 (59.63 to 60.53)	0.1 (−0.07 to 0.27)	1281.11 (1071.79 to 1481.9)	1213.3 (1211.16 to 1215.45)	0.02 (−0.07 to 0.11)
Andean Latin America	268.75 (196.73 to 339.03)	7440.38 (7329.78 to 7552.57)	0.38 (0.36 to 0.4)	20.54 (13.66 to 27.42)	522.99 (495.35 to 552.17)	−0.03 (−0.04 to −0.02)	1.95 (1.59 to 2.29)	40.55 (38.77 to 42.4)	−0.58 (−1.01 to −0.14)	43.83 (35.83 to 51.6)	896.15 (887.78 to 904.59)	−0.23 (−0.47 to 0.02)
Caribbean	273.93 (200.32 to 345.5)	7730.73 (7638.79 to 7823.73)	0.2 (0.18 to 0.22)	20.71 (14.04 to 27.27)	540.11 (517.22 to 564.02)	−0.13 (−0.14 to −0.12)	2.58 (2.28 to 2.89)	51.65 (49.66 to 53.7)	−0.43 (−0.72 to −0.14)	50.43 (41.67 to 58.85)	1037.43 (1028.35 to 1046.57)	−0.23 (−0.38 to −0.08)
Central Latin America	1232.88 (919.22 to 1563.43)	8030.41 (7979.63 to 8081.49)	0.1 (0.07 to 0.13)	95.78 (64.85 to 128.34)	578.84 (565.84 to 592.13)	−0.06 (−0.08 to −0.04)	9.68 (8.51 to 10.85)	48.48 (47.52 to 49.46)	−0.25 (−0.55 to 0.06)	208.52 (170.56 to 244.54)	1023.64 (1019.24 to 1028.06)	−0.09 (−0.24 to 0.07)
Southern Latin America	263.66 (211.06 to 315.16)	4123.79 (4062.15 to 4186.34)	−0.75 (−0.84 to −0.66)	19.13 (13.62 to 24.51)	259.25 (245.5 to 273.84)	−1.21 (−1.3 to −1.13)	3.18 (2.82 to 3.55)	38.11 (36.79 to 39.45)	0 (−0.33 to 0.33)	57 (49.47 to 64.76)	689.52 (683.87 to 695.21)	−0.55 (−1.1 to 0)
Tropical Latin America	1382.51 (1017.39 to 1727.1)	8392.58 (8340.55 to 8444.92)	−0.02 (−0.05 to 0)	102.8 (68.17 to 136.92)	620.76 (606.8 to 635.03)	−0.1 (−0.16 to −0.05)	10.8 (9.36 to 12.19)	51.66 (50.69 to 52.65)	−0.1 (−0.33 to 0.13)	228.96 (190.33 to 269.58)	1071.78 (1067.39 to 1076.19)	−0.06 (−0.17 to 0.05)
North Africa and Middle East	988.96 (751.13 to 1222.22)	4810.06 (4759.93 to 4860.68)	0.32 (0.28 to 0.36)	91.55 (64.83 to 119.63)	360.7 (348.19 to 373.67)	−0.13 (−0.16 to −0.1)	10.33 (8.83 to 11.76)	42.42 (41.58 to 43.27)	0.47 (0.19 to 0.74)	202.04 (169.49 to 233.94)	729.98 (726.69 to 733.28)	0.31 (0.16 to 0.45)
South Asia	5067.72 (3662.18 to 6498.47)	6624.15 (6585.69 to 6662.83)	0.07 (0.05 to 0.09)	453.12 (297.55 to 599.34)	566.69 (555.66 to 577.93)	0.06 (0.05 to 0.07)	32.94 (25.07 to 41.15)	39.32 (38.86 to 39.77)	1.53 (1.02 to 2.05)	818.27 (630.27 to 987.21)	823.03 (821.14 to 824.93)	0.8 (0.52 to 1.08)
East Asia	7611.6 (5678.87 to 9389.46)	4130.91 (4127.88 to 4133.94)	0.29 (0.13 to 0.46)	574.51 (394.08 to 768.17)	566.74 (558.38 to 575.21)	−0.31 (−0.44 to −0.18)	64.82 (52.62 to 77.38)	47.09 (46.71 to 47.47)	−0.43 (−0.73 to −0.12)	1334.49 (1094.2 to 1555.47)	818.62 (817.17 to 820.08)	−0.21 (−0.47 to 0.05)
Oceania	19.84 (14.68 to 24.53)	4813.23 (4740.58 to 4887.01)	0.12 (0.11 to 0.14)	1.66 (1.10 to 2.21)	574.69 (417.57 to 784.41)	−0.16 (−0.17 to −0.15)	0.13 (0.1 to 0.17)	42.71 (35.18 to 51.63)	−0.17 (−0.29 to −0.05)	3.47 (2.75 to 4.17)	899.76 (867.18 to 933.54)	−0.03 (−0.09 to 0.03)
Southeast Asia	2524.51 (1873.02 to 3205.04)	7917.46 (7867.36 to 7967.86)	0.20 (0.19 to 0.21)	206.55 (138.7 to 289.53)	441.20 (439.24 to 443.16)	0.00 (−0.01 to 0.02)	20.72 (17.51 to 23.9)	57.01 (56.22 to 57.81)	0.72 (0.64 to 0.8)	442.28 (365.69 to 513)	1058.02 (1054.81 to 1061.24)	0.43 (0.39 to 0.48)
Central Sub-Saharan Africa	105.64 (81.24 to 132.31)	4790.26 (4560.22 to 5031.51)	0.06 (0.06 to 0.07)	9.19 (6.06 to 12.51)	420.83 (353.83 to 499.75)	−0.04 (−0.05 to −0.03)	1.08 (0.67 to 1.48)	49.04 (45.8 to 52.54)	0.45 (0.37 to 0.54)	24.14 (17.57 to 30.78)	898.4 (885.71 to 911.32)	0.28 (0.24 to 0.33)
Eastern Sub-Saharan Africa	380.88 (281.95 to 470.7)	4819.33 (4710.52 to 4930.58)	0.32 (0.28 to 0.36)	30.01 (20.07 to 39.96)	409.16 (376.9 to 444.01)	0.19 (0.15 to 0.23)	2.74 (1.62 to 3.87)	34.62 (33.22 to 36.08)	0.15 (0.06 to 0.23)	67.27 (49.28 to 87.19)	701.9 (696.15 to 707.69)	0.18 (0.11 to 0.25)
Southern Sub-Saharan Africa	175.98 (125.16 to 221.87)	5961.43 (5743.7 to 6187.05)	0 (−0.01 to 0.01)	14.9 (9.38 to 19.62)	513.03 (450.11 to 584.22)	−0.01 (−0.02 to 0)	1.24 (1.09 to 1.38)	42.2 (39.66 to 44.93)	1.2 (0.87 to 1.53)	29.05 (24.14 to 34)	810.23 (800.14 to 820.49)	0.68 (0.48 to 0.88)
Western Sub Saharan Africa	449.66 (333.4 to 567.56)	4991.99 (4902.44 to 5083.12)	0.39 (0.38 to 0.39)	37.41 (24.33 to 50.75)	424.39 (398.29 to 452.11)	0.26 (0.25 to 0.27)	4.28 (3.62 to 4.95)	48.96 (47.42 to 50.55)	−0.02 (−0.11 to 0.06)	87.71 (71.85 to 104.06)	830.19 (824.34 to 836.08)	0.07 (0.01 to 0.13)

AAPC Average annual percent change, AF/AFL Atrial fibrillation and atrial flutter, ASR Age-standardized rate, DALYs Disability-adjusted life-years, UI Uncertainty interval

Globally, the number of incident AF/AFL cases increased from 1,231.97 thousand in 1990 to 2,885.69 thousand in 2021 (a 134.23% increase) (Table 1). Despite this rise in case numbers, the age-standardized incidence rate (ASIR) decreased from 446.09 to 427.09 per 100,000 population due to demographic changes (Table 1). In summary, the number of incident AF/AFL cases increased, while the age-standardized incidence rate showed a decreasing trend (AAPC = −0.13%; 95% CI = −0.19%, −0.07%) (Table 1).

AF/AFL mortality increased significantly from 106.03 thousand cases in 1990 to 322.25 thousand in 2021 (a 203.91% increase) (Table 1). Consistently, the age-standardized death rate (ASDR) also rose, from 45.56 to 47.03 per 100,000 population (Table 1). Overall, AF/AFL mortality showed an increasing trend (AAPC = 0.11%; 95% CI = 0.04%, 0.19%) (Table 1).

The number of AF/AFL DALYs increased from 2,555.22 thousand in 1990 to 6,703.66 thousand in 2021 (Table 1). The age-standardized rate (ASR) for AF/AFL DALYs in 2021 was 930.95 per 100,000 globally (Table 1). However, the trend in DALYs among those aged 65 years and older was not statistically significant compared to morbidity and mortality rates over the same period.

Regional and national levels

From 1990 to 2021, the ASPR of AF/AFL in older people increased most rapidly in high-income North America (AAPC 0.39%) and decreased most rapidly in Southern Latin America (AAPC − 0.75%) (Table 1). In 2021, high-income North America (11,815.10 per 100,000), Australasia (9,379.15 per 100,000), and Western Europe (8447.56 per 100,000) had the highest ASPR among those aged 65 years and older. Tropical Latin America had the highest ASIR at 620.76 per 100,000 (Fig. 1E-F; Table 1). While regions like the High-income Asia Pacific experienced varying degrees of decline in AF/AFL deaths and DALYs, South Asia saw substantial increases in both (Fig. 1G-H; Table 1). In 2021, Australiasia had the highest rates of AF/AFL deaths and DALYs among older people, while Central Asia and High-income Asia Pacific had the lowest (Fig. 1G-H; Table 1).

Fig. 1.

Distribution of AF/AFL disease burden globally, in territories with varying SDIs and across 21 GBD regions in 1990 and 2021. (A) Prevalent cases, (B) Incident cases, (C) Death cases, (D) DALYs, (E) Age-standardized prevalence rates, (F) Age-standardized incidence rates, (G) Age-standardized deaths rates, (H) Age-standardized DALY rates. AF/AFL Atrial fibrillation and atrial flutter, DALYs Disability-adjusted life-years, SDI Socio-demographic indices

At the national level, Sweden had the highest ASPR of AF/AFL in 2021 (13,521.89), while China had the highest number of cases (7,302.19 thousand) (Supplementary Fig. S2A, Supplementary Table S1). Between 1990 and 2021, Austria showed the largest increase in ASPR among those aged ≥ 65 years (AAPC 2.24%), followed by Sweden (AAPC 1.53%) and Czechia (AAPC 1.43%) (Fig. 2A). Seven countries (Qatar, Jordan, Kuwait, United Arab Emirates, Djibouti, Bahrain, and Guam) experienced prevalence increases exceeding 400% (Fig. 3A). In 2021, China and the United States had the highest morbidity, with 550.83 thousand and 392.14 thousand cases, respectively (Supplementary Fig. S2B, Supplementary Table S1). Qatar (718.89%) and Jordan (536.89%) showed the most substantial increases in morbidity since 1990, while Niue showed a decrease (−14.13%) (Fig. 3B). China (62.03 thousand) and Montenegro (185.71) had the highest number of deaths and the highest ASDR, respectively (Supplementary Fig. S2C, Supplementary Table S1). Most countries experienced increases in deaths, with the largest increases observed in Maldives (684.34%), Kuwait (665.53%), and Bhutan (642.59%) (Fig. 3C). Austria and Sweden showed the most rapid increases in age-standardized morbidity and mortality rates, respectively (Fig. 2B-C). Guam experienced the most substantial reduction in age-standardized DALYs (AAPC − 2.12%), followed by Cyprus (AAPC − 1.59%), while Sweden showed the most significant increase (AAPC 1.54%) (Fig. 2D). Montenegro had the highest rate of AF/AFL DALYs among the elderly, while China had the highest number of DALYs in 2021 (2,577.77 and 1,279.89 thousand, respectively) (Supplementary Fig. S2D, Supplementary Table S1). DALYs in Kuwait increased by over 500% between 1990 and 2021 (Fig. 3D).

Fig. 2.

Map showing AAPCs in AF/AFL prevalence, incidence, deaths, and DALYs among individuals aged ≥ 65 years (1990–2021). (A) Prevalence; (B) Incidence; (C) Deaths; (D) DALYs. AAPC Average annual percentage change, AF/AFL  Atrial fibrillation and atrial flutter, DALYs Disability-adjusted life years

Fig. 3.

Map showing percentage changes in AF/AFL prevalence, incidence, deaths, and DALYs among individuals aged ≥ 65 years (1990–2021). (A) Prevalence; (B) Incidence; (C) Deaths; (D) DALYs. AF/AFL Atrial fibrillation and atrial flutter, DALYs  Disability-adjusted life-years

Age and sex patterns

Globally, the ASPR of AF/AFL decreased among individuals aged ≥ 65 years between 1990 and 2021 (from 6149.18 to 6083.49 for men and from 4693.46 to 4605.5 for women) (Fig. 4J; Table 1). The decline was more gradual in men (AAPC − 0.04%) than in women (AAPC − 0.07%) (Table 1). Similarly, the ASIR decreased in both sexes, with a less pronounced decrease in women (AAPC − 0.31% vs. −0.08%) (Fig. 4J; Table 1). However, mortality rates increased for both men and women, more rapidly in men (AAPC 0.19% vs. 0.06%) (Fig. 4J; Table 1). The ASR of DALYs increased for men but decreased for women. YLLs and YLDs followed similar trends to mortality and DALYs (Fig. 4J; Table 1). In addition, the number of morbidities, deaths, and DALYs in females exceeded those in males (Fig. 4I; Table 1). Despite variations in socio-demographic factors, males consistently showed higher standardized rates than females, suggesting a potentially greater disease burden for males, particularly in countries with low to middle SDI.

Fig. 4.

Age- and sex-structured analysis of AF/AFL disease burden in 1990 and 2021. (A) Prevalent cases, (B) Incident cases, (C) Deaths cases, (D) DALYs, (E) Age-standardized prevalence rates, (F) Age-standardized incidence rates, (G) Age-standardized death rates, (H) Age-standardized DALY rates, (I) Cases, (J) Age-standardized rates. AF/AFL Atrial fibrillation and atrial flutter, DALYs Disability-adjusted life years, YLDs Years lived with disability, YLLs  Years of life lost

Globally, the prevalence, morbidity, mortality, and DALYs of AF/AFL at least doubled in each age subgroup of individuals aged ≥ 65 years between 1990 and 2021 (Fig. 4A-D). This trend showed a clear increase with age, with the largest increase (at least four-fold) observed in the 95 + years age subgroup (Fig. 4A-D). The overall trends in ASPR, ASIR, ASDR, and ASR of DALYs for older adults were not quite significant (Fig. 4E-H). In 2021, the mortality rate per 100,000 in the ≥ 65 years age group increased with age (from 4.5 in the 65–69 years group to 596 in the 95 + years group), with DALYs showing a similar pattern (Fig. 4G-H). The decline in morbidity was consistently greater in males than in females across age subgroups. Conversely, between ages 70 and 95 years, the increase in male mortality consistently exceeded that in females (Supplementary Table S2). These epidemiological trends are complicated by clinical factors like comorbidities and frailty, discussed further in the discussion.

Drivers of AF/AFL epidemiology: ageing, population growth and epidemiological change

To assess the impact of aging, population growth, and epidemiological changes on AF/AFL epidemiology, we performed a decomposition analysis of prevalence, morbidity, mortality, and DALYs. Globally, AF/AFL prevalence, mortality, and DALYs increased significantly, particularly in the high SDI quintile (Fig. 5A and C-D). Incidence also increased significantly, most notably in the middle SDI quintile (Fig. 5B). Population aging accounted for a 5.06% increase in the burden of AF/AFL prevalence between 1990 and 2021, while population growth contributed 95.17% (Table 2). The middle-high SDI quintile drove the majority of the increase in AF/AFL prevalence (99.26%), while population growth contributed the least in middle SDI countries (88.18%) (Table 2).

Fig. 5.

Analysis of changes in AF/AFL prevalence, incidence, deaths, and DALYs globally and across SDI quintiles (1990–2021), based on population aging, population growth, and epidemiological change. (A) Prevalence; (B) Incidence; (C) Deaths; (D) DALYs. AF/AFL Atrial fibrillation and atrial flutter, DALYs  Disability-adjusted life-years, SDI  Socio-demographic index

Table 2.

Decomposition analysis of age-standardized rates of prevalence, morbidity, mortality, and DALYs by SDI quintile

Determinants	Prevalence			Incidence			Deaths			DALYs
	Ageing	Population	EC	Ageing	Population	EC	Ageing	Population	EC	Ageing	Population	EC
Global	1171981.73 (5.06%)	22041808.98 (95.17%)	−52124.98 (−0.23%)	33124.83 (2%)	1674554.67 (101.26%)	−53958.97 (−3.26%)	45964.41 (21.26%)	164979.71 (76.3%)	5267.88 (2.44%)	466398.26 (11.24%)	3676919.79 (88.63%)	5120.79 (0.12%)
Low SDI	34582.12 (4.51%)	690833.59 (90.19%)	40582.11 (5.3%)	2177.38 (3.33%)	60375.89 (92.27%)	2877.48 (4.4%)	712.34 (11.13%)	4560.9 (71.23%)	1129.76 (17.64%)	9156.13 (6.37%)	117079.98 (81.48%)	17459.73 (12.15%)
Low-middle SDI	119894.97 (3.81%)	2899481.37 (92.25%)	123693.95 (3.94%)	7372.47 (2.75%)	253019.5 (94.35%)	7779.53 (2.9%)	2666.93 (9.99%)	18113.79 (67.85%)	5915.82 (22.16%)	32329.22 (5.61%)	458236.32 (79.58%)	85222.55 (14.8%)
Middle SDI	401574.61 (5.63%)	6288434.36 (88.18%)	441173.45 (6.19%)	26421.89 (4.8%)	513901.68 (93.36%)	10153.4 (1.84%)	12665.76 (21.52%)	45474.63 (77.26%)	717.28 (1.22%)	140451.81 (11.49%)	1047310.91 (85.65%)	35079.74 (2.87%)
High-middle SDI	245891.18 (5.27%)	4628761.61 (99.26%)	−211165.37 (−4.53%)	5825.74 (1.89%)	332106.67 (107.99%)	−30407.44 (−9.89%)	13245.73 (26.48%)	36516.02 (73%)	257.06 (0.51%)	127600.75 (14.31%)	795829.05 (89.27%)	−31985.13 (−3.59%)
High SDI	519301.38 (6.98%)	6704867.44 (90.14%)	213770.55 (2.87%)	−1087.64 (−0.24%)	474002.51 (102.87%)	−12123.72 (−2.63%)	22343.31 (30.18%)	52848.78 (71.37%)	−1148.07 (−1.55%)	218,209 (16.64%)	1116598.93 (85.17%)	−23724.73 (−1.81%)

AF/AFL Atrial fibrillation and atrial flutter, DALYs Disability-adjusted life-years, EC Epidemiological change, SDI Socio-demographic index

The incidence of AF/AFL significantly increased globally and across all SDI quintiles over three decades. Globally, population growth and aging increased AF/AFL incidence by 101.26% and 2.00%, respectively, between 1990 and 2021 (Table 2). Population growth in the middle-high SDI quintile was the primary driver. The effect of population aging on incidence was most significant in the middle SDI quintile (4.80%), followed by low SDI (3.33%), middle-low SDI (2.75%), middle-high SDI (1.89%), and least in the high SDI quintile (0.24%) (Table 2).

After adjusting for age and population, global AF/AFL DALYs and mortality rates showed an upward trend. Population growth contributed most to mortality in the middle-high SDI quintile (77.26%) and to DALYs in the same quintile (89.27%) (Table 2). Epidemiological changes (reflecting age- and population-adjusted changes in AF/AFL mortality and DALYs) increased globally, with more pronounced increases in the middle-low, low, middle, and middle-high SDI quintiles, and decreases in the high SDI quintile (Fig. 5; Table 2).

AF/AFL and socio-demographic development

From 1990 to 2021, the overall ASPR of AF/AFL increased with increasing SDI across 204 country regions, with ASIR following a similar trend (Fig. 6A-B). The ASDR increased with increasing SDI before stabilizing at an SDI of approximately 0.80; the age-standardized DALYs rate followed a similar pattern (Fig. 6C-D). These findings, supported by visualizations of health inequities, suggest a higher disease burden in populations with high SDIs (Supplementary Figure S3).

Fig. 6.

Age-standardized rates of AF/AFL across 204 countries and territories, by SDI in 2021. (A) Prevalence; (B) Incidence; (C) Deaths; (D) DALYs. AF/AFL Atrial fibrillation and atrial flutter, DALYs  Disability-adjusted life years, SDI  Socio-demographic index

Frontier analysis examining the correlation between AF/AFL dalys and the level of national development

Using 2021 DALYs and SDI data, we assessed the effective variance between countries and regions and the frontier. As SDI increased, the effective variance tended to decrease, with a relatively small variance overall. The top 15 countries with the worst performance (highest effective variance) included Montenegro, Sweden, Nauru, Austria, Germany, Greenland, Micronesia (Federated States of), New Zealand, Dominica, Iceland, Honduras, Slovakia, Saint Kitts and Nevis, the United States of America, and the Northern Mariana Islands (Fig. 7). These countries had much higher standardized AF/AFL DALY rates than other countries with similar SDI levels. Among countries with an SDI > 0.85, Sweden, Austria, Germany, Iceland, and the United States showed the largest effective differences (Fig. 7). Among countries with an SDI < 0.5, Somalia, Ethiopia, Burundi, Mali, and Niger showed the smallest effective differences (Fig. 7).

Fig. 7.

Frontier analyses of AF/AFL DALYs and SDI among individuals aged ≥ 65 years (1990–2021) across 204 countries and territories. (A) Frontier analysis based on SDI and age-standardized AF/AFL DALY rates. The color scale denotes years, with blue representing 1990 and grey representing 2021. Solid black lines are used to demarcate the frontier; (B) Frontier analysis based on SDI and age-standardized AF/AFL DALY rates for 2021. Countries and regions are represented by dots, with the solid black line denoting the frontier. The top 15 countries with the largest gap to the frontier AF/AFL DALY rate are highlighted in black font. Countries and regions with high SDI values (> 0.85) and significant effect differences relative to their development level are marked in red font, while frontier countries with low SDI values (< 0.5) and minimal effect differences are indicated in blue font. Red dots signify a decrease in age-standardized AF/AFL DALY rates between 1990 and 2021, whereas blue dots indicate an increase. AF/AFL Atrial fibrillation and atrial flutter, DALYs  Disability-adjusted life years, SDI  Socio-demographic index

Discussion

This study comprehensively analyzed global and regional trends in AF/AFL prevalence, incidence, mortality, and DALYs among individuals aged 65 and older from 1990 to 2021, stratified by age, sex, and SDI quintiles. Our findings reveal a substantial increase in the global burden of AF/AFL across all metrics. In 2021, we observed 38,959.43 thousand prevalent cases (146.61% increase), 2,885.69 thousand incident cases (134.23% increase), 930.95 thousand DALYs (162.35% increase), and 6703.66 thousand deaths (203.91% increase).

Geographical location and socioeconomic development significantly influence the AF/AFL burden. High-income North America showed the highest ASPR, while high-income Asia Pacific showed the lowest. This disparity may reflect differences in cardiovascular diseases (e.g., myocardial infarction [25], valvular heart disease [26], hypertension [27]), risk factors (e.g., alcohol consumption [28], diabetes [29, 30]), cultural factors, and access to healthcare [31]. Importantly, we observed a significant rise in ASPR and ASIR in regions with initially low prevalence, including sub-Saharan Africa and Eastern Europe.

Consistent with all-age studies, AF/AFL prevalence among those aged 65 + was positively associated with SDI, likely reflecting differences in risk factor prevalence, healthcare access, and AF awareness [32]. The increasing ASPR in populous Asian and African countries presents a major public health challenge, demanding attention from healthcare providers and policymakers. Strategies in high-SDI regions with stable prevalence should focus on disseminating new technologies and collecting robust data to inform future treatment. In regions experiencing rapid increases in morbidity and limited resources, prioritizing accurate diagnosis, treatment access, and infrastructure development (including basic medical facilities, healthcare worker training, and diagnostic tools) is crucial.

While AF/AFL incidence, prevalence, deaths, and DALYs increased significantly in all age subgroups over 65, a gradual decline in ASDR has been observed in some specific regions, particularly in high-SDI regions. This trend may be attributed to several factors, including the increased use of non-vitamin K antagonist oral anticoagulants (NOACs), improvements in comprehensive AF/AFL management, and technological advancements such as catheter ablation [33, 34]. Furthermore, emerging technologies (artificial intelligence, wearable devices [35–40]) hold promise for improving patient outcomes and reducing the disease burden.

Our study also revealed lower ASIR and ASPR of AF/AFL in women than in men aged 65 and older. This observation likely reflects sex-based differences in traditional risk factors, such as body size and higher rates of smoking and alcohol consumption typically seen in men [41, 42]. However, this unadjusted finding presents an incomplete picture and masks a well-documented paradox in clinical practice. Some studies suggest that after adjusting for factors like height and body size, women may have an intrinsically higher risk of developing AF/AFL, possibly due to differences in cardiac physiology [41]. This complexity extends beyond incidence to clinical outcomes, where the data appear conflicting. While the GBD data report higher mortality in men [32], some clinical studies highlight that women with AF/AFL often experience worse outcomes [43]. For instance, women tend to have a higher burden of comorbidities, are less likely to receive rhythm control therapy, and may suffer from higher rates of heart failure due to delayed diagnosis, as noted in a recent 2023 study [44]. These discrepancies may reflect variations in study design or unadjusted comorbidities. Taken together, these seemingly contradictory findings suggest that the impact of sex on AF/AFL is multifaceted, where the lower incidence rates observed in women in large-scale epidemiological data may be offset by worse clinical outcomes driven by a combination of biological predispositions and disparities in clinical management. Therefore, further investigation into gender-specific risk stratification and treatment strategies is urgently needed to address these disparities [45].

Decomposition analysis showed the middle SDI quintile experienced the most substantial rise in AF/AFL incidence over the past 30 years, potentially due to increasing risk factors, improved healthcare access, and greater awareness in newly industrialized nations. High-SDI regions, while exhibiting some decline in epidemiological trends, still contribute most to overall mortality and DALYs, largely driven by population aging and growth. These findings emphasize the need for comprehensive, region-specific strategies to enhance healthcare services, preventive measures, and treatment approaches to effectively manage the growing global burden of AF/AFL.

Frontier analysis revealed variations in AF/AFL DALY rates across SDI quintiles, with many countries showing values significantly different from the expected benchmark based on SDI. Among high-SDI countries, divergent outcomes are evident, such as lower DALY rates in Sweden compared to the U.S., likely reflecting Sweden’s universal healthcare system that facilitates early AF/AFL diagnosis and management, contrasted with the U.S.’s fragmented healthcare system and higher prevalence of obesity, a key risk factor [46, 47]. Conversely, some low-SDI countries, such as Ethiopia, demonstrate unexpectedly low AF/AFL DALY rates despite limited resources. Possible explanations include strengths such as effective community-based health systems or lower prevalence of Western lifestyle risk factors (e.g., smoking, obesity), but also methodological limitations, including underreporting due to limited diagnostic capacity, a younger demographic structure with fewer elderly at risk, and reduced survival among those with chronic comorbidities, which may mask the true burden [48]. Distinguishing between true policy efficiency and data limitations is therefore crucial before drawing broader inferences. These findings highlight the need for future research to evaluate healthcare system impacts in high-SDI settings and to validate low-SDI outcomes through improved surveillance and longitudinal studies [46–48].

Beyond epidemiological patterns, the clinical complexity of elderly patients with AF/AFL warrants particular consideration. While our analysis quantified the global burden of AF/AFL, prognosis in older adults is strongly influenced by factors not captured in large-scale datasets such as the GBD, including comorbidity burden and geriatric syndromes. For example, ischemic heart disease frequently coexists with AF/AFL and markedly worsens prognosis, particularly in the setting of acute coronary syndromes, where treatment often requires balancing anticoagulation and antiplatelet therapies [49, 50]. Equally critical is the role of frailty, a powerful predictor of adverse outcomes in older adults with cardiovascular disease [51]. Frailty creates a clinical paradox: patients at the highest risk of stroke are often under-prescribed anticoagulation due to concerns about falls and bleeding [52]. Although evidence supports the effectiveness and safety of anticoagulation in frail patients [53], decision-making remains highly individualized and cannot be fully reflected in population-level epidemiological estimates. Taken together, these considerations underscore that while AF/AFL cases continue to rise globally, the true challenge lies in managing the most clinically complex elderly patients. Future research that integrates large administrative datasets with granular clinical and geriatric assessments is essential to guide patient-centered strategies and optimize outcomes.

Despite the continuous updates and valuable health metrics provided by the GBD, several limitations warrant consideration. First, GBD data rely on statistical modeling rather than national clinical records, potentially reducing accuracy in low-resource settings with weak surveillance systems. Regional variations in disease management may further lead to misdiagnosis or underdiagnosis of AF/AFL, affecting data reliability [54]. Second, population heterogeneity in health systems, epidemiological transitions, and diagnostic capacities significantly limits the direct comparability of global AF/AFL burden across regions, particularly due to likely underreporting in low-resource settings such as sub-Saharan Africa [51]. Intra-group analyses within similar SDI categories, as conducted in our frontier analysis, yield more meaningful insights than generalized global comparisons. Third, the lack of differentiation between AF/AFL subtypes (e.g., paroxysmal, persistent, permanent AF; typical vs. atypical flutter) restricts clinical and policy applicability. Finally, attributing mortality and DALYs solely to AF/AFL without adjusting for comorbidities (e.g., heart failure, stroke, diabetes) or geriatric conditions like frailty may overestimate disease burden, potentially masking regional variations in prevalence or outcomes [55, 56]. Future studies with granular clinical data could address these limitations to enhance accuracy and applicability. Additionally, this study did not quantify the economic burden associated with AF/AFL, including healthcare costs and productivity losses. Future studies should address this limitation by estimating the economic impact of AF/AFL to inform resource allocation and policy development across diverse socioeconomic contexts.

This study utilized the GBD database, a leading source of global health data, to analyze AF/AFL trends from 1990 to 2021. The application of decomposition and frontier analyses provided further insights into the influence of demographic trends and facilitated cross-country comparisons.

Conclusions

The global burden of AF/AFL among individuals aged ≥ 65 years has increased substantially from 1990 to 2021, with marked regional disparities. Our findings reveal considerable rises in prevalence, incidence, mortality, and DALYs, particularly in regions that previously had lower prevalence. Population aging and growth are the main drivers, with epidemiological changes also contributing. These results underscore the urgent need for strengthening healthcare infrastructure, enhancing diagnostic capacity, and implementing efficient AF/AFL management strategies tailored to regional contexts. Although our study did not quantify the economic burden associated with AF/AFL, future research should address this important aspect, for example by estimating healthcare costs and productivity losses related to the increasing disease burden.

Abbreviations

AAPC

Average annual percentage change

AF

Atrial fibrillation

AFL

Atrial flutter

ASR

Age-standardized rate

ASDR

Age-standardized death rate

ASIR

Age-standardized incidence rate

ASPR

Age-standardized prevalence rate

DALYs

Disability-adjusted life years

ECG

Electrocardiogram

GBD

Global Burden of Disease Study

NOACs

Non-vitamin K antagonist oral anticoagulants

SDI

Socio-demographic index

UIs

Uncertainty intervals

YLDs

Years lived with disability

YLLs

Years of life lost

Authors’ contributions

Bo Yang and Shao Bo Shi provided ideas; Jing Lin Wang accessed and verified the data; Jing Lin Wang and Yi Ling Li organized the figure and wrote the manuscript; Bo Yang and Shao Bo Shi revised the manuscript. All authors reviewed the manuscript and agreed to publish.

Funding

The study was supported by The Interdisciplinary Innovative Talents Foundation from Renmin Hospital of Wuhan University (JCRCZN-2022-003) and the Fundamental Research Funds for the Central Universities (2042024YXB003).

Data availability

The data underlying this article will be shared upon reasonable request to the corresponding author.

Declarations

Ethics approval and consent to participate

This study did not involve human participants, biological samples, or direct interaction with individuals. Ethical approval and the requirement for informed consent were waived by the Institutional Review Board (IRB) at the University of Washington. This exemption is granted under U.S. federal regulations (45 CFR 46.104(d)(4)) for research involving the secondary analysis of publicly available, de-identified data from the Global Burden of Disease (GBD) Study, a fully anonymized dataset containing no personal identifiers. As this research exclusively utilized non-identifiable aggregated data, the requirements of the Declaration of Helsinki were not applicable. This exemption is explicitly permitted under Paragraph 32 of the Declaration for research involving anonymized public datasets and was formally confirmed by the University of Washington IRB in accordance with U.S. federal regulations (45 CFR 46.104(d)(4)).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.