The Latest Global Stroke Fact Sheet 1990–2023: Global Burden of Disease Study

Abstract

Background

Stroke remains a leading cause of death and long-term disability worldwide despite substantial reductions in age-standardized rates over recent decades. Updated and comprehensive assessments of global stroke burden are essential to inform prevention strategies, health-system planning, and policy responses, particularly in the context of widening inequities between regions and the emergence of new environmental and metabolic risk factors.

Methods

This review summarizes the most recent Global Burden of Disease (GBD) 1990–2023 estimates for stroke, drawing on data from the 2025 GBD cardiovascular disease publication and the Institute for Health Metrics and Evaluation visualization tools. Stroke burden was assessed for ischaemic stroke, intracerebral haemorrhage, and subarachnoid haemorrhage using standard GBD definitions. Estimates of incidence, prevalence, mortality, disability-adjusted life years (DALYs), and population-attributable fractions for 24 modifiable risk factors were analysed globally and across 21 GBD regions.

Results

In 2023, stroke was the second leading cause of death and the third leading cause of death and disability combined worldwide. Globally, there were 13.2 million incident strokes, 104.8 million people living with stroke, and 6.8 million stroke-related deaths, accounting for 156.5 million DALYs. Although age-standardized incidence, mortality, and DALY rates declined substantially since 1990, the absolute numbers of stroke events, deaths, and DALYs increased markedly, driven by population growth and ageing. Over 80% of stroke deaths and DALYs occurred in low- and middle-income countries. Ischaemic stroke accounted for 63.9% of incident cases, while haemorrhagic strokes contributed disproportionately to mortality and disability. Overall, 83% of the global stroke burden was attributable to modifiable risk factors, with high systolic blood pressure the leading contributor. Notably, the contribution of high ambient temperature, obesity, and lead exposure increased substantially between 1990 and 2023.

Conclusions

Stroke remains largely preventable, yet its global burden continues to rise in absolute terms, with pronounced regional disparities. Intensified, region-specific prevention strategies targeting dominant risk factors, alongside equitable access to acute care and rehabilitation, are urgently required to reduce avoidable stroke-related death and disability worldwide.

Introduction

Stroke remains one of the leading causes of death, disability, and loss of healthy life expectancy worldwide, despite substantial advances in prevention and acute care over recent decades [1]. Evidence from successive Global Burden of Disease (GBD) studies, led by the Institute for Health Metrics and Evaluation, has fundamentally reshaped understanding of stroke as a major, largely preventable contributor to global and national health loss. GBD analyses have consistently shown that stroke accounts for tens of millions of disability-adjusted life years (DALYs) annually, driven not only by mortality but also by long-term disability affecting cognition, physical functioning, communication, and social participation across the life course. The most recent published GBD estimates [2] showed that despite declining age-standardized rates, the global burden of stroke continues to rise due to population ageing, population growth, and increasing exposure to metabolic risk factors, with the vast majority of burden attributable to modifiable risks and disproportionately affecting low- and middle-income countries. There were substantial increases in DALYs attributable to high BMI (88.2% [53.4–117.7]), high ambient temperature (72.4% [51.1–179.5]), high fasting plasma glucose (32.1% [26.7–38.1]), diet high in sugar-sweetened beverages (23.4% [12.7–35.7]), low physical activity (11.3% [1.8–34.9]), high systolic blood pressure (6.7% [2.5–11.6]), lead exposure (6.5% [4.5–11.2]), and diet low in omega-6 polyunsaturated fatty acids (5.3% [0.5–10.5]) [3].

Findings from earlier GBD iterations showed that although age-standardized stroke mortality rates declined in many high-income countries during the late 20th and early 21st centuries, the absolute number of people experiencing stroke, living with its consequences, or dying from stroke continued to rise globally [2, 3]. This apparent paradox has been attributed to population growth, population ageing, and the sustained high prevalence of modifiable vascular and lifestyle risk factors, including elevated blood pressure, diabetes, obesity, smoking, unhealthy diet, physical inactivity, and harmful alcohol use [3].

GBD studies have also transformed understanding of stroke as a major cause of long-term disability, not merely a fatal acute event. Non-fatal stroke contributes substantially to YLDs through persistent impairments in mobility, cognition, communication, emotional regulation, and social participation, often affecting individuals for decades. Importantly, GBD evidence has shown that a large proportion of global stroke burden occurs in working-age adults, amplifying its economic and societal consequences [4].

Another consistent finding across GBD cycles is the marked geographic and socio-demographic inequality in stroke burden. Low- and middle-income countries now account for the majority of global stroke incidence, mortality, and DALYs, with disproportionately higher rates of premature stroke and more severe disability compared with high-income regions [1]. These disparities are closely linked to inequitable exposure to modifiable risk factors, limited access to effective prevention and acute care, and broader social and environmental determinants of health [5].

Despite the critical insights provided by earlier GBD analyses, there is a pressing need for updated global estimates of stroke burden. Risk-factor distributions, health-system performance, demographic structures, and social conditions have evolved rapidly in recent years. Rising global prevalence of obesity and diabetes, persistent gaps in hypertension control, widening socioeconomic inequalities may have substantially altered stroke incidence, severity, and outcomes. Earlier estimates may therefore underestimate current non-fatal burden, disability severity, and the true scale of preventable stroke worldwide.

Updated GBD stroke estimates are essential to provide contemporary, comparable evidence for global health planning and priority setting; to quantify the proportion of stroke burden attributable to modifiable risk factors; to track progress towards international non-communicable disease targets; and to inform the design and evaluation of population-level prevention strategies. In this context, updated GBD analyses are a critical foundation for guiding effective, equitable, and sustainable global responses to stroke prevention and brain-health protection.

Methods

The values contained in this Fact Sheet have all been extracted from the most current (2025) GBD publication on stroke burden in the Journal of American College of Cardiology [2] and online GBD Visualization Tool [6]. In GBD, stroke is considered as an aggregate category of ischaemic stroke (IS), intracerebral haemorrhage (ICH), and subarachnoid haemorrhage (SAH). GBD definition of IS includes brain imaging showing blood flow to part of the brain being occluded and according to WHO criteria of rapidly developing clinical signs of disturbance of cerebral function lasting >24 h or leading to death. GBD definition of ICH includes brain imaging showing bleeding into the tissue of the brain and according to WHO criteria of rapidly developing clinical signs of disturbance of cerebral function lasting >24 h or leading to death. For SAH, brain imaging or lumbar puncture indicating the rupture of a blood vessel resulting in bleeding into the subarachnoid space and according to WHO criteria of rapidly developing clinical signs of disturbance of cerebral function lasting >24 h or leading to death. Only non-traumatic events were included (traumatic ICH and SAH are not included in the definition of stroke). Only first-ever-in-a-lifetime strokes are included in the GBD analysis.

Results

The most recent Global Burden of Disease (GBD) 1990–2023 cardiovascular disease (CVD) and stroke burden and risk factors estimates [2] and Institute for Health Metrics and Evaluation GBD Compare/Results Website [6] showed that stroke remains the second leading cause of death (after ischaemic heart diseases) and the third leading cause of death and disability combined (as expressed by disability-adjusted life-years lost – DALYs) in the world. These data showed that although the age-standardized stroke death, DALYs, and incidence rates (per 100,000 people per year) have dropped significantly from 1990 to 2023 {52% (95% uncertainty interval [UI] 45–57), 48% (40–56), and 25% (23–27), respectively}, the number of people who either suffer or die from a stroke or live with a disability after a stroke has risen substantially worldwide between 1990 and 2023: incident strokes – by 73% (95% UI: 68–78), deaths from stroke – by 22% (7–39), and DALYs – by 18% (3–36), with the bulk of the global stroke burden (81.7% of deaths and 85.4% of DALYs) residing in low-income and low-middle-income countries.

In 2023, there were 104.8 (98.3–112.8) million prevalent, 13.2 (12.0–14.8) million incident, and 6.8 (6.1–7.5) million fatal strokes (Tables 1–4). Stroke burden varied markedly across GBD regions (Fig. 1), with the highest age-standardized DALY rate in Oceania (3,206 per 100,000; 95% UI: 2,498–4,050) and the lowest in Australasia (409 per 100,000; 95% UI: 362–451). Globally, stroke caused 3.50 million deaths in males and 3.29 million deaths in females. The number of incident strokes in males (7.2 million) was higher than in females (6.0 million). IS was the predominant subtype, accounting for 63.9% of incident cases, followed by ICH (29.3%) and SCH (6.9%).

Table 1.

GBD 2023 estimates for global incidence, prevalence, mortality, and DALYs for all stroke types combined

Metric of the burden	Number	Crude rate per 100,000 per year (95% UI)a	Age-adjusted rate per 100,000 per year (95% UI)a	Suggested statement for use in stroke awareness/advocacy materials
Incidence
Ages (all), sexes (both)	13,238,400	164.1 (148.5–183.9)	148.1 (134.7–165.6)	There are over 13 million new strokes each year. Globally, one in four people over age 25 will have a stroke in their lifetime
15–49 years	1,895,100	47.3 (40.7–55.6)	–	Each year, about 15% of all strokes occur in people 15–49 years of age
<70 years	6,893,800	91.6 (79.3–104.2)	–	Each year, about 52% of all strokes occur in people under 70 years of age
Men (all ages)	7,222,500	178.4 (160.0–201.7)	174.1 (157.4–195.6)	Each year, about 55% of all strokes occur in men
Women (all ages)	6,015,900	149.7 (135.0–166.3)	125.1 (134.7–165.5)	Each year, about 45% of all strokes occur in women
Prevalence
Ages (all), sexes (both)	104,756,100	1,299.0 (1,218.0–1,398.0)	1,169.8 (1,097.0–1,256.7)	Globally, there are almost 105 million people currently living who have experienced stroke
15–49 years	24,110,500	602.2 (558.0–651.0)	–	23% of people who have experienced a stroke and are currently living are people 15–49 years of age
<70 years	68,669,490	912.0 (852.0–977.0)	–	65% of people who have experienced a stroke and are currently living are under the age of 70
Men (all ages)	56,374,500	1,393.0 (1,305.1–1,494.0)	1,325.1 (1,241.4–1,419.9)	54% of people who have experienced a stroke and are currently living are men. Globally, men account for just over half (54%) of all persons who have experienced a stroke
Women (all ages)	48,381,600	1,204.0 (1.218.0–1,398.0)	1,034.8 (1,097.4–1,256.7)	46% of people who have experienced a stroke and are currently living are women
Deaths
Ages (all), sexes (both)	6,793,200	84.2 (75.2–92.6)	76.0 (67.8–83.6)	Almost seven million people die from stroke annually
15–49 years	410,100	10.2 (8.6–12.0)	–	About 6% of all deaths from stroke occur in people 15–49 years old
<70 years	2,390,000	31.8 (28.2–35.6)	–	35% of all deaths from stroke occur in people under 70 years old
Men (all ages)	3,500,300	86.5 (76.4–98.1)	87.5 (76.9–98.6)	52% of all deaths from stroke are in men
Women (all ages)	3,292,900	82.0 (70.0–96.4)	65.9 (56.4–77.7)	48% of all deaths from stroke are in women
DALYs
Ages (all), sexes (both)	156,519,800	1,941 (1,750–2,136)	1,738.4 (1,565.8–1,918.2)	Over 156 million years of healthy life are lost each year due to stroke-related death and disability
15–49 years	23,821,800	595.0 (496.0–692.5)	–	15% of healthy life lost due to stroke-related death and disability affects people 15–49 years old
<70 years	90,084,400	1,197 (1,048.0–1,343.0)	–	58% of healthy life lost due to stroke-related death and disability affects people under the age of 70 years
Men (all ages)	85,742,200	2,118.0 (1,874.0–2,419.0)	2,030.2 (1,793.9–2,313.1)	Men account for 55% of healthy life lost due to stroke-related disability
Women (all ages)	70,777,600	1,762.0 (1,538.0–2,076.0)	1,472.0 (1,273.6–1,741.8)	Women account for 45% of healthy life lost due to stroke-related disability

The 95th UI is determined by the 25th and 975th value of the 1,000 values after ordering them from smallest to largest. Larger UIs can result from limited data availability, small studies, and conflicting data, while smaller UIs can result from extensive data availability, large studies, and data that are consistent across sources.

^a95% UI represents a range of values that reflects the certainty of an estimate. In the GBD study, every estimate is calculated 1,000 times, each time sampling from distributions rather than point estimates for data inputs, data transformations, and model choice.

Table 4.

GBD 2023 estimates for global incidence, prevalence, mortality, and DALYs for SCH

Metric of the burden	Number	Crude rate per 100,000 per year (95% UI)	Age-adjusted rate per 100,000 per year (95% UI)	Suggested statement for use in stroke awareness/advocacy materials
Incidence
Ages (all), sexes (both)	908,500	11.3 (9.8–12.7)	10.3 (9.1–11.7)	There are almost 700,000 new SCHs each year. Globally, about 6% of all incident strokes are SCH
15–49 years	309,000	7.7 (6.2–9.4)	–	Each year, over 34% of all SCHs occur in people 15–49 years of age
<70 years	659,400	8.8 (7.4–10.1)	–	Each year, over 74% of all SCHs occur in people under 70 years of age
Men (all ages)	393,100	9.7 (8.4–11.0)	9.4 (8.2–10.6)	Each year, about 49% of all SCHs occur in men
Women (all ages)	515,400	12.8 (11.2–14.5)	11.3 (9.9–12.8)	Each year, about 51% of all SCHs occur in women
Prevalence
Ages (all), sexes (both)	10,362,500	128.5 (115.6–142.9)	116.4 (104.8–129.6)	Globally, there are almost 7.9 million people currently living who have experienced SCH
15–49 years	3,540,600	88.4 (79.4–100.7)	–	35% of people who have experienced an SCH and are currently living are people 15–49 years of age
<70 years	8,155,800	108.4 (97.5–121.0)	–	80% of people who have experienced an SCH and are currently living are under the age of 70
Men (all ages)	4,049,900	100.0 (89.7–111.5)	93.9 (84.2–104.9)	45% of people who have experienced an SCH and are currently living are men
Women (all ages)	6,312,600	157.1 (142.0–174.5)	137.8 (124.7–153.3)	55% of people who have experienced an SCH and are currently living are women
				Globally, women account for just over half (56%) of all persons who have experienced an SCH
Deaths
Ages (all), sexes (both)	357,400	4.4 (3.8–5.3)	4.0 (3.4–4.8)	Over 350,000 people die from SCH annually
15–49 years	57,000	1.4 (1.1–1.9)	–	About 16% of all deaths from SCH occur in people 15–49 years old
<70 years	197,000	2.6 (2.1–3.4)	–	55% of all deaths from SCH occur in people under 70 years old
Men (all ages)	177,300	4.4 (3.6–6.5)	4.3 (3.5–5.4)	49% of all deaths from SCH are in men
Women (all ages)	180,100	4.4 (3.6–6.1)	3.7 (3.0–5.1)	51% of all deaths from SCH are in women
DALYs
Ages (all), sexes (both)	11,257,000	139.6 (118.0–171.1)	127.3 (107.1–156.4)	Over 10 million years of healthy life are lost each year due to SCH-related death and disability. Globally, DALYs from SCH constituted approximately 7% of DALYs from all strokes combined
15–49 years	3,410,000	85.2 (67.9–110.2)	–	30% of healthy life lost due to SCH-related death and disability affects people 15–49 years old
<70 years	8,632,000	114.7 (93.5–144.8)	–	76% of healthy life lost due to SCH-related death and disability affects people under the age of 70 years
Men (all ages)	5,638,000	139.3 (113.3–176.6)	132.4 (107.0–168.3)	Men account for 52% of healthy life lost due to SCH-related disability
Women (all ages)	5,619,000	139.9 (112.3–191.4)	122.3 (107.1–156.4)	Women account for 48% of healthy life lost due to SCH-related disability

Fig. 1.

Age-standardized disability-adjusted life years (DALYs) rates per 100,000 people for stroke, 2023, both sexes combined (reprinted from Roth et al. [2] by permission). Map shows global variation in age-standardized DALY rates for stroke in 2023. Rates are expressed per 100,000 population and are age-standardized to allow comparison across countries with different population age structures. Darker colours indicate higher stroke burden. Data are from the Global Burden of Disease (GBD) Study 2023. DALYs represent the sum of years of life lost due to premature mortality and years lived with disability (reprinted from Roth et al. [2], by permission).

Table 2.

GBD 2023 estimates for global incidence, prevalence, mortality, and DALYs for IS

Metric of the burden	Number	Crude rate per 100,000 per year (95% UI)	Age-adjusted rate per 100,000 per year (95% UI)	Suggested statement for use in stroke awareness/advocacy materials
Incidence
Ages (all), sexes (both)	8,454,800	104.8 (91.0–136.4)	94.1 (82.2–108.2)	There are over 8.4 million new ISs each year. Globally, about 64% of all incident strokes are ISs
15–49 years	988,500	24.7 (18.9–31.0)	–	Each year, almost 12% of all ISs occur in people 15–49 years of age
<70 years	4,190,000	55.7 (45.7–67.1)	–	Each year, almost 50% of all ISs occur in people under 70 years of age
Men (all ages)	4,699,300	116.1 (99.5–136.4)	113.0 (97.9–131.2)	Each year, about 56% of all ISs occur in men
Women (all ages)	3,755,500	93.5 (82.5–107.4)	77.3 (68.3–88.7)	Each year, about 44% of all ISs occur in women
Prevalence
Ages (all), sexes (both)	77,778,100	964.3 (885.6–1,048.3)	862.0 (793.1–935.2)	Globally, there are almost 78 million people currently living who have experienced IS
15–49 years	13,721,600	342.7 (306.8–385.7)	–	Almost 18% of people who have experienced an IS and are currently living are people 15–49 years of age
<70 years	46,743,100	621.0 (565.5–679.1)	–	60% of people who have experienced an IS and are currently living are under the age of 70
Men (all ages)	43,374,200	1,071.4 (987.4–1,162.6)	1,021.3 (943.8–1,109.3)	Almost 56% of people who have experienced an IS and are currently living are men
Women (all ages)	34,403,900	856.4 (783.3–937.2)	723.4 (661.7–788.8)	Over 44% of people who have experienced an IS and are currently living are women
Deaths
Ages (all), sexes (both)	3,279,004	40.7 (35.6–45.7)	37.0 (32.4–41.7)	3.3 million people die from IS annually
15-49 years	60,652	1.50 (1.2–2.0)	–	Almost 2% of all deaths from IS occur in people 15–49 years old
<70 years	673,476	9.0 (7.6–10.5)	–	Over 20% of all deaths from IS occur in people under 70 years old
Men (all ages)	1,597,501	39.5 (34.3–45.1)	41.6 (36.0–47.6)	Almost 49% of all deaths from IS are in men
Women (all ages)	1,681,503	41.9 (34.1–49.7)	33.1 (26.9–39.3)	Over 51% of all deaths from IS are in women
DALYs
Ages (all), sexes (both)	67,043,715	831.2 (727.4–933.3)	743.3 (650.0–834.1)	Over 67 million years of healthy life is lost each year due to IS-related death and disability. Globally, DALYs from IS constituted about 43% of DALYs from all strokes combined
15–49 years	4,940,758	123.4 (98.4–152.0)	–	Over 7% of healthy life lost due to IS-related death and disability affects people 15–49 years old
<70 years	27,425,361	364.4 (308.8–417.7)	–	Almost 41% of healthy life lost due to IS-related death and disability affects people under the age of 70 years
Men (all ages)	35,472,001	876.2 (763.8–1,003.4)	860.9 (748.9–981.5)	Men account for 53% of healthy life lost due to IS-related disability
Women (all ages)	31,571,714	785.9 (636.6–925.2)	638.8 (516.8–752.9)	Women account for 47% of healthy life lost due to IS-related disability

Table 3.

GBD 2023 estimates for global incidence, prevalence, mortality, and DALYs for ICH

Metric of the burden	Number	Crude rate per 100,000 per year (95% UI)	Age-adjusted rate per 100,000 per year (95% UI)	Suggested statement for use in stroke awareness/advocacy materials
Incidence
Ages (all), sexes (both)	3,875,100	48.0 (41.5–53.6)	43.6 (37.7–48.5)	There are almost 3.9 million new ICHs each year. Globally, over 29% of all incident strokes are ICHs
15–49 years	597,600	15.0 (12.2–17.7)	–	Each year, over 15% of all ICHs occur in people 15–49 years of age
<70 years	2,044,000	27.2 (22.8–31.3)	–	Each year, almost 53% of all ICHs occur in people under 70 years of age
Men (all ages)	2,130,000	52.6 (45.4–58.8)	51.8 (44.9–57.9)	Each year, about 55% of all ICHs occur in men
Women (all ages)	1,745,000	43.4 (37.6–48.6)	36.5 (31.5–40.8)	Each year, over 45% of all ICHs occur in women
Prevalence
Ages (all), sexes (both)	17,267,000	214.1 (195.6–233.1)	198.6 (182.9–216.2)	Globally, there are over 17 million people currently living who have experienced ICH
15–49 years	6,905,000	172.5 (153.7–193.3)	–	40% of people who have experienced an ICH and are currently living are people 15–49 years of age
<70 years	14,089,000	187.2 (170.5–204.0)	–	82% of people who have experienced an ICH and are currently living are under the age of 70
Men (all ages)	9,304,000	229.8 (209.1–250.9)	218.3 (199.3–237.9)	54% of people who have experienced an ICH and are currently living are men. Globally, men account for just over half (56%) of all persons who have experienced an ICH
Women (all ages)	7,963,000	198.2 (181.7–215.7)	179.6 (165.3–195.9)	46% of people who have experienced an ICH and are currently living are women
Deaths
Ages (all), sexes (both)	3,157,000	39.1 (34.1–44.0)	34.9 (30.4–39.3)	Almost 3.2 million people die from ICH annually
15–49 years	292,000	7.3 (6.0–8.7)	–	Over 9% of all deaths from ICH occur in people 15–49 years old
<70 years	1,520,000	20.2 (17.4–23.0)	–	48% of all deaths from ICH occur in people under 70 years old
Men (all ages)	1,726,000	42.6 (35.9–50.7)	41.6 (35.2–49.5)	55% of all deaths from ICH are in men
Women (all ages)	1,431,000	35.6 (29.4–43.4)	29.1 (24.1–35.6)	45% of all deaths from ICH are in women
DALYs
Ages (all), sexes (both)	78,219,000	969.8 (834.0–1,092.7)	867.8 (745.9–979.8)	Over 78 million years of healthy life are lost each year due to ICH-related death and disability. Globally, DALYs from ICH constituted about 50% of DALYs from all strokes combined
15–49 years	15,471,000	386.4 (317.3–456.6)	–	19% of healthy life lost due to ICH-related death and disability affects people 15–49 years old
<70 years	54,027,000	717.8 (612.2–821.4)	–	69% of healthy life lost due to ICH-related death and disability affects people under the age of 70 years
Men (all ages)	44,632,000	1,102.5 (928.8–1,322.0)	1,036.9 (874.4–1,242.7)	Men account for 57% of healthy life lost due to ICH-related disability
Women (all ages)	33,587,000	836.0 (690.3–1,051.0)	710.9 (582.5–897.7)	Women account for 43% of healthy life lost due to ICH-related disability

The GBD study also showed that stroke is highly preventable, with 83% of the stroke burden in 2023 attributable to 24 modifiable risk factors – presenting a public health challenge and an opportunity for action (Table 5). Whilst from 1990 to 2023 there was a positive trend in the reduction of population-attributable fraction (PAF) of stroke DALYs due to low ambient temperature (−34.1%; −29.3 to −39.2), air pollution (−22.0%; −16.9 to −27.0), smoking (−17.1%; −1.7 to −30.1), and second-hand smoking (−16.2%; −0.7 to −28.4), there was a significant increase in PAF of stroke DALYs due to high ambient temperature (122.0%; 72.3–300.5), high BMI (82.6%; 31.3–173.8), and lead exposure (12.1%; 3.6–25.8). Stroke attributable to metabolic risks constituted 69.2% (95% UI 56.4–79.3) of all strokes, behavioural risks constituted 35.2% (26.9–44.7), and environmental risks constituted 34.8% (26.9–44.8). However, the level of contribution of the 24 stroke risk factors in PAF of age-standardized DALYs differed substantially in 21 GBD regions (Fig. 2). Globally, the 10 most significant stroke risk factors were high SBP (57.9%; 95% UI 41.6–70.5), dietary risks (17.1%; 8.0–29.9), ambient particulate matter pollution (16.4%; 10.1–22.6), high LDL cholesterol (13.1%; 4.9–21.9), smoking (13.0%; 10.9–15.7), kidney dysfunction (10.2%; 6.9–13.7), household air pollution from solid fuels (9.4%; 5.1–14.2), high fasting plasma glucose (9.0%; 6.7–12.4), high BMI (4.4%; 0.5–8.6), and low ambient temperature (4.1%; 3.5–4.9).

Table 5.

2021 GBD estimates for stroke burden (as measured by DALYs) attributable to risk factors, 2023, both sexes

1. Metabolic risks (high SBP, high BMI, high FPG, high total cholesterol, and low glomerular filtration rate) account for 69.2% (56.4–79.3) of stroke burden

2. Behavioural factors (smoking, poor diet, and low physical activity) account for 35.2% (26.9–44.7) of stroke burden and environmental risks (air pollution and lead exposure) 34.8% (26.9 to 44.8)

3. Metabolic risk factors contributed most to the stroke-related DALYs (72.0%; 95% UI 60.2–81.5) in high income World Bank countries, followed by the environmental risk cluster (air pollution and non-optimal ambient temperature) in low-income countries (46.1%; 32.9–57.3), and behavioural risks (37.8%; 29.0–48.3) in upper-middle income countries

4. Globally, 16 individually significant risk factors for stroke are high SBP (57·9% [95% UI 41.6–70.7] attributable DALYs), ambient particular matter (16.4% [10.1–22.5]), high LDL cholesterol (13.1% [4.9–21.9]), smoking (13.0% [10.9–15.7]), kidney dysfunction (10.2% [6.9–13.7]), diet high in sodium (9.8% [1.9–22.3]), household air pollution from solid fuels (9.4% [5.1–14.2]), high fasting glucose level (9.0% [6.7–12.4]), diet low in fruits (6.9% [0.5–12.3]), high BMI (4.4% [0.5–8.6]), second-hand smoking (4.1% [2.9–5.5]), low ambient temperature (4.2% [3.5–4.9]), high alcohol use (2.5% [0.3–5.8]), diet low in vegetables (2.2% [0.6–4.3]), low physical activity (2.0%; 0.5–3.6), and high ambient temperature (1.4%; 0.4–2.9)

5. All risk factors combined account for 83.0% (76.3–88.3) of the global stroke burden

SBP, systolic blood pressure; LDL, low density lipid; BMI, body mass index.

Fig. 2.

Ranking of 23 stroke risk factors by age-standardized percent of total DALYs in 21 GBD regions (countries in each GBD region are listed below) in 2023, both sexes. Heat map shows age-standardized percentage contribution (PAF) of 23 risk factors to stroke DALYs across 21 GBD regions in 2023. Darker colours indicate greater contribution. Risk factors are ranked within each region. Estimates are age standardized and derived from GBD 2023; contributions are not additive because of overlapping risk exposures. The list of GBD regions and their constituent countries is provided above.

The study findings of the large and significantly increased global effect of high outdoor temperatures on stroke burden is of significant importance for public health, especially for elderly people. Just in 2023, high outdoor temperature, globally, was responsible for over 2.2 million DALYs from stroke, particularly in North and Central Africa, Middle East, and South Asia [2]. For the first time, the GBD 2023 study revealed the significantly increased (45%, 95% UI 6.6–103.0) contribution of high particulate matter air pollution to SAH from 1990 to 2023 [2]. This Global Stroke Fact Sheet 2026 provides the most updated information that can be used to inform communication with all internal and external stakeholders.

Discussion

As for all GBD updates, the latest GBD 2023 stroke burden estimates supersede all previously published stroke burden estimates. This analysis highlights stroke as a major and persistent contributor to the global burden of CVD in 2023, with substantial variation across socio-demographic contexts and stroke subtypes. Despite advances in prevention and care in some regions, the overall burden of stroke remains disproportionately concentrated in low, low-middle, and middle socio-demographic index (SDI) settings, underscoring persistent global inequities in brain health and health system capacity. Compared with earlier GBD analyses (1990–2019 and 1990–2021) [3, 7], the 1990–2023 update does not indicate a sustained post-COVID increase in global stroke burden but rather a continuation of long-term declines in age-standardized incidence, mortality, and DALY rates alongside persistent growth in absolute numbers driven by population ageing and growth. However, the data suggest stagnation of prior gains in some regions and likely short-term pandemic-related fluctuations in stroke care and outcomes rather than a clear global trend reversal.

Age-standardized DALY rates for total stroke were highest in lower SDI regions, reflecting both a higher incidence of stroke and poorer outcomes following acute events. These disparities persist even after accounting for population age structure, indicating that differences in exposure to modifiable risk factors, access to timely diagnosis, and availability of effective acute and post-acute care play a central role. In many low-resource settings, limited access to diagnostic imaging, emergency transport, and specialized stroke units constrains the delivery of evidence-based acute stroke care, contributing to higher mortality and long-term disability. As evidenced by these data, the impact of stroke as the leading cause of loss of DALYs underscores its essential role in public health prevention efforts. Particularly in the lower age categories, loss of productivity and DALYs poses substantial threats to both patient well-being as well as a higher societal burden.

IS remained the dominant stroke subtype globally, accounting for the majority of incident stroke cases in 2023. Its predominance reflects the widespread distribution of vascular risk factors, particularly elevated systolic blood pressure, non-optimal diet, dyslipidaemia, diabetes, and obesity. While high SDI regions experienced substantial declines in age-standardized stroke burden during the early years of the analysis, these improvements have plateaued over the past decade. This stagnation suggests that gains achieved through conventional risk factor control and acute care improvements may be reaching their limits, highlighting the need for renewed emphasis on population-level prevention, early-life risk reduction, and more effective management of emerging metabolic risks. At the same time, sustained reductions in IS burden depend on the availability of well-organized stroke services, including rapid pre-hospital recognition, timely access to neuroimaging, and delivery of evidence-based acute treatments such as intravenous thrombolysis and mechanical thrombectomy. Equitable access to comprehensive stroke units and integrated post-acute rehabilitation is essential to reduce mortality, prevent avoidable disability, and maximize functional recovery following IS.

ICH continues to represent a disproportionately large share of stroke burden in low- and middle-SDI regions. The higher burden of ICH in these settings is closely linked to poorer control of hypertension, limited access to primary prevention and evidence-based stroke services, as well as delays in acute care. Given the generally worse prognosis of haemorrhagic stroke compared with IS, elevated ICH incidence translates directly into higher DALYs and mortality. These findings reinforce the central importance of blood pressure control as a cornerstone of stroke prevention globally, while also highlighting the need for system-level investments in rapid diagnosis, stroke unit care, neurosurgical capacity, and critical care where feasible.

SAH, although accounting for a smaller proportion of total stroke burden, remains a highly consequential subtype due to its high case fatality and long-term disability among survivors. As with ICH, outcomes for SAH are strongly influenced by access to timely diagnosis, specialized neurocritical and neurosurgical care, as well as rehabilitation services, all of which are unevenly distributed across regions. In lower SDI settings, gaps in these services likely magnify the long-term burden associated with SAH beyond what incidence alone would suggest. SAH incidence, as evidenced by the recent GBD publication on the matter, seemed to be decreasing in recent year, only to demonstrate a substantial impact on DALYs and a severe societal burden, particularly for younger patients, underscoring the need for screening and specialized treatment facilities.

The level of contribution of the 24 stroke risk factors to the PAF of age-standardized DALYs varies markedly across the 21 GBD regions, underscoring the need for region-specific prioritization of risk-reduction strategies. This substantial heterogeneity indicates that a uniform, “one-size-fits-all” prevention approach is unlikely to achieve optimal impact. Instead, stroke prevention policies and interventions should be explicitly tailored to target the dominant risk factors driving stroke burden in each region, whether they relate primarily to metabolic risks (such as high systolic blood pressure, obesity, or diabetes), behavioural risks (including poor diet, physical inactivity, smoking, or alcohol use), or environmental exposures. Prioritizing strategies in proportion to their contribution to stroke-related DALYs allows health systems to allocate resources more efficiently, maximize potential gains in population health, and reduce avoidable disability and premature mortality. Such targeted prioritization is particularly critical in low- and middle-income regions, where resources are constrained and the mismatch between prevailing risk exposures and implemented interventions often perpetuates high and inequitable stroke burden. Long-term global GBD data also indicate that environmental exposures are major contributors to stroke burden worldwide, with ambient particulate matter pollution, household air pollution, and non-optimal temperature together accounting for a substantial proportion of stroke DALYs. Particularly concerning is the marked rise in stroke burden attributable to high ambient temperature since 1990, underscoring that effective stroke prevention requires not only local public-health measures but also coordinated international action on air quality, clean household energy, climate-change mitigation, and heat adaptation, especially by large high-income countries whose emissions have disproportionate global effects.

Across all stroke subtypes, rehabilitation is a cornerstone of recovery, yet it remains one of the most under-resourced components of stroke care globally. In most health systems, access to best-practice, multidisciplinary rehabilitation after hospital discharge is limited to a small proportion of people living with stroke-related disability and typically for short, predefined periods only [8, 9]. Comprehensive services, encompassing physiotherapy, occupational therapy, speech and language therapy, and psychosocial support, are often unavailable beyond the early post-acute phase, despite ongoing functional needs. High out-of-pocket costs, shortages of trained rehabilitation professionals, geographic inequities, and a persistent health-system emphasis on acute care further constrain long-term access. As a result, many stroke survivors are discharged home with minimal support, contributing to sustained disability, reduced independence, caregiver burden, and increased societal costs [7, 9]. These limitations are particularly concerning because stroke recovery is a prolonged and dynamic process. While early neurological recovery may occur within weeks to months, evidence shows that meaningful improvements in mobility, strength, balance, communication, cognition, and activities of daily living can continue for many months and even years after stroke, especially when appropriate rehabilitation is available, and in particular for young patients [9, 10]. However, current service models in most countries are not designed to support this extended recovery trajectory, leading to missed opportunities for functional gains.

Only recently have scalable, low-cost, and universally accessible solutions begun to address this gap. A notable example is a free-to-use, online, self-management post-stroke rehabilitation and re-adaptation programme comprising 65 best-practice videos, developed by the New Zealand Stroke Education Charitable Trust (https://www.stroke.net.nz/). These videos are endorsed by the World Stroke Organization and the World Federation of Neurorehabilitation, reflecting alignment with international best-practice standards. The video set is grounded in role-model theory and covers a broad range of practical rehabilitation and daily-living challenges, including muscle strengthening; improving movements in the arms, hands, and legs; balance and mobility training; safe transfers (e.g., bed-to-chair and bed-to-toilet); turning and positioning of immobilized individuals in bed; personal care and hygiene; and essential caregiving tasks such as changing bed linens. These freely accessible resources are designed to complement, rather than replace, professional rehabilitation, empowering stroke survivors and caregivers to continue evidence-based rehabilitation activities beyond the limited period of formal service provision. To date, the videos have been used by more than 118,000 people with lived experience of stroke worldwide, demonstrating substantial real-world uptake. To improve equity and global reach, the videos are currently being translated into 22 of the world’s most widely spoken languages, making them accessible in native languages to over 93 million stroke survivors worldwide, and nearly all those in low- and middle-income countries [11]. Importantly, the feasibility and acceptability of using these videos as part of post-stroke self-management and rehabilitation have been demonstrated in a multi-country randomized controlled trial, showing that digitally delivered, guided rehabilitation support can be safely implemented across diverse health-care settings and cultural contexts [12, 13]. This evidence aligns with a growing body of literature supporting the role of digital and task-sharing approaches in extending rehabilitation reach where workforce capacity is limited [14, 15]. The need for such scalable innovations is likely to intensify. Projections indicate future global shortages in the health and rehabilitation workforce, driven in part by population ageing and declining fertility rates, which are expected to disproportionately affect low- and middle-income countries [15, 16]. Without sustainable solutions that extend rehabilitation beyond hospital- and clinic-based services, the global burden of long-term stroke disability will continue to rise. Freely accessible, evidence-based digital rehabilitation tools therefore represent a critical and timely complement to conventional rehabilitation services and an essential component of future stroke care systems.

Taken together, these findings emphasize that reducing the global burden of stroke will require a dual strategy: intensified prevention targeting dominant and emerging risk factors and substantial investment in equitable stroke care pathways, including acute treatment and rehabilitation. The marked heterogeneity in stroke burden across regions and subtypes also highlights the importance of tailoring policies and interventions to local risk profiles and health system capacities. Without coordinated global and regional action, the substantial and uneven burden of stroke is likely to persist, limiting progress in reducing overall CVD burden worldwide.

The World Stroke Organization (WSO) has identified major challenges and opportunities for reducing the global burden of stroke through evidence-based interventions at both policy and health-system levels [17]. Drawing on the WHO Noncommunicable Disease (NCD) Policy Best Buys [18] and the World Stroke Organization-Lancet Neurology Commission on pragmatic solutions to reduce stroke burden, these recommendations provide a comprehensive framework for governments to meet their NCD commitments. If implemented at scale, these interventions could reverse current trends through effective prevention, timely treatment, and improved rehabilitation.

Primordial prevention: addressing upstream determinants of stroke: primordial stroke prevention aims to reduce exposure to risk factors before they manifest as clinical conditions such as hypertension or diabetes. This requires addressing the social, commercial, and environmental conditions that shape health behaviours and risk. While many governments have implemented awareness campaigns targeting smoking, diet, and physical activity, these individual-focused strategies alone are insufficient.

As highlighted by Walsh et al. [19] in The Lancet Neurology Anniversary Lecture, interventions focused solely on individual behaviour underestimate the influence of commercial interests, social environments, and material circumstances. Editorial commentary [20] further emphasizes that addressing social and commercial determinants is essential for reducing stroke burden. These determinants include poverty, air pollution, food pricing and availability, urban design that discourages physical activity, inequality, and limited access to health information and education.

Governments can reduce stroke risk related to air pollution by regulating vehicle and industrial emissions, restricting traffic in urban centres, expanding public and green transport, and incentivizing clean energy use. India’s subsidies for clean cooking fuels illustrate how policy action can significantly reduce household air pollution and associated health risks.

Food systems, diet, and urban environments: unhealthy diets are a major driver of stroke risk and global mortality. The WHO estimates that around 11 million deaths annually are attributable to unhealthy diets [21], which also contribute substantially to environmental degradation. Governments can reduce diet-related stroke risk through measures such as subsidies for fruits and vegetables; taxes on sugar-sweetened beverages and alcohol; mandatory salt reduction in processed foods; bans on trans-fats; clear food labelling; regulation of food marketing; and provision of healthy food options in public institutions.

Urban planning also plays a critical role. Policies that enhance walkability, increase access to green spaces, separate residential areas from heavy traffic, and limit fast-food outlet density can support healthier lifestyles and reduce exposure to pollution.

Tobacco control and public awareness: WSO-recommended tobacco control measures include public education, advertising bans, age restrictions, taxation, smoke-free public spaces, and smoking cessation services. While awareness campaigns are common, particularly in high-income countries, further research is needed to understand which strategies most effectively drive sustained behaviour change. Given the extensive marketing of unhealthy products, governments should also regulate advertising, as exemplified by the UK’s ban on junk food advertisements before 9 p.m. Regular public education campaigns on stroke prevention and warning signs remain essential.

Population and individual approaches to primary stroke prevention: primary prevention focuses on early detection and management of modifiable risk factors, including hypertension, dyslipidaemia, obesity, diabetes, smoking, physical inactivity, and air pollution exposure. However, an exclusive focus on individuals at high risk is insufficient as most strokes occur in people at low to moderate risk [22]. Effective prevention therefore requires both population-wide strategies and targeted interventions for individuals with elevated risk. Blood pressure control, the single most important stroke prevention priority, requires population-level opportunistic screening, reliable data systems, and legislative action to reduce salt in processed foods. Once risk factors are identified, individuals must have access to affordable medicines, regular monitoring, and supportive environments that enable healthy choices.

Low- and middle-income countries face particular challenges in delivering comprehensive prevention services. Cost-effective digital tools, such as the Stroke Riskometer app [23], can support population-level risk assessment and individual behaviour change. The Stroke Riskometer app is free, available in 30 languages, used in 78 countries, and accessible to over 5.5 billion people. It is endorsed by major global and national cardiovascular and stroke organizations and recommended for worldwide use [16]. The app has been cross-culturally validated in 13 countries [24, 25]. Randomized controlled trials show that its use significantly reduces systolic blood pressure, improves cardiovascular and stroke risk, and increases stroke knowledge in stroke- and CVD-free adults [26–28] (equivalent to approximately 7% reduction in the risk of stroke and higher neurocognitive functioning) [29, 30], and significantly decreases CVD risk (measured by ASCVD [26] and HEARTS risk) compared to placebo/usual care at 9 months. Improved stroke knowledge independently predicts healthier behaviours [31–33]. Evidence suggests long-term effectiveness of stroke preventative strategies (blood metabolic and lifestyle risks control) is enhanced when combined with community health workers, polypill, primary care support, or other preventive strategies [34–37]. There are successful large-scale programmes in India and the WHO South-East Asia Region, including the India Hypertension Control Initiative (IHCI) [38], and the WHO HEARTS and SEAHEARTS programme [39, 40], which set country-specific targets for hypertension control, with the regional goal of placing more than 100 million people with hypertension and diabetes on effective treatment [17].

Successful prevention depends on health systems with strong data infrastructure, universal health coverage, trained staff, affordable medicines, political commitment, and collaboration with stroke support organisations. These efforts align with the WHO Intersectoral Global Action Plan on Epilepsy and Other Neurological Disorders and Sustainable Development Goal 3.4 [41]. Some disease burdens may also benefit from earlier screening and detection, such as unruptured intracranial aneurysms.

Data, Workforce, and Health System Capacity: reliable population-level data on cardiovascular risk factors are essential for effective prevention. Measurement should prioritize objective assessments, such as blood pressure readings, over self-reported data. Where resources are limited, data collection should focus on highly prevalent, strongly predictive, and modifiable risk factors. WHO STEPS surveys [42], alongside digital tools such as Stroke Riskometer and PreventS-MD [23], provide validated platforms for surveillance and risk management. Universal health coverage is critical to reducing inequities in access to prevention services. In its absence, individuals often prioritize treatment over prevention due to out-of-pocket costs. Task shifting from physicians to trained nurses and community health workers, supported by regulatory frameworks, can expand reach, particularly in rural and underserved areas. Addressing workforce shortages and maldistribution remains a central challenge.

Secondary stroke prevention: secondary prevention is essential as stroke and transient ischaemic attack survivors face a high risk of recurrence. Up to 16% experience another stroke within a year [43]. Early assessment and management of underlying conditions – such as hypertension, diabetes, atrial fibrillation, and carotid disease – are critical. Survivors require clear secondary prevention plans, regular follow-up, access to medications, and tailored support for lifestyle modification. Health professionals and stroke organizations play a vital role in supporting survivors to understand and manage their risk factors. Services should address practical challenges, including diet after swallowing impairment and physical activity after disability. Secondary prevention services should be universally accessible and publicly funded.

Acute care and advanced treatments: while effective treatments exist, access remains limited. Intravenous thrombolysis and especially mechanical thrombectomy offer exceptional benefits, surpassing many treatments for other NCDs. Central to improving outcomes for all patients with acute stroke, regardless of eligibility for reperfusion therapies, is access to organized stroke units, which represent the most effective and universally applicable intervention for reducing mortality and disability. Expanding access to these interventions requires investment in stroke units, 24/7 service networks, and workforce training. Although resource intensive, these investments yield long-term savings by reducing disability, rehabilitation costs, and societal burden. A good example of this strategy is the WHO South-East Asia Region (SEAR) Stroke Care Improvement Programme, which has built capacity and initiated stroke care services in Bhutan, Maldives, Myanmar, Timor-Leste, Nepal, and Sri Lanka based on country gap assessments and needs [44].

Evidence of feasibility: several examples demonstrate the feasibility of prevention strategies. South Africa’s sugar-sweetened beverage tax led to a 29% reduction in purchases overall and a 57% reduction in low-income households [45]. Complementary initiatives by the Heart and Stroke Foundation South Africa strengthened public education through evidence-based nutrition messaging and school-based health promotion. In Brazil, a national programme trained community health workers to deliver stroke prevention interventions using task-shifting and digital tools [1]. Participants demonstrated improved knowledge, behaviour change, reduced systolic blood pressure, and increased physical activity. In Australia, the Living Well After Stroke programme [46] showed high satisfaction, sustained behaviour change, and improved wellbeing, particularly in underserved regions.

In conclusion, stroke prevention is among the most effective and cost-efficient strategies to reduce global disease burden. Impactful interventions are feasible even in resource-limited settings. Governments must address social and environmental determinants, support healthy lifestyles, ensure universal health coverage, organize evidence-based stroke services in all hospitals admitting people with acute stroke, strengthen data systems, invest in community health workers and digital tools, deliver culturally appropriate prevention strategies, and sustain stroke awareness campaigns. The Global Stroke Action Coalition [47], launched by the World Stroke Organization in 2025, is a global advocacy initiative uniting partners from the public and private sector to drive urgent commitment, investment, and action on stroke prevention, treatment, and continuing care. The diverse membership facilitates inclusive representation across geographical regions, income levels, and disciplinary perspectives. The coalition aims to translate evidence into policy by aligning stakeholders around shared priorities, strengthening accountability, and supporting countries to implement scalable, equity-focused stroke solutions across the full continuum of care.

Acknowledgments

We would like to thank Prof. Simon Hay for the permission to use the latest GBD 1990–2023 CVD paper and IHME GBD Visualisation Tool for the paper. We would also like to thank Prof. Jeyaraj D. Pandian for review and suggestions for the manuscript. Translations of the stroke prevention and recovery videos into 22 languages were made possible through an unrestricted educational grant from the Neurological Foundation of New Zealand, which partially supported expenses related to this project and is gratefully acknowledged.

Conflict of Interest Statement

V.L.F. declares that the free Stroke Riskometer app and PreventS-MD are owned and copyrighted by Auckland University of Technology Ventures Ltd. and PreventS-MD Ltd., respectively, in which he is a shareholder and Chief Scientific Advisor. B.N. has received honoraria from Symbec-Orion, Moleac, and Merck companies. V.L.F., V.V., M.B., and B.N. declare that they are Review Board Members at Neuroepidemiology. V.L.F. is also Editor in Chief of Neuroepidemiology. None of the other authors declared any conflict of interest.

Funding Sources

No funding was received for this manuscript.

Author Contributions

Valery L. Feigin conducted the literature reviews, extracted relevant data from the Roth et al. GBD 1990–2023 paper [2], calculated the statistics from the IHME GBD Viz tool [6], and developed the draft Fact Sheets. Bo Norrving, Michael Brainin, Sheila Martins, Victor Volovici, and Greg A. Roth provided guidance, input, and edits throughout the development of this work. All authors reviewed and approved the final version of the manuscript.

Funding Statement

No funding was received for this manuscript.