Our analysis was based on the most recent data sourced from the GBD 2021,¹ which offered a comprehensive and comparative assessment of 288 estimates related to death causes, along with 87 risk factors, across 21 regions and 204 countries and territories during 1990–2021 (17). Our study specifically concentrated on ischemic strokes occurring in individuals aged 25 years and older. The participants were categorized into consecutive five-year age brackets (20–24, 25–29, 30–34… 90–94, and 95 + years). We also extracted the socio-demographic index (SDI) from the GBD 2021 study, a composite measure made for each country according to education, income per capita, and fertility rate. To classify all countries and territories, the SDI values in 2021 were utilized to divide them into five regions: low, low-middle, middle, high-middle, and high SDI regions.

The latest data comprised information on deaths, Disability-adjusted life year (DALYs) and corresponding age-standardized rate (ASR) for ischemic stroke, categorized by country, region, year, and age groups. DALYs were derived by aggregating the years lived with disability (YLDs) and the years of life lost (YLLs). YLDs were computed by multiplying the prevalence by the respective standardized disability weights for each health state. Additionally, YLLs were estimated based on a reference maximum observed life expectancy. ASRs were widely utilized to enhance the accuracy of disease burden comparisons, as they considered the age variation across populations.

Particulate matter pollution encompassed both ambient PM2.5 and household PM2.5. The former aspect, known as ambient particulate matter pollution, referred to as the annual mean concentration of particles with a diameter less than 2.5 micrometers, weighted by population (18). The latter aspect, commonly described as household air pollution stemming from solid fuels like coal, wood, dung, and charcoal, was estimated using both the proportion model and the PM2.5 mapping model (19). The impact of PM2.5 exposure on ischemic stroke was assessed using the proportional population attributable fraction, while simultaneously considering the potential effects of geography, year sex, and age.

To evaluate the temporal trends, we implemented two approaches, including the estimated annual percentage change (EAPC) and the average annual percentage change (AAPC). A linear regression model was applied to compute the EAPC, and two formulas were set as followed.

In these equation, X denoted the calendar year, ε represented the error term, β referred to the coefficient. We also employed joinpoint analysis to compute annual percentage change (APC) along with its associated 95% confidence interval (CI). The AAPC was measured as the weighted average of APCs to succinctly describe the trends over a predetermined fixed period. The equation employed to compute AAPC was set as followed.

where b symbolized the coefficient for the segment, w_i signified the length of the segment, i referred to the specific ith segment.

Decomposition analysis, motivated by Das Gupta’s method, was applied to gain insights into the disparities in disease burden across different regions (20). This approach attributed DALYs difference to the combined influences of three factors, such as population size, population aging, and epidemiological changes. We carried out the pearson’s correlation analysis to evaluate the relationship between ASRs and SDI. All statistics and visualization were executed via the R software (version 4.2.2). The establishment of statistical significance was predicated on a p value less than 0.05.

In 2021, the global impact of PM2.5 air pollution resulted in approximately 0.90 million ischemic stroke deaths and 18.29 million DALYs, representing an ASMR of 11.01 (95% UI 8.44–13.96) per 100,000 population and an ASDR of 215.64 (95% UI 168.84–266.03) per 100,000 population (Table 1). The ASRs in 2021 were less than those recorded in 1990, suggesting a downward trend as indicated by negative values for both EAPC (−2.19, 95%CI: −2.36 to −2.03 for ASMR; −2.04, 95%CI: −2.20 to −1.88 for ASDR) and AAPC (−2.13, 95%CI: −2.2 to −2.05 for ASMR; −1.98, 95%CI: −2.05 to −1.90 for ASDR) (Table 1; Supplementary Tables S1, S2; Supplementary Figures S1A,B). It indicated two periods of continuous decline in the ASMR and ASDR based on the joinpoint analysis (Figures 1A,D; Supplementary Tables S1, S2).

The ASMR and ASDR of ambient PM2.5-related ischemic stroke exhibited a slight downward trend, as evidenced by the negative EAPC and AAPC (Supplementary Tables S3–S5). It indicated a similar change pattern for both ASMR and ASDR according to the joinpoint analysis (Figures 1B,E). Specifically, the ASR decreased from 1990 to 2008, increased from 2009 to 2014, and then declined again from 2015 to 2021. The ASRs of household PM2.5-associated ischemic stroke decreased staggeringly over past three decades (Figures 1C,F; Supplementary Table S6). In 2021, household PM2.5 contributed a notably lower proportion to the overall air pollution-related ischemic stroke burden compared to ambient PM2.5 (Figures 1G,H; Supplementary Figures S1C,D).

With respect to the SDI region, household PM2.5 had a more pronounced impact on the ischemic stroke burden in low SDI region, whereas ambient PM2.5 made a more dominant contribution in middle, high-middle, and high SDI regions (Figures 1G,H). In all five SDI regions, the proportion of ischemic stroke death and DALYs attributed to household PM2.5 showed a decrease in 2021 compared to those recorded in 1990, while ambient PM2.5-related percentage rose (Supplementary Figures S1C,D). For ischemic stroke attributed to PM2.5 or household PM2.5, the ASRs tended to be lower in regions with higher SDI (Table 1; Figure 2). Both ASMR and ASDR of household PM2.5-associated ischemic stroke approached zero in high SDI region (Supplementary Table S6). In the low, low-middle, and middle SDI regions, the ASRs of ambient PM2.5-related ischemic stroke increased slowly over time, supported by positive values for both AAPC and EAPC (Figure 2; Supplementary Tables S3–S5).

With respect to the GBD region, Africa suffered the heaviest ischemic stroke burden attributable to PM2.5 in 2021 (Table 1). While the highest ASMR and ASDR were manifested in the Africa, East Asia, and North Africa and Middle East. The distribution of ischemic stroke burden attributable to ambient PM2.5 and household PM2.5 varied among the GBD regions (Supplementary Figures S1C,D). Ambient PM2.5-related ischemic stroke exceeded 95% of the total burden in some high-income regions, namely High-income Asia Pacific, High-income North America, Western Europe, and Australasia (Figures 1G,H). A substantial proportion of the disease burden was attributed to household PM2.5 in Central, Eastern, and Western Sub-Saharan Africa, as well as Oceania. The contribution of household PM2.5 to both ischemic stroke mortality and DALYs declined across all GBD regions from 1990 to 2021.

In addition, there was a notable decline in both ASMR and ASDR for ischemic stroke associated with PM2.5 and household PM2.5 in almost all GBD regions (Table 1; Supplementary Tables S1–S4, S6). However, both ASMR and ASDR for ambient PM2.5-related ischemic stroke significantly increased in Oceania, Sub-Saharan Africa, Central Asia, East Asia, South Asia, and Southeast Asia (Supplementary Tables S3–S5).

In the context of national level, the burden of PM2.5-related ischemic stroke displayed heterogeneity across 204 countries. ASMR and ASDR showed an upward trend in 15 and 12 countries, respectively. A downward trend was recorded for ASMR in 177 countries, and for ASDR in 179 countries (Supplementary Table S7). India, China, Bangladesh, and Indonesia ranked as the top four countries with the highest PM2.5-related ischemic stroke burden. The highest ASMR and ASDR were observed in several countries, such as Guinea-Bissau, Ghana, Gambia, Sierra Leone, Togo, Senegal, Benin, Liberia (West Africa), Egypt, Yemen (North Africa and Middle East), Haiti and Afghanistan (Figure 3A; Supplementary Figure S2A). Estonia and Portugal had the most minimal EAPC value, all less than −9.0 (Figure 4A; Supplementary Figures S3A, S4A, S5A). In contrast, the highest EAPC for both ASMR and ASDR were observed in Lesotho, Zimbabwe, Libya and Mozambique.

Ambient PM2.5-related ASMR decreased in 114 countries, increased in 69, and remained stable in 21 (Supplementary Table S8). ASDR declined in 116 countries, rose in 64, and stayed unchanged in 24. China, India, and Egypt were the top three countries with the highest burden of ambient PM2.5-related ischemic stroke. The maximal ASRs of ambient PM2.5-associated ischemic stroke were observed in the countries in North Africa and the Middle East, including Egypt, Iraq, Saudi Arabia, United Arab Emirates, Oman and Bahrain, along with those in Eastern Europe such as Bulgaria, North Macedonia, and Bosnia and Herzegovina (Figure 3B; Supplementary Figure S2B). Eastern and Southern Asia (Viet Nam, Mongolia, Bhutan, Timor-Leste, China, and India) and Africa (Cabo Verde, Tanzania, Lesotho, Angola, Sudan, Ghana, Kenya) exhibited the highest EAPC for both ASMR and ASDR (Figure 4B; Supplementary Figures S3B, S4B, S5B). More specifically, Viet Nam, Mongolia, and Equatorial Guinea emerged as the top three countries with the highest EAPC values above 4.0.

Over past 32 years, a declining trajectory was identified in household PM2.5-related ASMR across 184 countries and in household PM2.5-related ASDR across 185 countries (Supplementary Table S9). India, China, and Bangladesh were recognized as the top three countries bearing the highest cases of household PM2.5-related ischemic stroke. The majority of the top ten countries with the high ASRs consisted of Africa nations, such as Guinea-Bissau, Mozambique, Gambia, Sierra Leone, Madagascar, Togo, and Guinea (Figure 3C; Supplementary Figure S2C). Among the top ten countries with the lowest EAPC in ASRs, the majority were situated in Europe, encompassing Estonia, Portugal, Norway, Luxembourg, Austria, Finland, Sweden, United Kingdom, and Ireland (Figure 4C; Supplementary Figures S3C, S4C, S5C). The countries of Viet Nam, Mongolia, and Equatorial Guinea stood out as the top three with the highest EAPC in both ASMR and ASDR.

The age-specific death and DALYs rate for ischemic stroke associated with PM2.5 or ambient PM2.5 escalated with age increase (Figures 5A,B; Supplementary Figures S6A,B). For ischemic stroke associated with household PM2.5 exposure, the age-specific DALYs rate peaked in the 80–84 age group, while the age-specific mortality rate reached its highest point in the 90–94 age group (Figure 5C; Supplementary Figures S6C). Furthermore, the EAPCs of both death and DALYs rates exhibited negative values across all age groups for ischemic stroke attributable to PM2.5 or household PM2.5, signifying a notable declining trend (Figures 5D,F; Supplementary Figures S6D,F). For ambient PM2.5-associated ischemic stroke, the age-specific death and DALYs rate exhibited an upward trend among individual aged below 50, while demonstrating a decline among those aged above 60 (Figure 5E; Supplementary Figures S6E).

The findings of joinpoint analysis for ASMR and ASDR, stratified by gender, were shown in Supplementary Figure S7. Contrary to the dramatic drop in ASRs for ischemic stroke attributed to PM2.5 or household PM2.5, the ASRs trend for ambient PM2.5-related ischemic stroke remained relatively constant in the male populations. Additionally, the second segment of the ASRs temporal trends increased significantly in ischemic stroke associated with ambient PM2.5 for both female and male populations. For both ambient and household PM2.5, the mortality and DALYs rates were significantly higher in males than in females (Figures 5B,C; Supplementary Figures S6B,C). Interestingly, among individuals aged below 80, the mortality and DALYs rates for ischemic stroke attributed to household PM2.5 were nearly identical between males and females (Figure 5C; Supplementary Figure S6C). In the 65–95 age group, the age-specific rates of ambient PM2.5-related ischemic stroke exhibited a significant decline among females, whereas males demonstrated a relatively stable trend.

For ischemic stroke associated with PM2.5 or household PM2.5, both ASMR and ASDR decreased with increased SDI (Figures 6A,B). These observations were consistent that in the high-middle and high SDI regions for ischemic stroke associated with ambient PM2.5. However, in the regions with low, low-middle, and middle SDI, the ASRs related to ambient PM2.5 demonstrated an upward trend corresponding to the SDI rise. In the GBD regions with an SDI below 0.7, including Southern Sub-Saharan Africa and East, South, and Southeast Asia, a positive correlation was identified between SDI and ASMR linked with ambient PM2.5 (Supplementary Figure S8). The pattern of ASDR based on SDI at the regional level was consistent with that of ASMR (Supplementary Figure S9).

There was an inverse relationship between ASRs and SDI across 204 nations for both ischemic stroke attributed to PM2.5 (R = −0.51, p < 0.01 for ASMR in 1990; R = −0.82, p < 0.01 for ASMR in 2021; R = −0.57, p < 0.01 for ASDR in 1990; R = −0.75, p < 0.01 for ASDR in 2021) and household PM2.5 (R = −0.75, p < 0.01 for ASMR in 1990; R = −0.82, p < 0.01 for ASMR in 2021; R = −0.77, p < 0.01 for ASDR in 1990; R = −0.82, p < 0.01 for ASDR in 2021) (Supplementary Figures S10A,C, S11A,C, S12A,C, S13A,C). Both in 1990 and 2021, a unimodal curve was perceptible when examining the correlation between SDI and ASRs for ischemic stroke associated with ambient PM2.5. It is noteworthy that over this period, the peak of the correlation curve underwent a transition, moving from 0.8 down to 0.7 (Supplementary Figures S10B, S11B, S12B, S13B).

It presented a substantial worldwide escalation in DALYs attributed to PM2.5 and ambient PM2.5, but a significant reduction in DALYs linked to household PM2.5 (Figure 7; Supplementary Table S10). In the low and low-middle SDI regions, the DALYs increase of ischemic stroke related to PM2.5 and household PM2.5 was primarily driven by population growth and aging. The effect of epidemiological changes on DALYs growth showed a negative impact. All epidemiological changes, population growth and aging led to a considerable DALYs escalation linked to ambient PM2.5 across the low, low-middle, and middle SDI regions. In the high SDI region, the epidemiological changes led to a considerable DALYs reduction.