Global, Regional, and National Burden Attributed to Particulate Matter Pollution, 1990–2021: A Systematic Analysis for the Global Burden of Disease Study 2021

Global trends in PMP‑attributable deaths and DALYs

In 2021, the number of global PMP‑attributable deaths was 7,833,220.92 (95% UI, 6,479,473.76 to 9,263,395.35), with 4,718,812.24 (3,480,471.83 to 5,795,946.44) attributed to APMP and 3,112,926.41 (1,895,168.20 to 5,188,696.63) to HPMP. The global PMP‑attributable DALYs were 231,511,232.94 (194,538,892.20 to 270,855,451.37), with 120,004,672.01 (86,560,331.22 to 149,810,185.34) for APMP and 111,462,958.29 (75,085,852.45 to 163,710,710.53) for HPMP (Table 1). From 1990 to 2021, PMP‑attributable deaths increased globally, primarily due to a doubling in APMP‑attributable deaths for both females and males (Tables S1–S3, Figure S1‑3A). Despite a decrease in global PMP‑attributable DALYs, APMP‑attributable DALYs still doubled for both sexes (Tables S4–S6, Figure S1–3B). However, ASRs of PMP‑attributable deaths and DALYs significantly declined globally from 1990 to 2021, with similar trends for both APMP and HPMP in females and males (Tables S1–S6, Figures S4–S6).

Age‑ and gender‑specific patterns in PMP‑attributable deaths and DALYs.

There were two peaks in PMP‑attributable age‑specific deaths: one at age 0–6 days for both sexes and the other at age 70–74 in males and 80–84 in females (Figure 1A). The pattern of DALYs was slightly different, with the other peak at age 65–69 in males and age 70–74 in females (Figure 1B). Before ages 85–89 for deaths and 80–84 for DALYs, males had higher numbers than females, but this trend reversed afterwards. Except for the 0–4 years age group, the age‑specific rates of deaths and DALYs increased with age before age 90–94 in males, then declined, while they continually increased in females. Additionally, males consistently had higher age‑specific rates of deaths and DALYs than females. APMP‑ and HPMP‑attributable age‑specific deaths and DALYs showed similar patterns (Figures S7 and S8).

Figure 1

Regional variations in PMP‑attributable deaths and DALYs

Across the 21 GBD regions in 2021, the highest APMP‑attributable ASRs of deaths (102.51 per 100,000 people) and DALYs (2,399.07) were observed in North Africa and the Middle East (Tables S7 and S8). The lowest rates were seen in high‑income North America (7.75 deaths per 100,000 people) and Australasia (183.19 DALYs). The highest HPMP‑attributable ASRs were in Oceania (219.79 deaths and 5,997.06 DALYs per 100,000 people [Tables S9 and S10]). The lowest HPMP‑attributable rates were in high‑income North America, Australasia, high‑income Asia Pacific, and Western Europe. Overall, PMP burdens showed similar patterns to HPMP (Tables S11 and S12). The highest number of PMP‑attributable deaths and DALYs occurred in East Asia and South Asia. From 1990 to 2021, South Asia had the largest increase in ASRs of APMP‑attributable deaths (EAPC = 2.39%) and DALYs (1.74%), while high‑income North America had the largest decrease (EAPC = −5.77% for deaths and −5.10% for DALYs) (Tables S7 and S8). All 21 GBD regions saw a decrease in ASRs of HPMP‑ and overall PMP‑attributable deaths and DALYs (Tables S9–S12).

Country‑specific trends in PMP‑attributable deaths and DALYs

In 2021, Egypt had the highest ASRs of APMP‑attributable DALYs and deaths among the world’s 20 most populous countries, with 5,279.18 DALYs and 251.38 deaths per 100,000 people (Tables S13 and S14, Figure 2A, Figure S9A). The United States of America had the lowest rates, with 8.20 deaths and 224.24 DALYs per 100,000 people. For HPMP, the highest rates were in the Solomon Islands (337.76 deaths per 100,000 people) and the Central African Republic (8,596.53 DALYs) (Tables S15 and S16, Figures S10 and S11A). Among populous countries, the Democratic Republic of the Congo had the highest ASRs, with 208.91 deaths and 5,369.44 DALYs per 100,000 people. Overall, PMP burdens mirrored those of HPMP (Tables S17 and S18, Figure S12 and S13). From 1990 to 2021, Viet Nam (one of the world’s 20 most populous countries) saw the largest increases in APMP‑attributable deaths (EAPC = 5.13%) and DALYs (4.59%), while Estonia experienced the biggest decline (−8.31% for DALYs and −8.54% for deaths) (Tables S13 and S14, Figure 2B, Figure S9B). The United Kingdom had the largest decline among populous countries, with EAPCs of −5.69% for DALYs and −6.2% for deaths. Only the Northern Mariana Islands, Zimbabwe, and Lesotho saw increases in HPMP‑attributable deaths and DALYs from 1990 to 2021 (Tables S15 and S16, Figures S10 and S11B).

Figure 2

Leading causes of PMP‑attributable deaths and DALYs

In 2021, the leading four level 2 causes of PMP‑attributable deaths were CVDs (53.62 per 100,000 people; PAF = 15.15% from APMP and 7.64% from HPMP), chronic respiratory diseases (CRDs) (18.51; 19.09% and 15.48%), respiratory infections and tuberculosis (8.68; 2.76% and 3.57%), and maternal and neonatal disorders (8.03; 7.12% and 18.01%) (Table 1, Tables S2 and S3). The leading causes of PMP‑attributable DALYs were CVDs (1161.77; 14.65% and 8.33%), maternal and neonatal disorders (723.06; 6.61% and 16.73%), respiratory infections and tuberculosis (420.28; 3.29% and 6.34%), and CRDs (389.50; 16.09% and 14.01%) (Table 1, Tables S5 and S6). From 1990 to 2021, only diabetes and kidney diseases and neoplasms showed an increasing trend in age‑standardized DALY rate attributable to APMP (Table S5). Other GBD level 2 causes of APMP‑, HPMP‑, and overall PMP‑attributable deaths and DALYs are detailed in supplementary tables and figures (Tables S1–S6 and Figures S1–S6).

In 2021, among all GBD level 3 causes, the highest global age‑standardized death rates attributable to PMP were for ischemic heart disease (IHD) (29.88 per 100,000 people; PAF = 19.18% from APMP and 8.30% from HPMP), stroke (23.74; 16.90% and 10.24%), and chronic obstructive pulmonary disease (COPD) (18.51; 22.61% and 18.33%) (Tables S1–S3). The highest global age‑standardized DALY rates attributable to PMP were for neonatal disorders (723.06; 6.96% and 17.61%), IHD (638.48; 19.34% and 9.52%), stroke (523.30; 16.59% and 11.16%), lower respiratory infections (LRI) (420.09; 12.25% and 23.58%), COPD (389.50; 22.14% and 19.27%), diabetes mellitus (DM) (148.92; 11.23% and 5.01%), and tracheal, bronchus, and lung cancer (TBL) (102.08; 14.94% and 4.21%) (Tables S4–S6). From 1990 to 2021, only DM and TBL showed increases in age‑standardized DALY rates attributable to APMP (Figure 3).

Figure 3

Regional, gender, and age group variations in PMP‑attributable disease burdens

Figure 3 and Figures S14–S18 provide detailed visualization of the leading ten level 3 attributable causes in 2021 and their EAPCs in age‑standardized DALY and death rates from 1990 to 2021 for each region. IHD had the highest PMP‑attributable deaths and DALYs in high, high‑middle, and middle SDI regions, whereas neonatal disorders had the highest burdens in low or low‑middle SDI regions. From 1990 to 2021, burdens of DM, TBL, and IHD attributable to APMP significantly increased in middle‑, low‑middle, and low SDI regions. Only high SDI regions showed decreasing trends in ASRs of deaths and DALYs for the leading 10 causes attributable to PMPs.

The proportional contribution of APMP to age‑standardized DALYs was highest in NCDs, with the leading three being COPD, IHD, and stroke (Figure 4). In contrast, HPMP contributed most to DALYs in LRI, COPD, and neonatal disorders. Cataract DALYs were solely attributable to HPMP (Table S6). For most GBD level 3 causes, the proportional contribution of APMP to age‑standardized DALYs was higher in males than in females (9 of 13 GBD level 3 causes), except for sudden infant death syndrome (SIDS), encephalitis, upper respiratory infections (URI), and otitis media. Conversely, HPMP’s proportional contribution was higher in females (9 of 14 causes), except for neonatal disorders, meningitis, diarrheal diseases, and otitis media (Figure 4). Global rankings of fractions of level 3 causes in age‑standardized deaths attributable to APMP and HPMP across females and males in 2021 were provided in Figure S19.

Figure 4

There were substantial differences in the attributable proportions of age‑standardized DALYs due to PMPs across regions and age groups (Figure 5, Figures S20–S33). In 2021, Eastern and Western Sub‑Saharan Africa had the highest attributable proportions of age‑standardized DALYs due to PMP for the leading five GBD level 3 causes in both sexes, whereas high‑income North America and high‑income Asia Pacific had the lowest attributable proportions for the leading eight GBD level 3 causes in both sexes. By age group, notably, PMP was only associated with DALYs for neonatal disorders, diarrheal diseases, meningitis, SIDS, encephalitis, URI, and otitis media among children aged under 5 years.

Figure 5

Global burden attributable to ambient and household particulate matter pollution

Our study showed that in 2021, PMP posed a significant burden on global health, largely attributable to APMP, which has seen a substantial increase in both deaths and DALYs over the years. According to WHO population data and IHME reports, PMP‑related deaths accounted for over one‑tenth of global deaths and were a major cause of global DALYs [18, 19]. This aligns with a comprehensive analysis of 88 global risk factors, which identified PMP as the leading level 3 risk factor in 2021 and the primary contributor to the global health burden [4]. Initiatives under the United Nations (UN) Sustainable Development Goals (SDGs) to enhance access to clean household energy, such as grid electricity provision and cleaner cooking methods [20], have significantly reduced deaths and DALYs attributable to HPMP. However, the latest WHO data revealed that high levels of APMP persisted in large regions of the world, with 99% of the global population breathing air that exceeded WHO’s recommended air quality limits [21]. This may be mainly due to rapid industrialization and urbanization, especially in developing countries, leading to higher levels of APMP and increased urban population exposure, thus heightening the health burden [22]. Additionally, according to data from the UN and WHO, continuous global population growth and significant aging trends over the past 30 years have amplified the impact of PMP [9, 10].

Despite the increase in absolute deaths and DALYs caused by APMP, our study revealed that from 1990 to 2021, age‑standardized death and DALY rates attributable to global PMP significantly declined. These divergent trends reflected the competing influences of population growth, aging, and increased global environmental exposure on the burden of disease attributable to PMP. Moreover, the decline in age‑standardized death and DALY rates attributable to PMP can be attributed to several factors. First, over the past 30 years, significant efforts have been made globally to control PMP. WHO member states have implemented stringent air quality management policies and regulations, such as limiting industrial emissions, setting vehicle emission standards, and controlling emissions from coal‑fired power plants [23, 24]. Second, continuous advancements in socioeconomic conditions and scientific research have led to improvements in air quality monitoring systems, the development of clean energy technologies, and advancements in medical technologies and facilities, all contributing to the reduction of PMP and its associated disease burden [25, 26]. Additionally, increased public awareness of PMP pollution and its health risks, coupled with a heightened emphasis on environmental protection, has helped mitigate exposure risks [27]. These demonstrated the global capacity to adapt and improve in the face of public health challenges. However, the persistently high burden of disease attributable to PMP underscored its continued significance not only among environmental risk factors but also in comparison to dietary, lifestyle, and other modifiable risk factors, indicating that global efforts to control PMP remain insufficient.

Socio‑demographic differences in household particulate matter pollution and ambient particulate matter pollution

Our study found that global age‑standardized death and DALY rates attributable to HPMP decreased with increasing social development, likely due to differences in household fuel use and economic and sanitary conditions. In 2018, the Solomon Islands government initiated a pilot study to address cultural and developmental barriers in Oceania’s transition to clean energy practices, formulating a visionary energy transition plan that included clean energy for household cooking [27]. In 2017, the Central African Republic’s Ministry of Environment and Sustainable Development collaborated extensively with the Supporting National Planning initiative to establish comprehensive emission inventories and enhance the capacity to mitigate short‑lived climate pollutants like black carbon and methane [28]. Similarly, in 2017, the Democratic Republic of Congo joined the Climate and Clean Air Coalition, striving to fulfill its international environmental commitments [29]. Despite these three countries having the highest global disease burden attributable to HPMP, their efforts have led to a significant decline in this burden. Our results indicated that 95% of countries and regions have seen a significant reduction in the global disease burden attributable to HPMP, demonstrating the feasibility of current global policies addressing HPMP. However, the Northern Mariana Islands, Zimbabwe, and Lesotho continued to experience the fastest increase in this burden. In 2021, the WHO provided a household energy policy repository, suggesting that these countries should learn from others and tailor their environmental pollution policies [24]. Additionally, these low SDI countries often have poorer medical technology and healthcare services, also contributing to a higher disease burden [30].

However, countries and regions with middle SDI levels exhibited the highest global age‑standardized death and DALY rates attributable to APMP, consistent with the inverted U‑shaped relationship in Figure 6C, while those with high SDI showed the lowest rates. Moreover, middle and lower SDI regions showed rising trends in these rates, highlighting disparities in global development. Rapid industrialization, urbanization, and inadequate environmental regulations contribute to severe pollution in these areas [31]. For instance, Eastern and Western sub‑Saharan Africa had the highest proportion of age‑standardized DALYs attributable to PMP‑related diseases. The IHME also reported that PMP exposure rates in Asia, Africa, and the Middle East are nearly double the global average of 57.6 per 100,000 [15]. Nationally, our study identified Egypt as having the highest global disease burden attributable to APMP, with Viet Nam showing the fastest growth. According to a real‑time air quality tool, air pollution levels in Cairo, Egypt, exceed acceptable global standards by tenfold [32]. In 2020, Egypt launched the 2030 Sustainable Development Strategy to reduce particulate emissions by 50% as part of efforts to tackle urban pollution [33]. According to the International Trade Administration, Viet Nam ranked 14th among 118 countries with the most severe air pollution in 2022. In the same year, Viet Nam introduced the National Environmental Protection Strategy 2030, aiming to prevent pollution escalation and environmental degradation by 2050 [34]. Additionally, among the top 20 populous countries, the United States of America had the lowest global disease burden attributable to APMP, with the United Kingdom showing the most significant decline. An observational study revealed that these high SDI countries typically benefit from high‑quality healthcare systems, underscoring the critical need for research to improve healthcare in low‑ and middle‑income countries [35]. It is imperative for low and middle SDI countries to learn from other countries’ experiences, seek international cooperation and assistance, strengthen environmental regulation and governance, and enhance preventive healthcare services for related diseases.

Disparate impact of particulate matter pollution on diseases

Our study found significant heterogeneity in the spectrum of diseases attributable to PMP, with the leading level 2 causes being CVDs, maternal and neonatal disorders, respiratory infections and tuberculosis, and CRDs. Regarding CVDs, an exposure−response analysis supported our findings, linking long‑term PMP exposure to increased hospital admissions for first‑time IHD, cerebrovascular disease, heart failure, arrhythmias, cardiomyopathy, and thoracoabdominal aortic aneurysm, with the first four being particularly sensitive to PMP [36]. A meta‑analysis of cohort studies revealed significant associations between several PMP components and CVD mortality, such as black carbon (BC), sulfate (SO₄^2–), nitrate (NO₃^–), nickel (Ni), and zinc (Zn), with BC having the most substantial impact [37]. These components may affect cardiovascular health through systemic inflammation, vasoconstriction, changes in cardiac electrical activity, and thrombosis [38]. Another prospective cohort study suggested organic matter (OM) and chloride within PMP mostly impact CVDs, with OM potentially driving metabolic syndrome factors like obesity, high triglycerides, and high blood pressure [39]. Additionally, a fixed cohort study that integrated targeted biomarkers and multi‑omics results identified novel biomarkers, such as decay‑accelerating factor, cholesterol ester transfer protein, and polyprotein‑2, suggesting that ultrafine particles under 100 nm have unique mechanisms affecting cardiovascular health and may pose greater health risks than larger particles [40]. These studies implied that controlling specific PMP components and particle sizes could prevent CVDs and that intervening in metabolic risk factors may reduce the incidence of PMP‑related CVDs.

Disparate age group impact of particulate matter pollution

The global disease burden attributable to PMP exhibited age‑related differences. Our study showed that one of the global peaks in deaths and DALYs occurs in the infant and young child population. The impact of PMP on infants differs from that on adults. According to the Health Effects Institute Special Report for global air, in 2021, PMP became the second largest risk factor for deaths and DALYs in children under five, including neonates [15]. Systematic reviews indicated that maternal exposure to air pollution increases the risk of miscarriage, preterm birth, and low birth weight [41]. Experimental evidence suggested that PMP exerts cytotoxic effects by mediating stress response pathways associated with inflammation, metabolic changes, and apoptosis, leading to organelle damage and oxidative stress [42, 43]. Numerous studies have shown that inhalation of PMP pollutants has systemic adverse effects on infant health, affecting multiple organs, various cell types, and molecular mediators, thereby impairing their defenses against infections [44–46]. Our study also revealed that DALYs attributable to neonatal diseases, diarrheal diseases, meningitis, SIDS, encephalitis, URI, and otitis media occur only in children under five, with LRI being the leading cause of death in this age group. Moreover, prenatal and early childhood exposure to PMP increases lifelong risks of cancer and other diseases [1], including cardiovascular issues from secondhand smoke [47]. The UNSDGs should advocate substantially reducing under‑five mortality, especially in low SDI regions. Healthcare professionals and the public should address air pollution, particularly household air pollution, by adopting clean cooking and heating fuels and lighting technologies to benefit maternal and infant health.

Another global peak in deaths and DALYs occurred in the elderly, who bear the heaviest burden of PMP‑attributable NCDs such as COPD, diabetes, stroke, heart disease, lung cancer, and chronic lung diseases. Previous studies have established that COPD is one of the leading causes of morbidity and mortality among the elderly, with air pollution contributing to approximately 50% of COPD risk [48]. Meta‑analyses showed that certain PMP components (NO₃^–, silicon, and Zn) are significantly associated with respiratory disease mortality, with NO₃^– having the greatest impact [37]. Moreover, our results also indicated a significant impact of PMP on COPD DALYs in younger populations, and recent studies suggested that combined exposure to household air pollution and ozone exacerbates COPD risk in young people [49]. This underscores the importance of early COPD diagnosis in younger individuals to facilitate timely pharmacological interventions and the formulation of preventive and therapeutic measures based on pollutant components to slow disease progression and achieve better outcomes. According to the WHO, IHD, stroke, and COPD are the top three global causes of death, with lung cancer and diabetes among the top ten [50]. Air pollution significantly contributes to the high burden of these chronic NCDs, and individuals with these diseases are, in turn, more susceptible to PMP’s toxic effects [51]. Moreover, our findings revealed that from 1990 to 2021, the burden of NCDs increased, with diabetes and tracheal, bronchial, and lung cancer showing upward trends in APMP‑attributable age‑standardized DALY rates. Multi‑omics studies also highlighted the increasingly significant relationship between PMP and metabolic diseases, including type 2 diabetes mellitus, suggesting that PMP contributes to a growing number of chronic disease cases, leading to productivity loss and an overburdened healthcare system [52]. In 2019, the World Health Assembly included air pollution as a risk factor in the NCD framework [53], which should prompt more public health actions at the societal level, particularly among the elderly. Additionally, health education for individuals and families is crucial to mitigate lifestyle risk factors like poor diet, physical inactivity, smoking, and excessive alcohol consumption on PMP‑related disease burden.

Disparate gender impact of particulate matter pollution

Furthermore, the global disease burden of PMP exhibited gender differences. Our findings showed that overall death and DALY rates are consistently higher in males than females, with APMP contributing a higher proportion of age‑standardized DALYs in males for most GBD level 3 causes. A meta‑analysis of cohort studies supported our results, indicating that PMP components had a greater impact on males [37]. However, for specific diseases such as diabetes, PMP exposure had a more pronounced association in females [54]. A qualitative evaluation study showed that females, particularly the elderly or those assessed for residential exposure, are more strongly affected [55]. Studies on children indicated that boys were more affected in early life, while girls were more affected in later childhood. It remains unclear whether these observed differences are due to gender‑related biological responses (e.g., hormone composition, body size), behavioral patterns (e.g., smoking), or exposure patterns (e.g., high‑risk occupations, household cooking). More research should focus on emerging methods for quantitative gender analysis to elucidate the roots of these differences, enabling targeted interventions to control PMP and mitigate its adverse effects, thereby improving population health across different societies and genders.

Strengths and limitations of the study

This study is the first to comprehensively evaluate the levels and trends of deaths and DALYs attributable to PMP at global, regional, and national levels from 1990 to 2021, distinguishing between disease burdens attributable to APMP and HPMP. It also provided an in‑depth analysis of the burden differences across various diseases, genders, and age groups. However, this study has several limitations. First, the GBD study estimated PMP exposure based on population‑weighted annual mean concentrations, assuming spatial uniformity of PMP components, which vary by source and season, potentially affecting the effect size of PMP [4]. Second, the GBD study estimated PMP levels based on the proportion of households using solid fuels for cooking, combined with data from household and personal exposure measurement studies, but did not account for exposures from solid fuels used for heating, boiling water, or other residential tasks, nor exposures from burning liquid fuels (e.g., kerosene), potentially underestimating total exposure and disease burden in certain locations [4]. Third, due to a lack of universally accepted case‑finding methods and limited clinical healthcare access, particularly in low‑ and middle‑income countries, some diseases may be underdiagnosed [13]. Fourth, the role of PMP in other diseases, such as neurodegenerative diseases (e.g., Alzheimer’s) and chronic kidney disease, is still being researched, and these diseases are not currently included in GBD estimates [56, 57]. Fifth, given the high prevalence of comorbidities in the elderly population, the GBD study may overestimate the overall level of DALYs in this demographic [12].