Research in context
Air pollution is a complex mixture of gases and particles whose sources and composition vary spatially and temporally. Population-weighted annual mean concentrations of particle mass with aerodynamic diameter less than 2·5 μm (PM2·5) and tropospheric ozone are the two indicators used to quantify exposure to air pollution. PM2·5 is the most consistent and robust predictor of mortality in studies of long-term exposure.7,
Ozone, a gas produced via atmospheric reactions of precursor emissions, is associated with respiratory disease independent of PM2·5 exposure.
We estimated exposure to PM2·5 and ozone for the global population at spatial scales relevant to human exposure (appendix pp 2–4).
Global annual mean exposure to PM2·5 was estimated in 5-year intervals from 1990 to 2015, at 0·1 × 0·1° (∼11 km × 11 km at the equator) resolution using estimates from satellites combined with a chemical transport model, surface measurements, and geographical data. We aggregated gridded exposure concentrations to national-level population-weighted means using the corresponding grid cell population value.
National-level population-weighted mean concentrations and the 95% uncertainty interval (95% UI) around this mean were estimated by sampling 1000 draws of each grid cell value and its uncertainty distribution.
As in previous assessments,
we used a chemical transport model to calculate a running 3-month mean (of daily 1 h maximum values) ozone concentration for each grid cell over 1 year, from which we selected the maximum of these values, consistent with epidemiological studies that use a seasonal (summer) mean, while accounting for global variation in the timing of the ozone (summer) season. We estimated population-weighted mean ozone concentrations and 95% UIs for each country as described for PM2·5, assuming a normal distribution with a 95% UI within 6% either side of the estimated mean concentration.
Theoretical minimum risk exposure level
TMREL was assigned a uniform distribution of 2·4–5·9 μg/m3 for PM2·5 and 33·3–41·9 parts per billion for ozone, bounded by the minimum and fifth percentiles of exposure distributions from outdoor air pollution cohort studies (appendix pp 7, 11–14). The uniform distribution represents the uncertainty regarding adverse effects of low-level exposure.
We estimated the burden attributable to PM2·5 for ischaemic heart disease (IHD), cerebrovascular disease (ischaemic stroke and haemorrhagic stroke), lung cancer, chronic obstructive pulmonary disease (COPD), and lower respiratory infections (LRI), and the burden attributable to ozone for COPD.
Evidence linking these diseases with exposure to ambient air pollution was judged to be consistent with a causal relationship on the basis of criteria specified for GBD risk factors.
We developed integrated exposure–response functions (IERs) for each cause of death to estimate the relative risk of mortality over the entire global range of ambient annual mean PM2·5 concentrations using risk estimates from studies of ambient air pollution, household air pollution, and second-hand smoke exposure and active smoking (appendix pp 8–14).
IERs assign concentrations of PM2·5 to each type of exposure on an equivalent μg/m3 basis assuming that risk is determined by the 24-h PM2·5 inhaled dose regardless of the exposure source, consistent with previous findings.
We updated IERs from those used in GBD 2013 by adding additional risk estimates for air pollution (appendix pp 11–14) and active smoking.
An alternative method to estimate exposure to second-hand smoke was used that incorporated estimates of PM2·5 attributable to exposure per cigarette, breathing rate, and number of cigarettes smoked in the country where each study was done. Further details are provided in the appendix (pp 8–14).
The IER has the mathematical form:
where z is the level of PM2·5 and zcf is the TMREL, below which no additional risk is assumed, with
if z is greater than zcf and zero otherwise. Here, 1 + α is the maximum risk, β is the ratio of the IER at low to high concentrations, and γis the power of PM2·5 concentration.
Epidemiological evidence suggests that the relative risks for IHD and stroke decline with age.
We modified the particulate matter source-specific relative risk for both IHD and stroke mortality as described by Burnett and colleagues
and applied this age modification to the relative risks, fitting the IER model for each age group separately.
Estimation of PAF and burden
Role of the funding source
Figure 1 shows IERs for the five causes of death. The functions are all non-linear, with a greater change in relative risk for lower concentrations compared with higher values. We fit age-specific functions for IHD and cerebrovascular disease, and estimated decreasing relative risks as age increased from 25 years to 80 years.
Global population-weighted PM2·5 increased by 11·2% from 1990 (39·7 μg/m3) to 2015 (44·2 μg/m3), increasing most rapidly from 2010 to 2015 (figure 2). Among the world’s ten most populous countries, exposures since 2010 increased in Bangladesh and India and were stable but remained high in Pakistan and China. Exposures decreased substantially in Nigeria and were low and slightly decreased in the USA, Brazil, and Russia. Population-weighted concentrations were low and stable in Japan and Indonesia.
Long-term exposure to PM2·5 contributed to 4·2 million (95% UI 3·7 million to 4·8 million) deaths and to a loss of 103·1 million (90·8 million to 115·1 million) DALYs in 2015, representing 7·6% of total global deaths and 4·2% of global DALYs, which is an increase from 1990. In 2015, ambient PM2·5 was the fifth-ranked risk factor for global deaths and sixth-ranked risk factor for DALYs among the risk factors included in GBD 2015 (figure 3). DALYs attributable to long-term exposure to PM2·5 consisted of 99·2 million (95% UI 87·7 million to 111·0 million) years of life lost and 3·9 million (2·6 million to 5·2 million) years lived with disability in 2015.
Mortality from cardiovascular disease (IHD and cerebrovascular disease) accounted for most deaths and DALYs attributable to ambient PM2·5 air pollution (figure 4; table 1). Ambient PM2·5 air pollution contributed to 17·1% of IHD, 14·2% of cerebrovascular disease, 16·5% of lung cancer, 24·7% of LRI, and 27·1% of COPD mortality in 2015 according to GBD compare.
|Deaths, in thousands (95% UI)||Age-standardised deaths per 100 000 people (95% UI)||DALYs, in thousands (95% UI)||Age-standardised DALYs per 100 000 people (95% UI)|
|All causes||4241·1 (3698·0–4776·7)||66·0 (57·2–74·8)||103 066·2 (90 829·6–115 072·6)||1490·9 (1312·4–1665·6)|
|Lower respiratory infection||675·0 (491·9–889·0)||10·1 (7·4–13·4)||28 359·9 (21 141·8–35 796·9)||390·9 (290·9–494·3)|
|Lung cancer||283·3 (178·4–398·7)||4·4 (2·7–6·1)||6209·1 (3934·9–8689·3)||90·9 (57·5–127·3)|
|Ischaemic heart disease||1521·1 (1231·7–1821·2)||23·6 (18·9–28·5)||32 406·0 (27 078·2–37 427·4)||470·7 (394·6–543·0)|
|Cerebrovascular disease||898·1 (717·6–1083·6)||14·0 (11·0–17·1)||19 242·8 (16 095·9–22 679·7)||281·2 (234·4–331·4)|
|Chronic obstructive pulmonary disease||863·6 (538·5–1212·8)||14·0 (8·7–19·6)||16 848·2 (10 517·4–23 590·0)||257·2 (160·3–360·6)|
|Male||2455·4 (2140·2–2752·9)||83·9 (72·5–94·7)||62 894·7 (55 545·7–70 098·2)||1888·8 (1659·4–2113·6)|
|Female||1785·7 (1546·2–2049·2)||50·8 (44·0–58·4)||40 171·5 (35 205·5–45 382·8)||1127·4 (986·6–1275·4)|
|Children <5 years||202·6 (152·7–254·6)||30·1 (22·7–37·8)||17 431·1 (13 139·7–21 906·3)||2585·9 (1949·1–3249·5)|
|Elderly >70 years||2228·3 (1842·0–2653·9)||562·7 (465·1–670·8)||25 073·0 (20 775·2–29 511·1)||6302·2 (5226·3–7419·8)|
Age-standardised death and DALY rates due to exposure to PM2·5 were higher in males than females (table 1), as a result of higher all-cause mortality rates in males (1018·6 per 100 000 males vs 703·4 per 100 000 females
). They were also higher in elderly people (age >70 years) than in children (age <5 years; table 1), mainly because of age-related differences in mortality from non-communicable diseases (41·4 per 100 000 children aged 1–5 years vs 2914·4 per 100 000 adults aged 70–74 years
). Ambient PM2·5contributed to 202 000 (95% UI 152 700–254 600) deaths and 17·4 million (13·1 million to 21·9 million) DALYs from LRI in children younger than 5 years.
Deaths attributable to long-term exposure to PM2·5 in 2015 varied substantially among countries (figure 5). South and east Asia contributed 59% of the 4·2 million global deaths attributable to ambient PM2·5 in 2015 (1·36 million deaths [95% UI 1·19 million to 1·56 million] in south Asia and 1·14 million deaths [0·97 million to 1·31 million] in east Asia). In World Bank high-income countries, exposure to ambient PM2·5 contributed to 4·3% of total deaths in 2015 versus 9·0% in upper-middle-income, 8·7% in lower-middle-income, and 4·9% in low-income countries. These differences in attributable mortality mostly reflect the fraction of total deaths from cardiovascular disease among countries.
The highest age-standardised rates of death due to PM2·5 exposure were in southern Asia (133·4 per 100 000 population, 95% UI 114·2–152·6), central Asia (85·2 per 100 000 population, 72·0–98·9), and eastern Asia (83·2 per 100 000 population, 70·4–95·6). Rates in high-income North American (USA, Canada, and Greenland; 17·8 per 100 000 people [95% UI 13·6–22·9]), Asian (18·7 per 100 000 people [14·6–23·7]), and western European countries (19·9 per 100 000 [15·9–24·8]) were four to eight times lower (appendix pp 26–1078).
Table 2 provides 2015 mortality and DALY estimates for the world’s ten most populous countries in 2015. Ambient PM2·5 ranked among the top ten risk factors for mortality in each of the world’s most populous countries. China and India combined had the largest numbers of attributable deaths and DALYs: 52% and 50% of the respective global totals. Pakistan, India, and Bangladesh had the highest age-adjusted mortality rates, more than seven times higher than those of Japan and the USA (table 2; appendix pp 26–1078).
|Deaths, in thousands (95% UI)||Risk factor rank for deaths||Deaths per 100 000 people (95% UI)||DALYs, in thousands (95% UI)||DALYs per 100 000 people (95% UI)||Population-weighted mean PM2·5 (μg/m3; 95% UI)|
|China||1108·1 (948·7–1272·8)||1||84·3 (71·5–96·7)||21 778·7 (18 903·5–24 584·2)||1478·6 (1275·9–1675·6)||58·4 (58·1–58·7)|
|India||1090·4 (936·6–1254·8)||2||133·5 (112·8–154·9)||29 609·6 (25 923·3–33 562·7)||2922·1 (2527·3–3327·5)||74·3 (73·9–74·8)|
|USA||88·4 (66·8–115·0)||6||18·5 (14·2–23·7)||1485·9 (1166·3–1841·7)||337·1 (265·0–416·8)||8·4 (8·4–8·5)|
|Indonesia||78·6 (62·0–96·7)||7||49·9 (38·5–61·6)||2185·0 (1730·4–2716·2)||1081·1 (860·4–1324·2)||15·4 (15·1–15·7)|
|Brazil||52·3 (41·9–65·1)||9||30·9 (24·2–39·0)||1083·9 (884·0–1322·7)||573·7 (467·3–702·3)||11·4 (11·2–11·5)|
|Pakistan||135·1 (114·3–159·2)||4||136·3 (113·7–163·5)||4217·3 (3545·1–4916·3)||3114·2 (2651·3–3657·7)||65·0 (63·8–66·2)|
|Nigeria||50·9 (35·7–73·2)||10||68·9 (48·5–101·7)||2410·0 (1640·4–3387·0)||1581·0 (1107·6–2237·2)||38·0 (37·5–38·5)|
|Bangladesh||122·4 (103·2–144·4)||5||133·2 (111·8–158·4)||3408·0 (2920·3–3945·8)||2972·0 (2533·4–3469·1)||89·4 (87·3–91·7)|
|Russia||136·9 (111·3–161·1)||3||62·6 (51·8–73·2)||2601·6 (2194·8–3007·2)||1255·0 (1077·8–1431·1)||16·6 (16·2–17·0)|
|Japan||60·6 (44·5–81·4)||8||16·8 (12·8–21·9)||705·8 (561·2–891·0)||261·7 (212·8–319·2)||13·3 (13·1–13·6)|
Trends in PM2·5-attributable mortality at the global and national levels reflect the influence not only of changing air quality, but also of demography and underlying mortality rates. We calculated the contribution of changes in each of four factors—population growth, population ageing, age-standardised rates of mortality (IHD, cerebrovascular disease, COPD, lung cancer, and LRI), and exposure to ambient PM2·5—to the net change in mortality attributable to ambient PM2·5 between 1990 and 2015 globally and for the ten most populous countries (appendix pp 22–23). Figure 6 shows the changes in mortality attributable to ambient PM2·5 from 1990 to 2015 according to the contributions of these four factors. Age-standardised mortality decreased in all ten countries, with Nigeria, Russia, Brazil, Indonesia, Pakistan, and the USA also experiencing decreases in exposure. These decreases were offset by increases in population growth and population ageing in most countries. Consequently, net increases in attributable mortality were noted in all countries except Nigeria and the USA. In China, India, Bangladesh, and Japan, increases in exposure combined with increases in population growth and ageing resulted in net increases in attributable mortality. In Brazil, Russia, Indonesia, and Pakistan, despite decreasing exposure, population growth (except in the case of Russia) and the ageing of the population led to a net increase in attributable mortality. In the USA, reductions in exposure offset increases in population and ageing, leading to a net decrease in attributable burden.
estimated disease burden and mortality rates at the subnational level for China, the USA, and several other countries. When these data were combined with PM2·5 exposures estimated globally at fine (0·1 × 0·1°) resolution, we were able to estimate subnational burden attributable to PM2·5 exposure. In China, where ambient PM2·5 contributed to 1·1 million (95% UI 1·0 million to 1·8 million) deaths in 2015, the provincial-level PM2·5-attributable age-standardised rates varied by more than three times, from 132·1 deaths per 100 000 people (95% UI 97·6–172·0) in Qinghai to 40·6 deaths per 100 000 people (30·2–50·4) in Hong Kong. In the USA, where ambient PM2·5 contributed to 88 400 (95% UI 66 800–115 000) deaths in 2015, state-level PM2·5-attributable age-standardised death rates also varied by about three times, from 27·1 deaths per 100 000 (95% UI 21·2–34·1) in Mississippi to 8·1 deaths per 100 000 (5·1–11·7) in Hawaii.
Exposure to ozone contributed to 254 000 (95% UI 97 000–422 000) deaths globally and a loss of 4·1 million (1·6 million to 6·8 million) DALYs from COPD in 2015. In 2015, ambient ozone was the 34th-ranked risk factor for global deaths and 42nd-ranked risk factor for DALYs among the 79 risk factors assessed in GBD 2015. Exposure to ozone contributed to an estimated 8·0% (95% UI 3·0–13.3) of global COPD mortality in 2015, with China, India, and the USA experiencing some of the highest mortality rates (figure 7A; appendix pp 1079–1362). The ozone-attributable COPD mortality rate increased in many countries from 1990 to 2015. Global deaths and DALYs attributable to ozone exposure increased from 1990 to 2015, as a result of increases in both levels of ozone and COPD mortality (figure 7B; appendix pp 1079–1362).
Absolute numbers of attributable deaths and DALYs were higher in GBD 2015
than estimated in GBD 2013.
These differences are mainly a result of changes in the underlying disease burden estimates
and to updates to the IER (appendix pp 8–14), which estimated higher relative risks in 2015 than in 2013 for IHD, cerebrovascular disease, LRI, COPD, and lung cancer. National-level population-weighted exposure estimates also increased between GBD 2013 and GBD 2015 (appendix pp 2–6). Because of the updated data and methods described earlier, we consider the current estimates to be more accurate.
Our results assume that the toxicity of ambient PM2·5 depends only on the magnitude of concentration, but not on the source, such as coal burning or vehicular emissions, or chemical composition, which vary among and within countries.
However, despite substantial effort, neither epidemiological nor toxicological research has identified particular sources or components that uniquely determine the toxicity of the PM2·5 mixture, and therefore the evidence does not support the development and application of source-specific relative risk functions for burden estimation.
This issue remains an active area of research and is a source of uncertainty in our estimates.
In the past few years, other researchers have estimated the burden of disease due to air pollution using different data and methods. Recent estimates from WHO
of 3·0 million deaths in 2012 used the same exposure estimates as presented here, but an earlier (GBD 2013) version of the IER and somewhat different baseline disease burden estimates. Lelieveld and colleagues
analysed source sector contributions to air pollution and the resulting disease burden in 2010 and estimated the burden in 2050. These estimates used an older (GBD 2010) IER. Furthermore, the coarse spatial resolution (∼100 × 100 km) of the exposure estimates introduced errors via spatial misalignment between exposure and population density compared with our estimates.
As in any assessment of this scope, this study has limitations. Since the GBD will be regularly updated, we anticipate enhancements to the methodology in the future to address them. First, we have probably underestimated the complete burden of disease attributable to air pollution. Although the causes of mortality we included make up four of the five leading global causes of death in 2015,
findings from systematic reviews in the past 10 years have shown that PM2·5 exposure is also associated with low birthweight and preterm birth,
and type 2 diabetes.
Future updates of GBD estimates will consider these other causes of mortality and morbidity should they meet GBD inclusion criteria.
Second, our estimate of the importance of ambient PM2·5 assumes that exposure does not affect the prevalence of other mortality risk factors. However, if long-term exposure to PM2·5 causes high blood pressure, then some amount of the PM2·5 burden would be mediated by its effect on high blood pressure. Mediation analysis was used in GBD 2015 to more accurately apportion the burden attributable to other risk factors such as diet and high blood pressure, but an absence of longitudinal studies precludes such analyses for ambient PM2·5.
Third, because large-scale cohort studies of PM2·5 and mortality are absent in the most polluted countries, the IERs were developed to estimate the effects of exposure at levels above those observed in air pollution cohort studies done in the USA, Canada, and western Europe, but the magnitude of the excess relative risk from PM2·5 exposure at high levels of PM2·5 remains uncertain. In Chinese cohort studies from the past 5 years, other metrics were used, such as total suspended particles and PM10,
and findings from a few analyses that converted these metrics to PM2·5 suggest that the IERs provide reasonable estimates of effects at high levels of ambient pollution.
Fourth, although we included estimates of the effect of seasonal ozone exposure on COPD mortality, less evidence is available for this relationship than that linking PM2·5 with COPD or the other causes of mortality. However, a causal link between increased COPD mortality and long-term exposure to ozone is, in our view, supported by a large body of evidence linking ozone exposure mortality to adverse effects on the respiratory system, including chronic changes in lung structure and function in human beings and non-human primates, and increased morbidity and mortality from COPD due to short-term and long-term exposure, especially in the warmer seasons.
In conclusion, ambient air pollution contributes substantially to the global burden of disease, which has increased over the past 25 years, as a result of both demographic and epidemiological trends and increasing levels of air pollution in low-income and middle-income countries. Should these trends continue, major reductions in pollution levels will be needed to avoid increases in disease burden. Moreover, the non-linear IERs imply modest reductions in burden in the most polluted countries unless PM2·5concentrations decline markedly.
As a result, the challenges for future reductions in the burden of disease attributable to air pollution are substantial. For example, using earlier attributable burden estimates and future mortality predictions, Apte and colleagues
estimated that air pollution levels in 2030 in China would need to decline by 29%, and those in India by 20%, to maintain per-person mortality at 2010 levels, although the economic
and public health benefits of even incremental reductions would probably be substantial in view of the large populations affected.
Exposure to ambient air pollution and its associated burden of disease can potentially be lowered for entire populations via policy action at the national and subnational levels. As the experience in the USA suggests,
changes in ambient PM2·5 associated with aggressive air quality management programmes, focused on major sources
of air pollution including coal combustion, household burning of solid fuels, and road transport, can lead to increased life expectancy over short timeframes.
2009 final report: integrated science assessment for particulate matter.
(accessed Dec 22, 2015).
Air quality guidelines—global update.
(accessed Dec 22, 2015).
Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015.
Lancet. 2016; 388: 1659-1724
Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks in 188 countries, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013.
Lancet. 2015; 386: 2287-2323
Ambient air pollution exposure estimation for the Global Burden of Disease 2013.
Environ Sci Technol. 2016; 50: 79-88
Data integration model for air quality: a hierarchical approach to the global estimation of exposures to ambient air pollution.
ArXiv1609.00141 Stat. 2016; (published online Sept 1.) (accessed Nov 4, 2016)
A systematic review of the relation between long-term exposure to ambient air pollution and chronic diseases.
Rev Environ Health. 2008; 23: 243-297
Long-term air pollution exposure and cardio-respiratory mortality: a review.
Environ Health Glob Access Sci Source. 2013; 12: 43
- Integrated science assessment of ozone and related photochemical oxidants (second external review draft).US Environmental Protection Agency, Washington, DC; 2011
(accessed March 3, 2017).
Long-term ozone exposure and mortality.
N Engl J Med. 2009; 360: 1085-1095
- Gridded population of the world, version 4 (GPWv4): population count.NASA Socioeconomic Data and Applications Center (SEDAC), Palisades, NY; 2016
(accessed March 16, 2017).
An integrated risk function for estimating the global burden of disease attributable to ambient fine particulate matter exposure.
Environ Health Perspect. 2014; 122: 397-403
A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010.
Lancet. 2012; 380: 2224-2260
Meta-analysis methods to estimate the shape and uncertainty in the association between long-term exposure to ambient fine particulate matter and cause-specific mortality over the global concentration range.
Risk Anal Off Publ Soc Risk Anal. 2016; 36: 1813-1825
National Particle Component Toxicity (NPACT) Initiative: integrated epidemiologic and toxicologic studies of the health effects of particulate matter components.
Res Rep Health Eff Inst. 2013; 177: 5-13
Attributing health effects to apportioned components and sources of particulate matter: an evaluation of collective results.
Atmos Environ. 2011; 45: 5655-5663
Mortality in relation to smoking: 40 years’ observations on male British doctors.
BMJ. 1994; 309: 901-911
The age-specific quantitative effects of metabolic risk factors on cardiovascular diseases and diabetes: a pooled analysis.
PLoS One. 2013; 8: e65174
Global, regional, and national life expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980–2015: a systematic analysis for the Global Burden of Disease Study 2015.
Lancet. 2016; 388: 1459-1544
Global, regional, and national disability-adjusted life-years (DALYs) for 315 diseases and injuries and healthy life expectancy (HALE), 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015.
Lancet. 2016; 388: 1603-1658
(accessed March 1, 2017).
Ischemic heart disease mortality and long-term exposure to source-related components of U.S. fine particle air pollution.
Environ Health Perspect. 2016; 124: 785-794
Burden of disease attributable to coal-burning and other air pollution sources in China.Date: Aug 20, 2016
(accessed Nov 2, 2016).
WHO global urban ambient air pollution database (update 2016).
(accessed March 3, 2017).
The contribution of outdoor air pollution sources to premature mortality on a global scale.
Nature. 2015; 525: 367-371
Ambient air pollution, birth weight and preterm birth: a systematic review and meta-analysis.
Environ Res. 2012; 117: 100-111
Effect of long-term exposure to air pollution on type 2 diabetes mellitus risk: a systemic review and meta-analysis of cohort studies.
Eur J Endocrinol. 2014; 171: R173-R182
Association between long-term exposure to outdoor air pollution and mortality in China: a cohort study.
J Hazard Mater. 2011; 186: 1594-1600
Particulate air pollution and mortality in a cohort of Chinese men.
Environ Pollut Barking Essex 1987. 2014; 186: 1-6
Air pollution and life expectancy in China and beyond.
Proc Natl Acad Sci USA. 2013; 110: 12861-12862
Long-term exposure to ambient ozone and mortality: a quantitative systematic review and meta-analysis of evidence from cohort studies.
BMJ Open. 2016; 6: e009493
Long-term ozone exposure and mortality in a large prospective study.
Am J Respir Crit Care Med. 2016; 193: 1134-1142
Addressing global mortality from ambient PM2·5.
Environ Sci Technol. 2015; 49: 8057-8066
Health benefits of air pollution abatement policy: role of the shape of the concentration–response function.
J Air Waste Manag Assoc 1995. 2015; 65: 516-522
- The cost of air pollution: strengthening the economic case for action.World Bank, Washington, DC; 2016
(accessed March 3, 2017).
Effect of air pollution control on life expectancy in the United States.
Epidemiology. 2013; 24: 23-31
Figure 1Integrated exposure–response functions
Figure 2Trends in population-weighted mean concentrations of particle mass with aerodynamic diameter less than 2·5 μm
Figure 3Leading level 3 Global Burden of Diseases global risk factors for deaths (A) and disability-adjusted life-years (B), 1990 and 2015
Figure 4Deaths attributable to ambient particulate matter pollution by year and cause
Figure 5Deaths attributable to ambient particulate matter pollution in 2015
Figure 6Changes in mortality attributable to ambient particulate matter pollution according to population-level determinants by country from 1990 to 2015
Figure 7Proportion of deaths attributable to ozone (A) in 2015 and percentage change from 1990 (B)
- Table 1Global deaths, disability-adjusted life-years, and age-standardised rates attributable to ambient particulate matter pollution in 2015
- Table 22015 estimates of mortality and disability-adjusted life-years attributable to ambient particulate matter pollution and population-weighted mean particulate matter pollution in the world’s ten most populous countries