International Journal of Public Health

, Volume 62, Issue 7, pp 729–738 | Cite as

Short-term effects of fine particulate matter pollution on daily health events in Latin America: a systematic review and meta-analysis

  • Laís Fajersztajn
  • Paulo Saldiva
  • Luiz Alberto Amador Pereira
  • Victor Figueiredo Leite
  • Anna Maria Buehler
Review

Abstract

Objectives

Ambient air pollution is among the leading risks for health worldwide and by 2050 will largely overcome deaths due to unsafe sanitation and malaria, but local evidence from Latin America (LA) is scarce. We aimed to summarize the effect of short-term exposure to fine particulate air pollution (PM2.5) on morbidity and mortality in Latin America and evaluate evidence coverage and quality, using systematic review and meta-analysis.

Methods

The comprehensive search (six online databases and hand-searching) identified studies investigating the short-term associations between PM2.5 and daily health events in LA. Two reviewers independently accessed the internal validity of the studies and used random-effect models in the meta-analysis.

Results

We retrieved 1628 studies. Nine were elected for the qualitative analysis and seven for the quantitative analyses. Each 10 µg/m3 increments in daily PM2.5 concentrations was significantly associated with increased risk for respiratory and cardiovascular mortality in all-ages (polled RR = 1.02, 95% CI, 1.02–1.02 and RR = 1.01, 95% CI , 1.01–1.02, respectively).

Conclusions

Short-term exposure to PM2.5 in LA is significantly associated with increased risk for respiratory and cardiovascular mortality. Evidence is concentrated in few cities and some presented high risk of bias.

Keywords

Air pollution Particulate matter Fine particulate matter PM2.5 Mortality Latin America Systematic review and meta-analysis 

Introduction

Ambient air pollution is among the leading risks for health worldwide (GBD 2015 Risk Factors Collaborators 2016), and by 2050 will largely overcome deaths due to unsafe sanitation and malaria (Organisation for Economic Co-operation and Development 2012).

Particulate matter (PM) is a complex mixture of solid and liquid compounds widely used as an indicator for air pollution. Its composition and size vary substantially according to emission sources and prevailing weather conditions. The smaller its size the greater its potential to impact health, given the increased chance of penetration in the respiratory tract (Health Effects Institute 2002). Epidemiological evidence of the effects of fine particulate matter [PM with aerodynamic diameter ≤ 2.5 μm (PM2.5)] on morbidity and mortality is consistent, particularly on cardiovascular and respiratory systems (Anderson et al. 2004; Pope et al. 2006). Biological evidence supporting plausible mechanisms is clear and includes pulmonary oxidative stress, inflammation and altered cardiac autonomic function (Pope et al. 2006). However, evidence in Latin America (LA) is scarce and systematic reviews and meta-analysis (SRMA) on the effects of PM2.5 on daily health events rarely pool evidence from LA (Adar et al. 2014; Shah et al. 2015). The World Health Organization (WHO) is updating current ambient Air Quality Guidelines, which establishes standards for pollutants at concentrations considered sufficiently safe for human health (World Health Organization 2015). Transferability of existing concentration response functions (CRF) developed for European and North American cities to other countries with significant differences in air pollution levels are being questioned. According to the experts consulted by WHO, re-evaluation of the existing CRFs for PM2.5 is necessary, given the observed changes in the risk at extreme concentrations. Experts also underscored that current evidence shows association of short-term PM2.5 exposure with mortality at levels below the current guideline, among other new evidences that need to be evaluated (World Health Organization 2015).

When the WHO air quality guidelines were last updated in 2005 (World Health Organization 2006), monitoring systems were scarce in cities from developing nations, including LA’s cities. Most of the data available in developing regions were exclusively for total suspended particles concentrations, an overcome measure of particulate air pollution that includes particulates of bigger size. After a decade, evidence in LA remains delayed compared to North America and Europe. While developed countries are concerned with the health effects of fine (PM2.5), ultrafine (PM with aerodynamic less than 100 nm) and coarse [PM with aerodynamic diameter 10–2.5 (PM10−2.5)] particles, few cities in LA depict evidence for these pollutants. Only eight Latin American countries have set National Air Quality Standards for PM2.5 (Green and Sanchéz 2012), all above the final target of 25 µg/m3 for 24-hour average, current suggested by WHO (World Health Organization 2006). In 2012, the ESCALA multi-city study has established methodological foundation for future research in LA and provided high-quality evidence from nine cities to support air pollution control policies in the region (Romieu et al. 2012). However, the evidence was limited to the effects of PM10 on mortality.

Given the very high levels of PM2.5 pollution in Africa, Asia and Middle East (Brauer et al. 2012), air pollution in LA receives less attention in global analysis. However, between 1990 and 2013, PM2.5 annual mean level trends increased in parts of South America and decreased in parts of Southeast Asia (Brauer et al. 2012).

In this systematic review, we aimed to summarize the evidence of the short-term effects of PM2.5 on daily health events in Latin American populations. We also aimed to identify the Latin American cities covered with local evidence on short-term health effects of PM2.5 and to evaluate the quality of the evidence identified.

Methods

This study followed a protocol previously registered at PROSPERO (registration CRD42015029673) and is reported according to both: PRISMA Statement recommendations for transparent reporting of systematic reviews (Moher et al. 2009) and MOOSE Guidelines for meta-analyses and systematic reviews of observational studies (Stroup 2000).

Search strategy and eligibility criteria

We searched six online databases (PubMed, Embase, Scopus, Web of Science, Cochrane and Lilacs) up to September 2015 to identify studies conducted in LA that analysed the associations of short-term exposure to PM2.5 with daily health events (mortality, hospital admission and emergency room visits). Search combined terms for exposure (e.g. “air pollution”, or “particulate matter”), outcome (e.g. “hospital admission”) and population (e.g. “Latin America” or name of countries). Controlled vocabulary of databases was included in the search string. We did not impose language or date restrictions to the search. We searched the reference lists of relevant reviews and studies for additional manuscripts and used Mendeley software for uploading the references. For detailed search strategy, see the Supplemental Material, S1.

Eligibility criteria included

  1. (a)

    Original published time-series or case cross-over studies with a minimum period evaluated of one uninterrupted year, conducted in LA.

     
  2. (b)

    Studies that reported quantitative measurements of daily short-term exposure (up to a lag of 7 days) to ambient PM2.5.

     
  3. (c)

    Studies that reported daily emergency room visits, hospital admissions, and/or mortality for all natural causes and/or all or selected respiratory causes and or all or cardiovascular causes. Related International Code of Disease (ICD 9th and 10th revision) for the causes included were ICD-9 001–799, ICD-10 A00-T98 and Z00-Z99 (all natural causes), ICD-9 460–519, ICD-10 J00-J99 (respiratory causes) and, ICD-9 390–459, ICD-10 I00-I99 (cardiovascular cases).

     
  4. (d)

    Studies that calculated or provided enough information to estimate the relative risk of the daily health event associated with an increment of 10 µg/m3 in PM2.5 levels.

     
  5. (e)

    Studies that reported the risk exclusively by subgroups other than age, daily event or disease groups of interest for this review (e.g. by season or chemical composition of the particle) were excluded.

     

Study selection

First, two independent reviewers screened all abstracts and titles for studies that potentially met the inclusion criteria. Then, reviewers evaluated full studies for final inclusion in the SR (Systematic Review) and recorded reasons for exclusion (see the Supplemental Material). Similar to reviews on global (Atkinson et al. 2014) and European evidence (Anderson et al. 2004), we selected only one study by setting to avoid overlapping of populations (see protocol at PROSPERO, registration CRD42015029673). We solved disagreements between reviewers by consensus along the stages.

Data collection process and items

Data items included citation information (e.g. year of publication) and characteristics of the location (e.g. main air pollution sources), population (e.g. age), exposure (e.g. PM2.5 levels) and outcomes (e.g. total number of events) (see Table 1). We also collected information necessary to conduct meta-analysis of the relative risks related with 10 µg/m3 increments in PM2.5 levels (e.g. association measurements and its 95% Confidence Interval (CI)) and methodological characteristics necessary to assess the quality of the included studies (see Table 2). One reviewer extracted the data into Microsoft Excel 2013 and a second reviewer confirmed the information. Disagreements were solved by discussion.

Table 1

Characteristics of the studies included in the systematic review of the effects of fine particulate matter on daily health events in Latin America

References

Study period

City (Country)

Air pollution sources

Inhabitants reported

PM2.5 mean (SD) in µg/m3

Population in years

Outcome (cause)

Number of events

Pollutant model

Risk of bias

Dales et al. (2010)ŧ

1998–2005

Santiago metropolitan region (Chile)

NI

5.370.000

32.9 (20.2)

All-ages

HA(CV :I00-I99)

282.645

Both

Low

Cesar et al. (2013)

2011–2012

Piracicaba (Brazil)

Biomass, automotive

350.000

26.8 (16.7)

<10

HA(R:J00-J98)

437

Single

High

Ignotti et al. (2010)

2000–2004

Alta Floresta (Brazil)

Biomass

101.278

20.4 (32.6)

<5

HA(R:J00-J98)

481

Single

High

Silva et al. (2013)

2005

Cuiabá (Brazil)

Biomass

550.562

7.5 (10.4)

<5

HA(R:J00-J98)

1.152

Single

Low

Loomis et al. (1999)ŧ

1993–1995

Ciudad de México (Mexico)

Automotive, industrial

2.500.000

27.4(10.5)

<1

M(A:A00-T98, Z00-Z99)

2.798

Both

Low

Borja-Aburto et al. (1998)

1993–1995

Ciudad de México (Mexico)

Automotive, industrial

2.500.000

27.0 (11.0)

All-ages, >65

M(A,R:J00-J98),CV:I00-I99)

1.152

Both

Low

Sanhueza et al. (1998)

1988–1993

Santiago (Chile)

NI

NI

81.0 (NI)

All-ages, >65

M(A:A00-T98, Z00-Z99)

76.442 (A,>65)

Both

Low

Valdes et al. (2012)

1998–2007

Santiago (Chile)

NI

NI

34.0 (NI)

All-ages

M(R:J00-J98),CV:I00-I99)

92.891

Single

High

Reyna et al. (2012)

2003–2007

Mexicali (Mexico)

Automotive, unpaved roads

856.000

60.6 (32.3)

All-ages

M(A*:A00-T98, Z00-Z99)

872.350

Single

Low

SD standard deviation, ICD international code of disease 10th revision, NI not informed, PM particulate matter, M mortality, HA hospital admissions, E emergency room visits, A all non-external causes, CV cardiovascular disease, R respiratory disease, Single single-pollutant model, Multi multi-pollutant model, Both single and multi-pollutant model. Not all studies evaluated the whole population of the city investigated, therefore some inhabitants reported reflect the population of the area covered by the hospitals included in the analysis, not all the city

*Reyna et al. (2012) also evaluated respiratory mortality, but did not present the associations with short-term exposure to PM2.5 for this outcome. Study evaluated specific cardiovascular disease I26 an I80 according to the International Classification of Disease, 10th Revision (ICD-10). ŧNot included in the quantitative analysis. Dales et al. (2010) was the only study that evaluated cardiovascular hospitalizations and Loomis et al. (1999) the only study that evaluated mortality in children

Risk of bias in individual studies

Since there is no specific tool to evaluate methodological quality of time-series and case cross-over studies, two independent reviewers conducted a qualitative assessment, which classified studies as high or low risk of bias, according to criteria previously registered at the protocol of this study. We also verified funding sources and the conflicts of interest reported, which are recently being considered potential source of bias (Lam et al. 2014).

Summary measures

We calculated summary effects associated with PM2.5 levels in the statistical software RevMan 5.1 by daily event type (emergency room visits, hospital admissions and mortality), cause (total non-external causes, respiratory and cardiovascular causes) and age group (all-ages, elderly and children), if at least two association measurements were available from two different studies. Given that we calculated a polled effect for an entire continent, variations regarding population, outcomes and exposure characteristics among the cities included in the quantitative analysis were expected. For this reason, we used random effects model to estimate the summary measures and present them as Relative Risk (RR) with a 95% Confidence Interval (95% CI) for each increment of 10 μg/m3 in PM2.5 24-hour average levels.

The lag (number of days between the exposure and the health event) reported by studies varied greatly and it was not possible to group the studies by lag structure. The lag used to calculate the summary measure was the lag reported. We chose the lag with the highest effect, if more than one lag was reported.

We used Cochrane Q test (significance level: 0.1) and I2 statistic to test statistical heterogeneity. I2 statistic quantifies heterogeneity by calculating the proportion of variation occurred by heterogeneity rather than by chance (Higgins et al. 2003). For I2, we considered that values ranging from 0 to 30, 30 to 50 and >50 indicated low, moderate and high heterogeneity, respectively. We also reduced potential heterogeneity due to populations and outcome cause differences by polling evidence grouped by cause/age pairs. We planned to perform subgroup analysis by pollutant model (single vs. multi-pollutant), pollution source (rural vs. urban) and risk of bias between studies (combining low risk of bias studies only), if heterogeneity was statistically significant. Non-quantitative data were presented descriptively.

Publication bias assessment

If 10 or more studies are included in the review, we accessed publication bias through funnel plot and Egger test (Egger et al. 1997).

Results

Study selection

Search retrieved 1628 studies, nine were elected for the qualitative analysis and seven for the quantitative analyses (Fig. 1). Elected studies were conducted in three countries (Chile, Brazil and Mexico), distributed in six cities, mostly with less than a million inhabitants. None of the studies investigated the effects of PM2.5 on emergency room visits.

Fig. 1

Flow diagram describing search, screening and eligibility of the studies on short-term effects of fine particulate matter on daily events in Latin America

Automotive was the main source of air pollution reported, followed by biomass burning and industrial sources. PM2.5 daily mean varied from 7.5 to 81.0 µg/m3 among the included studies (Table 1).

The most studied decade was the 2000s (44%), followed by the 1990s (35%), 2010s (23%) and 1980s (5%). All elected studies were time-series studies published as articles. Only one was a multi-city study (evaluated the metropolitan area). Elected studies were mostly published in English (89%), and 56% of the included studies used single-pollutant models exclusively (see Supplemental Material, TableS1).

The quality assessment revealed that 33% of the included studies presented high risk of bias, 43% if we consider only studies elected for the quantitative analysis. Frailties in statistical adjustments (e.g. not using the Akaike criteria), in exposure characterization (not using monitoring networks) and controlling for confounders (not controlling for short trends such as for day of the week) were the main reasons why studies were classified as high risk of bias. When reported, conflict of interest was absent and financial sources were related to governmental agencies and University-related grants (Table 2).

Table 2

Quality assessment of the included studies in the systematic review of the effects of fine particulate matter on daily health events in Latin America

 

Selection

Exposure

Confounders

Statistical adjustments of the model

 

Reference

ICD

Source

Source◊◊

Network

Daily

Long trends

Short trends

Temp

Hum

Autocorrelation

Akaike

Risk of bias**

Founding source

Conflict of interest

Dales et al. (2010)ŧ

Y

Y

N

Y

Y

N

N

Y

Y

Y

Y

Low

NI

No conflict

Cesar et al. (2013)

Y

Y

Y

N*

Y

N

N

Y

N

N

N

High

Governmental research agency

No conflict

Ignotti et al. (2010)

Y

Y

Y

N*

Y

N

N

Y

Y

N

N

High

Governmental research agencies

NI

Silva et al. (2013)

Y

Y

Y

N*

Y

Y

Y

Y

Y

N

N

Low

NI

No conflict

Loomis et al. (1999)ŧ

Y

Y

Y

N

Y

Y

N

Y

N

Y

Y

Low

NI

No conflict

Borja-Aburto et al. (1998)

Y

Y

Y

N

Y

Y

N

Y

Y

Y

Y

Low

NI

NI

Sanhueza et al. (1998)

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

N

Low

Governmental research agency and University-related grants

NI

Valdes et al. (2012)

Y

Y

Y

N

N

Y

Y

Y

Y

N

N

High

NI

No conflict

Reyna et al. (2012)

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

N

Low

University-related grants

NI

ICD International code of disease informed, Y yes, N no, U unclear, NI not informed, Network monitoring network, Daily daily measurements, Akaike Akaike criteria, Temp temperature, Hum humidity

Studies included in the quantitative analysis, N* satellite, Source of the outcome data, ◊◊Source of the exposure data, ŧnot included in the quantitative analysis. Dales et al. (2010) was the only study that evaluated cardiovascular hospitalizations and Loomis et al. (1999). **Studies > 3 N were classified as high risk of bias.

Measurements combined for meta-analysis were derived from single-pollutant models, except for total mortality in elderly, where evidence was available for single- and multi-pollutant models. All meta-analysis combined evidence from two different studies, except from hospital admissions where evidence from three were combined. Given the low number of studies combined and their characteristics, we did not perform subgroup analysis by pollution source (rural vs. urban) and risk of bias (low vs. high risk of bias).

Mortality

Meta-analysis of the risks for respiratory mortality in all-ages associated with PM2.5 levels showed a significant 2% increased risk (RR per 10 µg/m3 = 1.02, 95% CI, 1.02–1.02; I2 = 0%) (Fig. 2). Meta-analysis of the risk for cardiovascular mortality associated with PM2.5 levels showed a significant 1% increased risk (RR per 10 µg/m3 = 1.01, 95% CI, 1.01–1.02; I2 = 0%) (Fig. 2). One of the studies combined presented high risk of bias.

Fig. 2

Forest plots for the risks for daily mortality by cause associated with short-term exposure to PM2.5 levels in Latin America

Risk for total mortality was not significant neither for all-ages (RR per 10 µg/m3 = 1.01, 95% CI, 1.00–1.01; I2 = 0%) (Fig. 2), nor for elderly (RR per 10 µg/m3 = 1.01, 95% CI, 1.00–1.02; I2 = 55%) The risk for total mortality in elderly remained not significant when we combined measures based on multi-pollutant models, all adjusted for O3: RR per 10 µg/m3 = 1.01, 95% CI, 0.99–1.03; I2 = 73% (see Supplemental Material, Figure S1). All studies combined presented low risk of bias.

For hospital admissions, meta-analysis showed no statistical association between PM2.5 levels and daily hospital admissions and presented significant high heterogeneity: RR per 10 µg/m3 = 1.03, 95% CI 0.99–1.08; I2 = 68%) (Fig. 3). The studies combined used satellite-derived measurements and two depicted high risk of bias.

Fig. 3

Forest plots for the risks for daily respiratory hospitalization in children associated with short-term exposure to PM2.5 levels by pollutant model in Latin America

Discussion

Our results showed significant associations between short-term exposure to PM2.5 in LA and increased risk for respiratory and cardiovascular mortality in all-ages. The risks for each 10 µg/m3 increment in PM2.5 levels ranged between 1 and 2% and heterogeneity between studies was low. On the other hand, risks associated with short-term exposure to PM2.5 for total mortality in all-ages and in elderly and respiratory hospitalization in children were not significant. Imprecision might explain why some meta-analysis were not statistically significant in our review, since CIs of the combined studies tended to be larger in the non-significant meta-analysis (particularly for hospitalizations meta-analysis), compared to significant meta-analysis (respiratory and cardiovascular mortality). Frequently, multi-city studies find larger effects when the analysis is restricted to elderly populations (Romieu et al. 2012), because elderly are more vulnerable to air pollution adverse health effects. However, our meta-analysis for total mortality in elderly was not significant. Because populations are exposed to a complex mixture of air pollutants, not single pollutants, researchers (and policy makers) have been encouraged to move towards a multi-pollutant approach to air quality, rather than single-pollutant models (Dominici et al. 2012). In our review, most of the risks reported were based on single-pollutant models. Meta-analysis for multi-pollutant models (adjusted for O3) was just feasible for total mortality in elderly, but the related risks were not significant in both pollutant models.

Our results are partially in accordance with other meta-analysis on the short-term association of PM2.5 and daily health events in the world (Adar et al. 2014; Atinkson et al. 2014), pointing out that, while local evidence is scarce, extrapolating evidence from other regions to control particulate pollution in LA might be reasonable.

Compared to the present review, Atkinson and colleagues study (2014) identified fewer Latin American studies and did not evaluate all the outcomes considered in this review for Latin American region, possible due to differences in the search strategy adopted. We searched Lilacs, a specific database for scientific literature of Latin American, and other databases up to September 2015, while Atkinson’s and colleagues review (2014) searched global databases exclusively up to May 2011. While our results are mostly in accordance with Atkinson’s global findings, our results contrast with Atkinson’s findings for LA for total, respiratory and cardiovascular mortality in all-ages. Both reviews combined evidence from two studies to perform meta-analysis, but in our review, CIs of the included studies were more precise.

Evidence on the short-term effects of PM2.5 in daily health events conducted in LA is scarce, concentrated in few locations and publication trend is not increasing. The scarcity of studies in the continent might be reflecting lack of exposure data, since only half of Latin American cities with monitoring stations for PM10, also monitors PM2.5 (Green and Sanchéz 2012). Moreover, in some cities, PM2.5 measurements started in recent years and occur in a few percentage of the existing monitoring stations (Ambiente 2014). In our review, all Brazilian studies included estimated population exposure through satellite estimations. Sophisticated satellite-based estimates are a promising tool to overcome the scarcity of exposure data, given that correlation with ground-level PM2.5 mass might reach 81% (Van Donkelaar et al. 2016) and The Global Burden of Disease Study is already using satellite-based PM2.5 estimates (Brauer et al. 2012).

One of the strengths of our review was to evaluate the internal validity of the included studies, which is not frequent in most SRMAs of epidemiological time-series studies, since, to our knowledge, no validated tool evaluates methodological quality of time-series. Despite advances in methodology of SRMAs involving epidemiological observational studies in general (Atkinson et al. 2014)1 and of epidemiological studies of environmental health (EH) in particular (Sheehan and Lan 2015), it is still difficult to adapt the recommendations to assess the risk of bias in time-series studies. Cochrane Collaboration recommend tools for randomized controlled trials and non-randomized studies of intervention (Higgins and Green 2011), both not applicable to time-series design. Yet, time-series studies have been playing a relevant role in air pollution regulatory process, because they estimate the burden of disease attributable to air pollution exposure and the related CRF can be used in cost–benefit analysis (Bell et al. 2004). Time-series studies use regression models to assess the effects of short-term changes in PM2.5 levels on acute health effects by estimating associations between day-to-day variations in both air pollution and in mortality and morbidity counts. The study design uses aggregated level data for exposure and outcome. Confounding bias arise from factors that vary in short time scales and may be associated with particulate and health (e.g. temperature). Statistical specificities (e.g. accounting for serial correlation in the residuals) also play a role (Bell et al. 2004).

Our quality assessment followed an a priori criteria and two reviewers independently assessed key methodologically potential sources of bias in human epidemiological observational studies in EH in general (e.g. exposure characterization and controlling for confounders) (Sheehan and Lan 2015) and of time-series design in particular (e.g. statistical adjustments) (Bell et al. 2004). Frailties in at least one of the potential source of bias domains were frequent; underlining that internal validity of time-series study design on EH SRMAs needs further investigations.

Since we retrieved less than 10 studies for each health endpoints (outcome/cause/age), we did not evaluate publication bias, a limitation of this review.

Multi-city studies standardize statistical approach, and therefore minimize potential risk of bias due to differences in statistical methodology of time-series studies (Bell et al. 2004). To the extent that time-series design usually uses secondary dada relatively accessible from public sources, efforts to conduct high-quality multi-city studies might be preferable than conducting meta-analysis based on individual studies data retrieved in a SR. However, PM2.5 monitoring stations are scarce in LA (Green and Sánchez 2012) and the multi-city study for Latin American region8 did not evaluate the effects of PM2.5 on health, making this review relevant.

In summary, evidence in LA is scarce, concentrated and trend of publication is not increasing. Short-term exposure to PM2.5 in the continent is significantly associated with increased risk for respiratory and cardiovascular mortality in all-ages in region. Our results are in accordance with evidence from other parts of the world, pointing out that, while local evidence is scarce, extrapolating evidence from other regions to control particulate pollution in LA might be reasonable. Quality assessment identified frailties in key methodological sources of bias in observational studies in EH in general (exposure characterization and controlling for confounders) and in time-series design in particular (statistical adjustments), thus assessing the internal validity of time-series study design in SRMA on EH needs further investigations.

Notes

Compliance with ethical standards

Ethical statement

The present manuscript is an original work, has not been previously published whole or in part, and is not under consideration for publication elsewhere. We did not fabricate data. We did not present data, text or theories from other authors as if they are ours (no “plagiarism”). All authors contributed sufficiently to the scientific work, read the manuscript, agreed the work is ready for submission to a journal and accepted responsibility for the manuscript’s content. All authors do not present actual or potential conflicts of interest regarding the submitted.

Supplementary material

38_2017_960_MOESM1_ESM.docx (316 kb)
Supplementary material 1 (DOCX 315 KB)

References

  1. Adar SD, Filigrana PA, Clements N, Peel JL (2014) Ambient coarse particulate matter and human health: A systematic review and meta-analysis Curr Environ Heal Reports 1:258–274; doi:10.1007/s40572-014-0022-z.CrossRefGoogle Scholar
  2. Anderson H, Atkinson R, Peacock J, Marston L, Konstantinou K (2004) Meta-analysis of time-series studies and panel studies of particulate matter (PM) and ozone (O3). Rep. a WHO Task Gr. 1–68.Google Scholar
  3. Atkinson RW, Kang S, Anderson HR, Mills IC, Walton HA (2014) Epidemiological time series studies of PM2.5 and daily mortality and hospital admissions: a systematic review and meta-analysis. Thorax 69:660–665. doi:10.1136/thoraxjnl-2013-204492 CrossRefPubMedPubMedCentralGoogle Scholar
  4. Bell ML, Samet JM, Dominici F (2004) Time-series studies of particulate matter. Annu Rev Public Health 25:247–280. doi:10.1146/annurev.publhealth.25.102802.124329 CrossRefPubMedGoogle Scholar
  5. Borja-Aburto VH, Castillejos M, Gold DR, Bierzwinski S, Loomis D (1998) Mortality and ambient fine particles in southwest Mexico City, 1993–1995. Environ Health Perspect 106:849–855CrossRefPubMedPubMedCentralGoogle Scholar
  6. Brauer M, Amann M, Burnett RT, Cohen A, Dentener F, Ezzati M et al (2012) Exposure assessment for estimation of the global burden of disease attributable to outdoor air pollution. Environ Sci Technol 46:652–660. doi:10.1021/es2025752 CrossRefPubMedPubMedCentralGoogle Scholar
  7. Cesar ACG, Nascimento LFC, De Carvalho JA, Gobbo Cesar AC, Nascimento LFC, de Carvalho Jr JÁ (2013) Association between exposure to particulate matter and hospital admissions for respiratory diseases in children. Rev Saude Publica 47:1209–1212. doi:10.1590/S0034-8910.2013047004713 CrossRefPubMedPubMedCentralGoogle Scholar
  8. Dales RE, Cakmak S, Vidal CB (2010) Air pollution and hospitalization for venous thromboembolic disease in Chile. J Thromb Haemost 8:669–674. doi:10.1111/j.1538-7836.2010.03760.x CrossRefPubMedGoogle Scholar
  9. Dominici F, Peng RD, Barr CD, Bell ML (2012) Single-pollutant to a multi-pollutant approach. Epidemiology 21:187–194. doi:10.1097/EDE.0b013e3181cc86e8.Protecting CrossRefGoogle Scholar
  10. Egger M, Davey Smith G, Schneider M, Minder C (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ 315:629–634. doi:10.1136/bmj.316.7129.469 CrossRefPubMedPubMedCentralGoogle Scholar
  11. GBD 2015 Risk Factors Collaborators (2016) 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 388(10053):1659–1724. doi:10.1016/S0140-6736(16)31679-8
  12. Green J, Sánchez S (2012) Air quality in Latin America: an overview. Clean air Inst. 1–28. doi:10.1017/CBO9781107415324.004
  13. Health Effects Institute (2002) Understanding the Health Effects of Components of the Particulate Matter Mix: Progress and Next StepsGoogle Scholar
  14. Higgins J, Green S (2011) Cochrane handbook for systematic reviews of interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration.Google Scholar
  15. Higgins JPT, Thompson SG, Deeks JJ, Altman DG (2003) Measuring inconsistency in meta-analyses BMJ. Br Med J 327:557–560. doi:10.1136/bmj.327.7414.557 CrossRefGoogle Scholar
  16. Ignotti E, Hacon SS, Junger WL, Mourão D, Longo K, FreitasS, et a. (2010) Air pollution and hospital admissions for respiratory diseases in the subequatorial Amazon: a time series approach Cad Saúde Pública 26(4):747–761; doi:10.1590/S0102-311X2010000400017
  17. Lam J, Koustas E, Sutton P, Johnson PI, Atchley DS, Sen S et al (2014) The Navigation Guide - evidence-based medicine meets environmental health: integration of animal and human evidence for PFOA effects on fetal growth. Environ Health Perspect 122:1040–1051. doi:10.1289/ehp.1307923 PubMedPubMedCentralGoogle Scholar
  18. Loomis D, Castillejos M, Gold DR, McDonnell W, Borja-Aburto VH (1999) Air pollution and infant mortality in Mexico City. Epidemiology 10:118–123. doi:10.1097/00001648-199903000-00006 CrossRefPubMedGoogle Scholar
  19. Ministério do Meio Ambiente (2014) 1° diagnóstico da rede de monitoramento da qualidade do ar no Brasil. 267.Google Scholar
  20. Moher D, Liberati A, Tetzlaff J, Altman DG, Grp P (2009) Preferred reporting items for systematic reviews and meta-analyses: The PRISMA Statement (Reprinted from Annals of Internal Medicine). Phys Ther 89:873–880. doi:10.1371/journal.pmed.1000097 PubMedGoogle Scholar
  21. Organisation for Economic Co-operation and Development (2012) Environmental outlook to 2050: the consequences of inaction. The Organisation for Economic Co-operation and Development [online], http://www.oecd.org/environment/oecdenvironmental outlookto2050theconsequencesofinaction.htm
  22. Pope CA, Dockery DW, Chow JC, Watson JG, Mauderly JL, Costa DL, et al. (2006) Health Effects of Fine Particulate Air Pollution: Lines that Connect J. Air Waste Manage Assoc 56:1368–1380; doi:10.1080/10473289.2006.10464545 CrossRefGoogle Scholar
  23. Reyna MA, Bravo ME, López R, Nieblas EC, Nava ML (2012) Relative risk of death from exposure to air pollutants: a short-term (2003–2007) study in Mexicali, Baja California, México. Int J Environ Health Res 22:370–386. doi:10.1080/09603123.2011.650153 CrossRefPubMedGoogle Scholar
  24. Romieu I, Gouveia N, Cifuentes LA, de Leon AP, Junger W, Vera J, et al (2012) Multicity study of air pollution and mortality in Latin America (the ESCALA study) Res Rep Health Eff Inst 5–86Google Scholar
  25. Sanhueza PH, Vargas CR, Jiménez J.P (1998) Daily mortality in Santiago and its relation with air pollution. Rev Méd Chile 127:235–242Google Scholar
  26. Shah AS V, Lee KK, McAllister DA, Hunter A, Nair H, Whiteley W, et al. (2015) Short term exposure to air pollution and stroke: systematic review and meta-analysis BMJ 350:h1295. doi:10.1136/bmj.h1295 CrossRefPubMedPubMedCentralGoogle Scholar
  27. Sheehan MC, Lam J (2015) Use of systematic review and meta-analysis in environmental health epidemiology: a systematic review and comparison with guidelines Curr Environ Heal Rep 2:272–283; doi:10.1007/s40572-015-0062-z CrossRefGoogle Scholar
  28. Silva AM da, Mattos IE, Ignotti E, de Hacon SS (2013) Particulate matter originating from biomass burning and respiratory. Rev Saude Publica 47:345–352. doi:10.1590/S0034-8910.2013047004410 CrossRefPubMedGoogle Scholar
  29. Stroup DF (2000) Meta-analysis of observational studies in epidemiology: a proposal for reporting. Jama 283:2008. doi:10.1001/jama.283.15.2008 CrossRefPubMedGoogle Scholar
  30. Valdes A, Zanobetti A, Halonen JI, Cifuentes L, Morata D, Schwartz J (2012) Elemental concentrations of ambient particles and cause specific mortality in Santiago, Chile: a time series study. Environ Health 11:82. doi:10.1186/1476-069X-11-82 CrossRefPubMedPubMedCentralGoogle Scholar
  31. Van Donkelaar A, Martin RV, Brauer M, Hsu NC, Kahn RA, Levy RC et al (2016) Global estimates of fine particulate matter using a combined geophysical-statistical method with information from satellites, models, and monitors. Environ Sci Technol 50:3762–3772. doi:10.1021/acs.est.5b05833 CrossRefPubMedGoogle Scholar
  32. World Health Organization (2006) WHO Air quality guidelines for particulate matter, ozone, nitrogen dioxide and sulfur dioxide: global update 2005: summary of risk assessment. Geneva World Heal Organ 1–22; doi:10.1016/0004-6981(88)90109-6
  33. World Health Organization (2015) WHO Expert Consultation: available evidence for the future update of the WHO Global Air Quality Guidelines (AQGs).10Google Scholar

Copyright information

© Swiss School of Public Health (SSPH+) 2017

Authors and Affiliations

  • Laís Fajersztajn
    • 1
    • 2
  • Paulo Saldiva
    • 1
    • 2
  • Luiz Alberto Amador Pereira
    • 3
  • Victor Figueiredo Leite
    • 4
  • Anna Maria Buehler
    • 5
  1. 1.Laboratory of Experimental Air Pollution (LIM05), Department of Pathology, School of MedicineUniversity of São PauloSão PauloBrazil
  2. 2.Institute for Advanced Studies of the University of São Paulo-IEASão PauloBrazil
  3. 3.Collective Health Pos-Graduation ProgramCatholic University of SantosSantosBrazil
  4. 4.Institute of Physical Medicine and RehabilitationUniversity of São PauloSão PauloBrazil
  5. 5.Health Technology Assessment Unit, Institute of Health Education and ScienceGerman Hospital Oswaldo CruzSão PauloBrazil

Personalised recommendations