Health effects from Sahara dust episodes in Europe: Literature review and research gaps

Environment International 47 (2012) 107–114

Contents lists available at SciVerse ScienceDirect

Environment International journal homepage: www.elsevier.com/locate/envint

Review

Health effects from Sahara dust episodes in Europe: Literature review and research gaps A. Karanasiou a,⁎, N. Moreno a, T. Moreno a, M. Viana a, F. de Leeuw b, X. Querol a a b

Institute of Environmental Assessment and Water Research (IDAEA-CSIC), Barcelona, Spain National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands

a r t i c l e

i n f o

Article history: Received 21 December 2011 Accepted 23 June 2012 Available online 15 July 2012 Keywords: Health effects Sahara dust Desert dust Fine particles Coarse particles PM10

a b s t r a c t The adverse consequences of particulate matter (PM) on human health have been well documented. Recently, special attention has been given to mineral dust particles, which may be a serious health threat. The main global source of atmospheric mineral dust is the Sahara desert, which produces about half of the annual mineral dust. Sahara dust transport can lead to PM levels that substantially exceed the established limit values. A review was undertaken using the ISI web of knowledge database with the objective to identify all studies presenting results on the potential health impact from Sahara dust particles. The review of the literature shows that the association of fine particles, PM2.5, with total or cause‐specific daily mortality is not significant during Saharan dust intrusions. However, regarding coarser fractions PM10 and PM2.5–10 an explicit answer cannot be given. Some of the published studies state that they increase mortality during Sahara dust days while other studies find no association between mortality and PM10 or PM2.5–10. The main conclusion of this review is that health impact of Saharan dust outbreaks needs to be further explored. Considering the diverse outcomes for PM10 and PM2.5–10, future studies should focus on the chemical characterization and potential toxicity of coarse particles transported from Sahara desert mixed or not with anthropogenic pollutants. The results of this review may be considered to establish the objectives and strategies of a new European directive on ambient air quality. An implication for public policy in Europe is that to protect public health, anthropogenic sources of particulate pollution need to be more rigorously controlled in areas highly impacted by the Sahara dust. © 2012 Elsevier Ltd. All rights reserved.

Contents 1. 2.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . Materials and methods . . . . . . . . . . . . . . . . . . . . . 2.1. Description of Sahara dust episodes in Europe . . . . . . . 2.2. Study identification and selection. . . . . . . . . . . . . 3. Results and discussion . . . . . . . . . . . . . . . . . . . . . 3.1. European studies on health effects of Sahara dust . . . . . 3.2. Confounders used . . . . . . . . . . . . . . . . . . . . 3.3. Methods used for the identification of Sahara dust episodes 3.4. Health effects from the Sahara dust fine particles, PM2.5 . . 3.5. Health effects from PM10 and PM2.5–10 . . . . . . . . . . 3.6. Relationship between chemical species and health effects . 3.7. Prediction and mitigation measures. . . . . . . . . . . . 4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

⁎ Corresponding author. E-mail address: [email protected] (A. Karanasiou). 0160-4120/$ – see front matter © 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.envint.2012.06.012

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

107 108 108 108 109 109 109 111 111 111 111 112 113 113 114 114

108

A. Karanasiou et al. / Environment International 47 (2012) 107–114

1. Introduction During the last years there has been an increasing interest in the study of atmospheric aerosols given their confirmed impact on human health. A large number of epidemiological studies demonstrated that atmospheric particulate matter (PM) has a clear correlation with the number of daily deaths and hospitalizations as a result of respiratory and cardiovascular diseases. The USA six cities study (originally conducted by Schwartz et al., 1996 and replicated by Klemm and Mason, 2000) demonstrated a clear association between elevated fine particulate matter (PM) concentrations and daily mortality. Nevertheless health effects from coarse particles are still of concern (Brunekreef and Forsberg, 2005). Recently, special attention has been given to PM air pollution due to mineral dust. The main global source of atmospheric mineral dust is the Sahara desert, which produces about half of the annual mineral dust. Approximately 12% of the Saharan dust moves northwards to Europe, 28% westwards to the Americas and 60% southwards to the Gulf of Guinea (Engelstaedter et al., 2006). Sahara dust transport may greatly increase the ambient levels of PM recorded in air quality monitoring networks (Querol et al., 2009). This is especially relevant in Southern Europe (Escudero et al., 2005, 2007; Gerasopoulos et al., 2006; Kallos et al., 2007; Koçak et al., 2007; Mitsakou et al., 2008; Querol et al., 1998; Rodríguez et al., 2001) and in some Atlantic islands (Arimoto et al., 1997; Chiapello et al., 1995; Coudé-Gaussen et al., 1987; Prospero and Nees, 1986; Prospero et al., 2008; Viana et al., 2002). Sahara dust has attracted increasing attention due to the important influences on nutrient dynamics and biogeochemical cycling in both oceanic and terrestrial ecosystems in North Africa and far beyond, due to frequent long-range transport across the Atlantic Ocean, the Mediterranean Sea and the Red Sea, to the Americas, Europe and the Middle East. Furthermore, Sahara dust particles may have considerable climatic significance through a range of possible influences and mechanisms (Goudie and Middleton, 2001). Sahara dust outbreaks may also cause health impacts due to the high levels of PM and to the transport of anthropogenic pollution (Erel et al., 2007; Rodríguez et al., 2011) and also to the possible transport of micro-organisms (Polymenakou et al., 2008). However, only a limited number of studies have been carried out to establish the health effects due exclusively to the Saharan dust source.

For this reason, the objective of this work is to review the literature to verify if Sahara dust particles have a relevant impact on human health. Additionally, this study expresses the need for determining the possible effects of exposure to Sahara dust particles. The current review was carried out within the frame work of The European Topic Centre for Air Pollution and Climate Change Mitigation (ETC/ACM), http://acm. eionet.europa.eu/. The results of the present review may be considered to establish the objectives and strategies of a new European directive on ambient air quality. 2. Materials and methods 2.1. Description of Sahara dust episodes in Europe The transport of Saharan dust into Europe has a clear seasonality, being more frequent from February to June, and from late autumn to early winter (Escudero et al., 2005), although dust events can be distributed throughout the year. The Mediterranean countries are frequently affected by Sahara dust episodes (Gerasopoulos et al., 2006; Kallos et al., 2007; Koçak et al., 2007; Mitsakou et al., 2008). The reason for this is the low precipitation in the Mediterranean basin that favours the long residence time of PM in the atmosphere with the consequent impact on air quality. Furthermore, >70% of the exceedances of the PM10 daily limit value (2008/50/CE European directive) in most regional background EMEP sites of Spain have been attributed to dust outbreaks (Querol et al., 2009). Fig. 1 (reprint from Querol et al., 2009) shows the number of exceedances of the EU PM10 limit value in 21 regional background sites in Europe calculated from long-term sampling campaigns (details in Querol et al., 2009). Regional background PM10 levels across the Mediterranean show clear increasing trends from north to south and from western to eastern of the Basin. These trends are almost coincident with the PM10 African dust load (Fig. 2, reprint from Querol et al., 2009). Querol et al. (2009) revealed a higher probability of elevated PM levels (>100 μg/m3) occurrence in the Eastern Mediterranean Basin rather than in the Western Mediterranean Basin owing to the higher occurrence and intensity of African dust intrusions. The particle size of desert dust lies in the coarse mode as 48% of the total suspended particles (TSP) are particles with an aerodynamic diameter ≤ 10 μm (Goudie and Middleton, 2006), with suspension being the main mechanism of dust transport. Thus, silt is expected

Fig. 1. Map of 21 regional background (EMEP) sites in Southern Europe with the number of daily exceedances of the PM10 limit value marked for African days (right circle) and non African days (left circle). Source: Querol et al., 2009.

A. Karanasiou et al. / Environment International 47 (2012) 107–114

109

Fig. 2. Mass contribution (in μg/m3) of African dust in PM10 for the 21 regional background (EMEP) sites in Southern Europe. Source: Querol et al., 2009.

to be a dominant component, although clay and sand fractions can also be present, as dust deposits tend to get finer with increasing distance from their source regions. There has been a considerable amount of work in characterising African aerosol chemistry, mainly using major element variations (e.g. Moreno et al., 2006), and on chemical variations linked to particle size fractionation. The chemical composition of dust is mostly dominated by SiO2 (60%), Al2O3 (14%), Fe2O3 (7%), CaO (4%), MgO (2.6%) and K2O (2.4%) (Goudie and Middleton, 2006). Mineralogically, this composition relates to dominance of quartz, magnetite/hematite and carbonates. In the finer fractions clays such as palygorskite, illite and kaolinite can also be important (Alastuey et al., 2005), although there are big differences in clays' abundance depending on the source area of the dust. Thus, dust from North and West Sahara contains abundant illite, whereas kaolinite is dominant in dust from the South and Central Sahara (Goudie and Middleton, 2006). Saharan dust particles may also serve as carries of anthropogenic pollutants (Erel et al., 2007; Rodríguez et al., 2001) and micro-organisms (Polymenakou et al., 2008) as they are transported from North Africa to Europe. 2.2. Study identification and selection An electronic search on the ISI web of knowledge database was conducted using various combinations of the following key words: “particulate matter”, “aerosol”, “PM10,” “PM2.5,” “health,” “dust storm,” “sand storm,” “African dust,” “Sahara dust,” “dust events,” “yellow dust”, “Europe” without restrictions of publication type or publication date. In addition recent articles in relevant journals were collected. Finally, the reference lists of the identified publications were checked for additional studies. In the present study only the European studies reporting on the health effects from the Sahara desert have been reviewed. Studies conducted in other regions other than the European continent were excluded from the study. All the relevant information from each study e.g.: risk factors with their 95% confidence interval were extracted and tabulated. 3. Results and discussion 3.1. European studies on health effects of Sahara dust Unlike the chemical composition and transport pattern of dust events, the possible health effects have been rarely studied. In the

literature only 12 studies report on the health effects of Sahara dust particles in the region of Europe. All these studies have been conducted in south European countries. The majority of these works are epidemiological studies (Jiménez et al., 2010; Mallone et al., 2011; Middleton et al., 2008; Pérez et al., 2008; Sajani et al., 2011; Samoli et al., 2011a, 2011b; Tobías et al., 2011a, 2011b) that investigate the association between Sahara dust outbreaks and mortality/morbidity. One study is a toxicological study where the microbial quality of the aerosols over the eastern Mediterranean region during an African storm was examined (Polymenakou et al., 2008), while one study reports on the inhalation dose during dust episodes (Mitsakou et al., 2008). Finally, Dadvand et al. (2011) have studied the impact of Saharan dust episodes on pregnancy complications (preeclampsia and bacteriuria) and outcomes (birth weight and gestational age at delivery), Samoli et al. (2011b) have investigated the paediatric asthma exacerbation caused by air pollution and dust events while Tobías et al. (2011b) report on the risk of meningococcal meningitis due to Sahara dust. A number of studies (Dadvand et al., 2011; Jiménez et al., 2010; Middleton et al., 2008; Samoli et al., 2011a; Samoli et al., 2011b) use regression models constructed for days with Saharan dust intrusions and for control days when no dust event was observed. Many studies (Mallone et al., 2011; Pérez et al., 2008; Sajani et al., 2011; Tobías et al., 2011a, 2011b)concerning Sahara dust health effects use case-crossover designs. This design uses the day on which the outcome of interest (mortality) occurs as a case day. Exposure on case days is compared with exposure on days on which the outcome of interest does not occur (control days). The epidemiological studies concerning Sahara dust health effects include a time-series analysis on daily mortality/morbidity among a specific age group (above 75 years: Jiménez et al., 2010) or for all ages (Pérez et al., 2008; Sajani et al., 2011; Samoli et al., 2011a; Tobías et al., 2011a). The causes of mortality examined in these studies are classified as total causes (except accidents), cardiovascular, respiratory and cerebrovascular causes. Three studies exist where the association between daily morbidity and dust storms is examined (Middleton et al., 2008; Samoli et al., 2011b; Tobías et al., 2011b). The daily mean PM (PM10, PM2.5, PM2.5–10) levels were used as independent variables. Table 1 provides an overview of the studies conducted in Europe that investigate the health effects of Sahara dust particles. Estimated effects are reported as percentage increases in risk of death (IR %), and 95% confidence intervals (95% CI) associated with an increase of 10 μg m−3 in the PM fractions.

110

Table 1 Summary results of all European studies relating Sahara dust events to mortality, morbidity, and pregnancy implications. % Risk per 10 μg/m3 IR% (95% CI)

Sahara dust days

Non-Sahara dust days

2.9 (−1.1, 6.9) 0.0812 2.8 (0.1, 5.8) 0.016 0.0 (−3.5, 3.6) 0.55 −0.8 (−5.9, 4.6) 0.61 −0.2 (−10.8, 11.5) 0.69 2.5 (−0.9, 6.1) 0.31 1.1 (−4.6, 7.2) 0.92 −2.5 (−9.1, 4.8) 0.51 −0.7 (−5.5, 4.4) 0.53 6.6 (−10.0, 27.0) 0.37 3.0 (−0.0, 6.0) 0.91 8.9 (3.5, 14.5) 0.02 1.6 (−4.7, 8.4) 0.72 5.5 (0.9, 10.2) 0.13 2.5 (−11.9,19.3) 0.80 1.1 (−0.6, 2.7) 0.50 4.9 (2.2, 7.8) 0.01 3.5 (0.6, 6.7) 0.53 4.0 (1.6, 6.5) 0.04 9.8 (0.2, 21.3) 0.35 −0.1 (−0.6, 0.4) p b 0.005 0.2 (−0.4, 0.9) p b 0.005 0.2 (−0.5, 2.7) p b 0.005

2.6 (0.6, 4.6) 0.6 (−1.1, 2.4) 0.8 (0.0–1.6) 0.3 (−0.9, 1.5) 1.6 (−0.9, 1.4) 0.7 (−0.9, 2.3) 0.8 (−1.7, 3.4) −0.1 (−3.4, 3.4) 0.9 (−1.3, 3.2) −1.7 (−9.9, 7.6) 2.8 (1.2, 4.4) 1.9 (−0.7, 4.6) 3.0 (−1.7, 7.9) 1.7 (−0.6, 4.0) 4.6 (−2.6, 12.4) 1.7 (0.7, 2.7) 0.5 (−1.3, 2.2) 2.3 (−0.1, 4.8) 1.1 (−0.4, 2.7) 4.4 (−2.1, 11.7) 1.0 (0.7, 1.4) 1.4 (1.8, 2.9) 1.5 (0.3, 2.8)

No statistical significant effects 2.7 (1.4, 4.1) b0.05 4.0 (1.7, 6.3) b0.05 3.5 (0.9, 6.1) b0.05

2.0 (1.0, 4.0) 3.0 (3.0, 4.0) 3.0 (0.0, 6.0) No statistical significant effects

5.0 (0.5, 9.7) 0.56 8.4 (1.5, 15.8) 0.05 4.8 (0.7, 9.0) NA

Location

Population

Outcome

Identification of Sahara dust days

PM fraction

Causes

Tobías et al. (2011a)

Madrid, Spain

All ages

Mortality

Back trajectoty analysis

Sajani et al. (2011)

Emilia–Romagna, Italy, Aug 2002–Dec 2006

All ages

Mortality

PM number concentrations, back‐trajectory analysis

PM2.5 PM2.5–10 PM10

Mallone et al. (2011)⁎

Rome, Italy, Feb 2001–Dec 2004

Age >35

Mortality

Lidar observations, PM10/NO2 >0.6

TotM TotM TotM CVD RESP TotM CAR CER CIRC RESP TotM CAR CER CIRC RESP TotM CAR CER CIRC RESP TotM CVDa RESa TotM CIRC RES TotM CIRC RESP TotM TotM MORD CVD-MORD

10.4 (−4.7, 27.9) NA

PM2.5

PM10

PM2.5–10

Samoli et al. (2011a)

Athens, Greece, 2001–2006

All ages

Mortality

Jiménez et al. (2010)

Madrid, Spain, Jan 2003–Dec 2005

Age >75

Mortality

Pérez et al. (2008)

Barcelona, Spain, Mar. 2003–Dec. 2004 Nicosia, Cyprus, Jan 1995–Dec 2003

All ages

Mortality

All ages

Morbidity

Middleton et al. (2008)

Samoli et al. (2011b) Tobías et al. (2011b) Dadvand et al. (2011)

Athens, Greece, 2001–2004 Barcelona, Spain Barcelona, Spain, 2003–2005

Aerosol optical depth, back‐trajectory analysis, PM10 remote/PM10 urban > annual median Events identified by the Spanish Ministry for the Environment and Rural & Marine Habitats

PM10

PM2.5 PM10

Back‐trajectory analysis, PM10 remote ≥ 0.5 × PM10 urban Meteorological observations (visibility), PM10 criteria (PM10 ≥ 100 μg/m3)

PM2.5 PM2.5–10 PM10

Age: 0–14

Morbidity

Same as Samoli et al. (2011b)

PM10

PAA

4.1 (0.1, 8.3) 0.41

3.5 (1.6, 5.5) 1.4 (0.8, 3.4) 5.5 (3.5, 7.6) for the highest PM10 levels observed 6.3 (0.0, 15.0) for the highest PM10 levels observed 2.1 (−1.0, 5.2)

All ages

Morbidity Pregnancy

NA PM10, NO2 trends, back‐trajectory analysis

NA PM10

MENG BWb GADb

39.2 (15.2, 68.1) b0.05 −2.1 (−5.8, 1.7) 0.28 0.5 (0.4, 0.6) b0.01

NA NA NA

CI: confidence interval. NA: not available. p-Value: p-value of the interaction between PM and the Saharan dust indicator if b0.05 statistical significant model results. Mortality causes: CAR: cardiac, CER: cerebrovascular, CVD: cardiovascular, CIRC: circulatory, RESP: respiratory, TotM: total mortality. MORD: morbidity, CVD-MORD: cardiovascular morbidity. PAA: paediatric asthma admissions, BW: birth weight, GAD: gestational age at delivery, MENG: meningococcal meningitis. a Among the elderly, age > 75. b Regression coefficients for one episodic day increase for the whole pregnancy.

A. Karanasiou et al. / Environment International 47 (2012) 107–114

% Risk per 10 μg/m3 (95% CI) p-value

Reference

A. Karanasiou et al. / Environment International 47 (2012) 107–114

3.2. Confounders used Tobías et al. (2011a), Sajani et al. (2011), Pérez et al. (2008) and Mallone et al. (2011) adjusted for several confounders: temperature, humidity, pressure, flu epidemic weeks, and heat waves. In the study of Jiménez et al. (2010) the confounders were: ambient pollutants (chemical, biotic and acoustic); temperature; trend; seasonality; influenza epidemics; and autocorrelations between mortality series. Samoli et al. (2011a) investigated the effect of desert dust events on mortality using an indicator variable and its interaction with PM concentrations. The confounders used were long-term trends and seasonality as well as meteorological variables, temperature, humidity, correlation and potential confounding effects from other pollutants, SO2, NO2 or O3 were also included in the model. Middleton et al. (2008) report on the effect of changes in levels of air pollutants on the number of daily admissions in hospitals by means of Poisson regression models. The model controlled for long- and short‐term trends, temperature and relative humidity. Dadvand et al. (2011) used regression models to study the effect of Sahara dust outbreaks on pregnancy complications (preeclampsia and bacteriuria) and outcomes (birth weight and gestational age at delivery). The confounders used might influence the results of the epidemiological studies. The occurrence of African dust events is largely driven by seasonal trends, resulting in a spring–early summer maximum over the Eastern Mediterranean Basin, and a clear summer maximum in the Western Mediterranean Basin (Querol et al., 2009). This means that in Spain and Italy these episodes coincide with the maximum ambient temperatures while in Greece moderate temperatures prevail during these events. Indeed two studies conducted in Spain (Jiménez et al., 2010; Tobías et al., 2011a, 2011b) and one study conducted in Italy (Mallone et al., 2011) report higher temperature and ozone levels during Sahara dust days. Moreover, Pérez et al. (2008) report on the potential confounding effect of the summer 2003 heat wave. On the other hand the study conducted in Greece (Samoli et al., 2011a, 2011b) found no difference in the temperature and ozone mean values between days with and without Sahara dust. The results concerning hospital admissions are consistent since the studies conducted in eastern Mediterranean basin found higher effects when temperature and ozone and SO2 levels are higher (Middleton et al., 2008; Samoli et al., 2011b). 3.3. Methods used for the identification of Sahara dust episodes In the reviewed studies dust episodes were identified by one of the following methods: • by back-trajectory analysis (Hysplit model) using NRL, SKIRON and BSC-DREAM dust maps and satellite images provided by the NASA SeaWiFS (Tobías et al., 2011a) • by the combination of Lidar observations with operational models and the changes of the ratio PM10/NO2. A PM10/NO2 ratio larger than 0.6 revealed the intrusion of Sahara dust particles where PM10/NO2 ratios ≤ 0.6 were found to be characteristic of traffic air pollution (Mallone et al., 2011) • by back-trajectory analysis in combination with a data driven criterion, based on the ratio of the PM10 concentration in the most remote fixed monitoring station to a centrally located one exceeding the median of this ratio during that year (Samoli et al., 2011a, 2011b) • by the combination of air mass back-trajectories calculated using the FLEXTRA model and high hourly number concentration of coarse (diameter > 1 μm) aerosol particles, considered to be a marker of dust transport from the Sahara desert (Sajani et al., 2011) • by the data provided from the Spanish Directorate General for Environmental Quality & Assessment (Dirección General de Calidad and Evaluación Ambiental), Ministry for the Environment and Rural & Marine Habitats (Ministerio de Medio Ambiente y Medio Rural y Marino) available at: http://www.calima.ws/ (Jiménez et al., 2010)

111

• by back trajectories analysis using information obtained from NRL, SKIRON and BSC-DREAM dust maps and satellite images in order to identify the transport of air masses of Sahara–Sahel region. In addition days on which air mass transport occurred were classified as Saharan dust days if levels of PM10 concentrations measured at a reference remote rural monitoring site reached at least 50% of the PM10 levels measured of the urban site (Pérez et al., 2008). It should be noted that especially for Eastern Mediterranean countries dust episodes might occur due to the transport of desert dust from the Arabian Peninsula (Kubilay et al., 2000). In fact, Samoli et al. (2011a) identified one dust event originating from the Arabian Peninsula. However, these episodes are not as frequent as the Saharan dust intrusions and could not affect significantly the results and conclusions of the epidemiological studies. The methods used for the identification of Sahara dust episodes might influence the outcome of the epidemiological studies. It is possible that some days are erroneously identified as episodic days e.g.: the use of PM data of solely urban sites includes a high level of uncertainty as these sites are highly affected by other sources of dust mainly anthropogenic (Pey et al., 2010). The presence of demolition, construction dust and/or the resuspension of road dust could provide rather high ratios of PM10/NO2 (higher than 0.6). Episodes of high resuspension might coincide with the occurrence of air masses transported from North Africa (that could be confirmed by back-trajectories analysis) but they cannot be characterised as Sahara dust episodes. To identify an African dust episode apart from back‐trajectory analysis the comparison of PM levels at an urban site and at remote site closely located at the urban site should be conducted (Escudero et al., 2007). 3.4. Health effects from the Sahara dust fine particles, PM2.5 Only four works report on the health effects of the fine fraction of the Sahara dust (Jiménez et al., 2010; Mallone et al., 2011; Pérez et al., 2008; Tobías et al., 2011a). These studies produced similar results, the association (if any) of fine particles with total or cause specific mortality was not statistically significant (p > 0.05). In Barcelona (Spain) the effects of short-term exposure to PM2.5 were not statistically significant (p for interaction = 0.56) during Saharan dust days. A daily increase of 10 μg/m3 in PM2.5 increased daily mortality by 5.0% (95% CI 0.5–9.7) during Saharan dust days compared with 3.5% (95% CI 1.6–5.5) during non-Saharan dust days (Pérez et al., 2008). The study of Jiménez et al. (2010) conducted in Madrid (Spain) found that the daily mean PM2.5 concentrations displayed a significant statistical association with daily mortality for total mortality, circulatory and respiratory causes on non-Saharan dust days, while this association was absent for Saharan dust days. A recent study conducted in the same region (Madrid, Spain) reports that PM2.5 health effects were similar during days with and without Sahara dusts 2.9% and 2.6% risk respectively, with p for interaction= 0.08 (Tobías et al., 2011a). In Rome, (Italy) Mallone et al. (2011) found that PM2.5 effect estimates during Sahara dust days were positive for natural causes mortality, cardiac and respiratory mortality, but not statistically significant with p values of the interaction terms greater than 0.15 and always above 0.30. 3.5. Health effects from PM10 and PM2.5–10 Although all studies reach the same conclusion regarding fine particles the results for PM10 fraction are not in agreement. Some of the published studies state that PM10 particles during Sahara dust days increase mortality while other studies find no association between mortality and PM10. Recently, Sajani et al. (2011) in Emilia Romagna, (Italy) found no evidence of an effect modification of dust events on relationship between PM10 and daily deaths. Similarly, a recent study conducted in Athens (Greece) reports that the effect of PM10 on mortality during non-desert dust events was higher than the effect identified

112

A. Karanasiou et al. / Environment International 47 (2012) 107–114

during the whole studied period and essentially null during dust event days. The analyses indicated that traffic‐related particles, which prevail on non-desert dust days, have more toxic effects than the ones originating from long-range transport, such as the Sahara dust (Samoli et al., 2011a). On the other hand, few studies (Jiménez et al., 2010; Mallone et al., 2011) found a significant association between mortality and PM10 during the Sahara dust events. In the study of Mallone et al. (2011) PM10 concentrations during the Sahara dust days caused an increase of 8.9% of cardiac mortality compared to 1.9% on dust free days, (p= 0.02). Jiménez et al. (2010) found that daily PM10 concentrations in Madrid displayed a significant statistical association with daily mortality for all causes on days with the Saharan dust while this association was not significant on non-Sahara dust days. Regarding coarse particles PM2.5–10, it should be noted that sometimes PM2.5–10 is calculated by subtracting a direct measurement of PM2.5 from a direct measurement of PM10. The disadvantage of this is that coarse particle measurement is then affected by two measurement errors rather than one. In the study of Jiménez et al. (2010) conducted in Madrid, Spain, the PM2.5–10 fraction did not cause any significant effect in daily mortality with or without the presence of Sahara dust intrusions. On the contrary, for the same area (Madrid), Tobías et al. (2011a) reached the conclusion that PM2.5–10 effect was higher during Saharan dust days than during non-Saharan dust days. In addition, Pérez et al. (2008) showed that coarse particles during Sahara dust days significantly increased daily mortality in Barcelona, Spain. A daily increase of 10 μg/m3 of PM2.5–10 increased daily mortality by 8.4% (p=0.05) compared to 1.4% determined for the non-Sahara dust days. Mallone et al. (2011) report higher mortality effects of coarse particles during desert dust episodes for specific cause mortality. They found that associations of PM2.5–10 with cardiac mortality were stronger on Saharan dust days (4.9%, p = 0.005) than on non-Sahara dust days (0.5%). Regarding the morbidity studies the results are rather consistent. The analysis of morbidity in Nicosia, Cyprus (Middleton et al., 2008) showed an increased risk (10.4%) of hospitalisation on Saharan dust storm days, particularly due to cardiovascular causes but the magnitude of the effect was comparable to that seen on the quartile of non-desert days with the highest levels of PM10 (6.3%). The authors stated that inference from these associations was limited by the small number of dust storm days in the study period. Another study, on paediatric asthma exacerbation (Samoli et al., 2011b) reports higher admissions during Sahara dust episodes.

The differences between the epidemiological studies concerning PM10 and coarse fraction might lie in the different sections of the Sahara desert that affected the studied areas and also the different route that air masses followed. A major source area for transport of African dust to western Mediterranean was identified in southernmost Algeria, and Western Sahara (Middleton and Goudie, 2001). On the other hand, the main atmospheric scenario responsible for the transport of desert dust over the Eastern Mediterranean is the travel of air masses originating from Libya and Egypt (Kallos et al., 2006). The different regions of the Sahara desert have distinct mineralogical characteristics (Moreno et al., 2006). Thus, dust from North and West Sahara contains abundant illite, whereas kaolinite is dominant in dust from the South and Central Sahara (Goudie and Middleton, 2006). This fact might also affect the toxicological properties of the Sahara dust particles. Moreover the air path that these air masses follow influences its chemical composition as they might be mixed with anthropogenic emissions of industrial regions (e.g.: North African countries) and acts as carriers of industrial pollutants (Rodríguez et al., 2001). The studies that report on PM2.5–10 (Mallone et al., 2011; Pérez et al., 2008; Tobías et al., 2011a) concluded that Saharan dust outbreaks may have adverse health effects and pointed out the necessity of further investigation into the role of coarse particles and the mechanism by which Saharan dust increases mortality. 3.6. Relationship between chemical species and health effects Concerning the potential impact of specific chemical species on health in Europe, Pérez et al. (2008) studied the chemical composition of particulate matter to explain changes of health effects. They obtained chemical composition of PM2.5 and PM10 particles including: nonmineral carbon (nmC), total carbon, crustal and marine aerosol elements (sodium and chloride), inorganic secondary components (sulphate, nitrate, and ammonium), metals and trace elements. The chemical composition of PM2.5–10 was obtained by subtracting the chemical composition of PM2.5 from PM10. Comparison of changes in adjusted mass and chemical concentration during Saharan and non-Saharan dust days was carried out by multivariate linear regression. Fig. 3 (reprint from Pérez et al., 2008) provides the mass concentration of major species groups during Saharan and non-Saharan dust days. During Saharan dust days, PM2.5 was dominated by nonmineral carbon (40% of the particulate matter) and secondary inorganic aerosols (34%), with lesser amounts of

Fig. 3. Mean concentrations (with 95% CI) of the major group of elements in PM2.5 and PM10–2.5 during non-Saharan dust days (n = 80) and Saharan dust days (n = 9). Asterisk indicates p value b 0.05 for comparison within each fraction of the mass-adjusted concentrations during Saharan and non-Saharan dust days. Source: Pérez et al., 2008.

A. Karanasiou et al. / Environment International 47 (2012) 107–114

crustal elements (23%). In contrast, PM10-2.5 was dominated by crustal elements (65%), with lesser amounts of secondary aerosols (18%), marine aerosols (8%), and nonmineral carbon (8%). The chemical composition varied equally in PM2.5 and PM2.5–10 during Saharan dust days. In addition, metals involved in oxidative stress pathways, such as iron, copper, lead, and zinc were similarly abundant during Saharan dust days and non-Saharan dust days in PM2.5–10. The authors concluded that some factor associated with PM2.5–10 and transported by Sahara dust such as pesticides or industrial byproducts, though not detected, could be responsible for increased mortality. They also suggest that the biogenic load of coarse particles in Saharan dust is a possible explanation of observed effects. Although coarse particles seem to be more hazardous during Saharan dust days, differences in chemical composition could not explain these observations. Further work should be conducted regarding the chemical composition of coarse fraction during Sahara dust episodes. The study of Pérez et al. (2008) raises concern over possible underestimation of toxicity from coarse particles when they come from desert sources (Sandstrom and Forsberg, 2008). Evidence exists from previous studies that allergenic and inflammatory effects are stronger in coarse particles than in fine particles (Happo et al., 2007, Alexis et al., 2006) while Griffin et al. (2001) detected the presence of bacteria and fungi in Saharan dust particles. The study of Polymenakou et al. (2008) reveals the presence of numerous pathogens at aerosol particles during an intense Sahara dust event in the Eastern Mediterranean. The transport of Sahara dust mixed with industrial pollutants of North Africa has also been confirmed (Rodríguez et al., 2001). 3.7. Prediction and mitigation measures There is evidence and scientific consensus that climate change is real. Over the past years, many countries across the European region have experienced increasing numbers of heat waves, floods and/or droughts (Trenberth et al., 2007). Moreover, these climate events may add to the existing problems of desertification, while also introducing new threats to human health. Specifically, the European continent has large areas of dry lands around 299.7 Mha, susceptible to desertification (UNEP, 1991). Desertification processes affect between 8 and 10% of the total European land from low to high degrees of degradation. The Mediterranean region is particularly sensitive to desertification processes due to its fragile environmental conditions (geomorphology favouring soil erosion processes, increasing forest fires, etc.). Steps should be taken to ensure relevant health protection, as climate change and desertification process will probably cause an increase of dust outbreaks. Further work is needed on the important questions of monitoring, prediction and forecasting of Sahara dust intrusions. Dust forecasts could be used to inform the public so as to reduce health risks. Alert messages and general recommendations to the public and especially the elderly such as to reduce outdoor activities during dust storms could improve public health. Another issue to consider is that sedimentation of African dust during dust outbreaks may increase resuspension during a number of days after the episode. These resuspension contributions may considerably enhance PM levels for a longer period than the duration of the dust intrusion. To abate the resuspension of deposited dust after intensive African dust outbreaks, mitigation measures could be adapted by the national authorities of countries suffering by frequent episodes. Street sweeping and cleaning could be efficient in reducing Sahara dust particles available for resuspension. Recent studies provide evidence that these methods reduce resuspension from paved road surfaces (Amato et al., 2010; Karanasiou et al., 2011). 4. Conclusions In Europe the possible health effects of dust outbreaks have been rarely studied. The review of the literature shows that the association

113

of fine particles, PM2.5, with total or cause‐specific daily mortality is not significant during Saharan dust intrusions. However, regarding coarser fractions PM10 or PM2.5–10 a clear answer cannot be given. Some of the published studies state that PM10 or PM2.5–10 increases mortality during Sahara dust days while other studies find no association between mortality and these size fractions. The reasons for this discrepancy between the published studies might lie in: • The different size fraction studied (PM10 or PM2.5–10) and the different ways to determine coarse concentration, e.g.: PM2.5–10 is calculated by subtracting a direct measurement of PM2.5 from a direct measurement of PM10. The disadvantage of this is that coarse particle measurement is then affected by two measurement errors rather than one. Studies examining PM10 showed inconsistent results. Several studies showed that the association of Sahara dust episodes and daily mortality is not significant (Sajani et al., 2011; Samoli et al., 2011a); however, Jiménez et al. (2010) found that PM10 concentrations displayed a significant statistical association with daily mortality on days with Saharan dust while there was no statistical effect attributable to PM10 exposure on non-Sahara dust days. Most studies that report on PM2.5–10 (Mallone et al., 2011; Pérez et al., 2008; Tobías et al., 2011a) are in agreement with each other, a strong association was found between concentration levels and daily mortality. Only the study of Jiménez et al. (2010) found no effect of PM2.5–10 on mortality during days with and without Sahara dust. • The different time periods studied, or the different meteorological conditions of each region during the Sahara dust episodes resulting in poor adjustment of meteorology e.g.: heat waves (Spain in 2003). • The different studied areas. As environmental conditions are often not evenly distributed over the years, the episodic nature of dust events makes them difficult to study. Epidemiological studies require sufficient statistical power to detect marginally significant effects while the low frequency of Saharan dust days restricts analysis of cause-specific mortality. As the occurrence and intensity of Sahara dust intrusions is higher in the Eastern Mediterranean Basin rather than in the Western Mediterranean Basin the statistical power of epidemiological studies conducted in these regions is probably different. However, Tobías et al. (2011a, 2011b) and Jiménez et al. (2010) both conducted in Madrid (Spain) for the same time period (Jan. 2003– Dec. 2005) reached different results. • The different methodologies used for the definition of episodic days that include combinations of back trajectories analysis, PM10 concentration levels, Lidar observations and PM10/NO2 values. In some cases depending on the intensity of the episodes, it is difficult to classify a day as a Sahara dust episode or not. Moreover the use of ratio PM10/NO2 could give misleading results as its value during construction or resuspension of road dust could be high. Given that the chemical composition of Sahara dust aerosols is distinct from that of other crustal sources this could prove a useful tool for the researchers. However, none of the studies have applied any criterion related to the chemical composition of the studied aerosols. Researchers should conclude to a common protocol for the definition of Sahara dust episodes to produce comparable results. • The different sections of the Sahara desert that affected the studied areas and the different routes of air masses. A major source area for transport of African dust to western Mediterranean is southernmost Algeria and Western Sahara while air masses from Libya and Egypt transfer desert dust to the Eastern Mediterranean. The different regions of the Sahara desert have distinct mineralogical characteristics. Moreover the air path that these air masses followed influences its chemical composition as they might be mixed with anthropogenic emissions of industrial regions (e.g.: North African countries) and act as carriers of industrial pollutants. The main conclusion of this review is that the health impact of Saharan dust outbreaks needs to be further explored, especially in

114

A. Karanasiou et al. / Environment International 47 (2012) 107–114

Southern European areas, the most impacted by these episodes using comparable exposure and statistical methodologies. Given that some evidence exists for the health risks of coarse particles, PM2.5–10 during Sahara dust events, future studies should focus on the chemical characterization of the coarse particle mode during these episodes. The toxic effects of coarse particles when they originate from desert sources should be further investigated. The outcome of this paper may be considered to establish the objectives and strategies of a new European directive on ambient air quality. The Sahara dust outbreaks, which are beyond control from the Member States, mostly affect Mediterranean countries. An implication for public policy in Europe is that to protect public health, anthropogenic sources of particulate pollution need to be more rigorously controlled in areas highly impacted by the Saharan dust (Pérez and Künzli, 2011). Acknowledgements This work was supported by the European Environment Agency (EEA) and the European Topic Centre on Air Pollution and Climate Change Mitigation (ETC-ACM). The authors would like to thank Laura Perez and Aurelio Tobias for their recommendations to improve the manuscript. References Alastuey A, Querol X, Castillo S, Escudero M, Avila A, Cuevas E, et al. Characterisation of TSP and PM2.5 at Izaña and Sta. Cruz de Tenerife (Canary Islands, Spain) during a Saharan dust episode. Atmos Environ 2005;39:4715–28. Alexis NE, Lay JC, Zeman K, Bennett WE, Peden DB, Soukup JM, et al. Biological material on inhaled coarse fraction particulate matter activates airway phagocytes in vivo in healthy volunteers. J Allergy Clin Immunol 2006;117:1396–403. Amato F, Querol X, Johansson C, Nagl C, Alastuey A. A review on the effectiveness of street sweeping, washing and dust suppressants as urban PM control methods. Sci Total Environ 2010;408:3070–84. Arimoto R, Ray BJ, Lewis NF, Tomza U, Duce RA. Mass-particle size distributions of atmospheric dust and the dry deposition of dust to the remote ocean. J Geophys Res 1997;102:15867–74. Brunekreef B, Forsberg B. Epidemiological evidence of effects of coarse airborne particles on health. Eur Respir J 2005;26:309–18. Chiapello I, Bergametti G, Gomes L, Chatenet B. An additional low layer transport of Sahelian and Saharan dust over the North-Eastern Tropical Atlantic. Geophys Res Lett 1995;22:3191–4. Coudé-Gaussen G, Rognon P, Bergametti G, Gomes L, Strauss B, Gros JM, et al. Saharan dust over Fuerteventura Island (Canaries), chemical and mineralogical characteristics, air mass trajectories and probable sources. J Geophys Res 1987;92:9753-9711. Dadvand P, Basagaña X, Figueras F, Amoly E, Tobias A, de Nazelle A, et al. Saharan dust episodes and pregnancy. J Environ Pregnancy 2011;13(11):3222–8. Engelstaedter S, Tegen I, Washington R. North African dust emissions and transport. Earth Sci Rev 2006;79:73-100. Erel Y, Kalderon-Asael B, Dayan U, Sandler A. European atmospheric pollution imported by cooler air masses to the Eastern Mediterranean during the summer. Environ Sci Technol 2007;41:5198–203. Escudero M, Castillo S, Querol X, Avila A, Alarcon M, Viana MM, et al. Wet and dry African dust episodes over Eastern Spain. J Geophys Res 2005;110(D18S08): 10.1029. Escudero M, Querol X, Ávila A, Cuevas E. Origin of the exceedances of the European daily PM limit value in regional background areas of Spain. Atmos Environ 2007;41:730–44. Gerasopoulos E, Kouvarakis G, Babasakalis P, Vrekoussis M, Putaud J-P, Mihalopoulos N. Origin and variability of particulate matter (PM10) mass concentrations over the Eastern Mediterranean. Atmos Environ 2006;40(25):4679–90. Goudie AS, Middleton NJ. Saharan dust storms: nature and consequences. Earth Sci Rev 2001;56:179–204. Goudie AS, Middleton NJ. Desert dust in the global system. Heidelberg: Springer; 2006. Griffin DW, Garrison VH, Herman JR, Shinn EA. African desert dust in the Caribbean atmosphere: microbiology and public health. Aerobiologia 2001;17:203–13. Happo MS, Salonen RO, Halinen AI, Jalava PI, Pennanen AS, Kosma VM, et al. Dose and time dependency of inflammatory responses in the mouse lung to urban air coarse, fine, and ultrafine particles from six European cities. Inhal Toxicol 2007;19(3): 227–46. Jiménez E, Linares C, Martínez D, Díaz J. Role of Saharan dust in the relationship between particulate matter and short-term daily mortality among the elderly in Madrid (Spain). Sci Total Environ 2010;408(23):5729–36. Kallos G, Papadopoulos A, Katsafados P, Nickovic S. Transatlantic Saharan dust transport: model simulation and results. J Geophys Res 2006;111:D09204. http://dx.doi.org/ 10.1029/2005JD006207.

Kallos G, Astitha M, Katsafados P, Spyrou C. Long-range transport of anthropogenically and naturally produced particulate matter in the Mediterranean and North Atlantic: current state of knowledge. J Appl Meteorol Climatol 2007;46:1230–51. Karanasiou A, Moreno T, Amato F, Lumbreras J, Narros A, Borge R, et al. Road dust contribution to PM levels — evaluation of the effectiveness of street washing activities by means of positive matrix factorization. Atmos Environ 2011;45:2193–201. Klemm RJ, Mason Jr RM. Aerosol Research and Inhalation Epidemiological Study (ARIES): air quality and daily mortality statistical modelling — interim results. J Air Waste Manag Assoc 2000;50:1433–9. Koçak M, Mihalopoulos N, Kubilay N. Contributions of natural sources to high PM10 and PM2.5 events in the eastern Mediterranean. Atmos Environ 2007;41:3806–18. Kubilay N, Nickovic S, Moulin C, Dulac F. An illustration of the transport and deposition of mineral dust onto the Eastern Mediterranean. Atmos Environ 2000;34:1293–303. Mallone S, Stafoggia M, Faustini A, et al. Saharan dust and associations between particulate matter and daily mortality in Rome, Italy. Environ Health Persp 2011. http://dx.doi.org/ 10.1289/ehp. 1003026. Middleton NJ, Goudie AS. Saharan dust: sources and trajectories. Trans Inst Br Geogr NS 2001;26:165–81. Middleton N, Panayotis Y, Kleanthous S, Kolokotroni O, Schwartz J, Dockery DW, et al. A 10-years time-series analysis of respiratory and coardiovascular morbidity in Nicosia, Cyprus: the effect of short term changes in air pollution and dust storms. Environ Health 2008;7:39. Mitsakou C, Kallos G, Papantoniou N, Spyrou C, Solomos S, Astitha M, et al. Saharan dust levels in Greece and received inhalation doses. Atmos Chem Phys 2008;8: 11967–96. Moreno T, Querol X, Castillo S, Alastuey A, Cuevas E, Herrmann L, et al. Geochemical variations in aeolian mineral particles from the Sahara–Sahel dust corridor. Chemosphere 2006;65:261–70. Pérez L, Künzli N. Saharan dust: no reason to exempt from science or policy. Occup Environ Med 2011;68:389–90. Pérez L, Tobias A, Querol X, Künzli N, Pey J, Alastuey A, et al. Coarse particles from Saharan dust and daily mortality. Epidemiology 2008;19:800–7. Pey J, Querol X, Alastuey A. Discriminating the regional and urban contributions in the North-Western Mediterranean: PM levels and composition. Atmos Environ 2010;44:1587–96. Polymenakou PN, Mandalakis M, Stephanou EG, Tselepides A. Particle size distribution of airborne microorganisms and pathogens during an intense African dust event in the Eastern Mediterranean. Environ Health Perspect 2008;113:292–6. Prospero JM, Nees RT. Impact of the North African drought and El Niño on mineral dust in the Barbados trade winds. Nature 1986;320:735–8. Prospero JM, Blades E, Naidu R, Mathison G, Thani H, Lavoie MC. Relationship between African dust carried in the Atlantic trade winds and surges in pediatric asthma attendances in the Caribbean. Int J Biometeorol 2008;52:823–32. Querol X, Alastuey A, Puicercus JA, Mantilla E, Miro JV, Lopez-Soler A, et al. Seasonal evolution of suspended particles around a large coal-fired power station: particulate levels and sources. Atmos Environ 1998;32:1963–78. Querol X, Pey J, Pandolfi M, Alastuey A, Cusack M, Pérez N, et al. African dust contributions to mean ambient PM10 mass-levels across the Mediterranean Basin. Atmos Environ 2009;43:4266–77. Rodríguez S, Querol X, Alastuey A, Kallos G, Kakaliagou O. Saharan dust contributions to PM10 and TSP levels in Southern and Eastern Spain. Atmos Environ 2001;35: 2433–47. Rodríguez S, Alastuey A, Alonso-Pérez S, Querol X, Cuevas E, Abreu-Afonso J, et al. Transport of desert dust mixed with North African industrial pollutants in the subtropical Saharan Air Layer. Atmos Chem Phys 2011;11:6663–85. Sajani SZ, Miglio R, Bonasoni P, et al. Saharan dust and daily mortality in Emilia–Romagna (Italy). Occup Environ Med 2011;68:446–51. Samoli E, Kougea E, Kassomenos P, Analitis A, Katsouyanni K. Does the presence of desert dust modify the effect of PM10 on mortality in Athens, Greece? Sci Total Environ 2011a;409:2049–54. Samoli E, Nastos PT, Paliatsos AG, Katsouyanni K, Priftis KN. Acute effects of air pollution on pediatric asthma exacerbation: evidence of association and effect modification. Environ Res 2011b;111:418–24. Sandstrom T, Forsberg B. Desert dust: an unrecognized source of dangerous air pollution? Epidemiology 2008;19(6):808–9. Schwartz J, Dockery DW, Neas LM. Is daily mortality associated specifically with fine particles? J Air Waste Manag Assoc 1996;46:927–39. Tobías A, Cayla JA, Pey J, Alastruey A, Querol X. Are Saharan dust intrusions increasing the risk of meningococcal meningitis? Int J Infect Dis 2011a;15:503. Tobías A, Pérez L, Díaz J, Linares C, Pey J, Alastruey A, et al. Short-term effects of particulate matter on total mortality during Saharan dust outbreaks: a case-crossover analysis in Madrid (Spain). Sci Total Environ 2011b;412:386–9. Trenberth KE, et al. Observations: surface and atmospheric climate change. In: Solomon S, et al, editor. Climate change 2007: the physical science basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate ChangeCambridge: Cambridge University Press; 2007. UNEP. Status of desertification and implementation of the United Nations plan of action to combat desertification. Report of the executive director. UNEP/GCSS.III/3. Nairobi; 1991. Viana M, Querol X, Alastuey A, Cuevas E, Rodríguez S. Influence of African dust on the levels of atmospheric particulates in the Canary Islands air quality network. Atmos Environ 2002;36:5861–75.