Introduction

The term dust refers to minute solid particles emitted into the air from various sources and which are found to have settled onto outdoor objects and surfaces due to either wet or dry deposition (Ferreira-Baptisa and De Miguel 2005; Gill et al. 2006). Urban street dust in particular is a complex mixture consisting of suspended particles (atmospheric aerosol) and displaced soil and biogenic materials (e.g. tree leaves, debris and other plant matter) that can be easily mobilised by moving vehicles (Al-khashman 2007; Charlesworth et al. 2011; Fergusson and Kim 1991; Shi et al. 2008). In addition, emissions from a range of anthropogenic sources (e.g. vehicular exhausts particles, tyre wear, brake lining wear particles, municipal waste incineration as well as construction and building renovations) can all contribute to urban dust composition (Omar et al. 2007; Duzgoren-Aydin 2007; Han et al. 2008; Wei et al. 2010; Wei and Yang 2010; Manasreh 2010). Urban dusts have high-surface area and are easily transported and deposited thereby contributing significantly to potentially toxic elements (PTEs) load (Celis et al. 2004; Allout et al. 1990; Zhao et al. 2006; Irvine et al. 2009).

Dusts can be seen to pose more risk to human health when compared to other environmental matrices like soil; this is due to its pervasive and omnipresent nature (Banerjee 2003). Potential threats to human health due to elevated concentrations of PTEs in urban street dust are now well recognised (Schwar et al. 1988; Shi et al. 2010), and concerns have been expressed about the acute or prolonged long-term adverse effects on human health and ecosystems in general (Lu et al. 2003; Lough et al. 2005; Steiner et al. 2007). In addition, the presence of PTEs in urban street dusts can have an adverse effect on air and water quality (Duong and Lee 2009).

Urban street dust could cause potentially adverse health effects since opportunities exist for a variety of different exposure pathways. For example, exposure could occur as a result of the association of dust particles with consumed food (e.g. eating outdoors) and dust particle adherence to hands followed by hand-to-mouth contact both leading to ingestion of urban street dust (Sezgin et al. 2004; Abrahams 2002) or alternatively inhalation. The inhalation pathway is not the subject of this paper though could be a significant contributor to human health risk, particularly on precipitation-free windy days when urban street dust will be at its most mobile (Laidlaw and Taylor 2011).

Children have the greatest risk considering their outdoor activities and their hand-to-mouth behaviour (Laidlaw and Filippelli 2008; Mielke et al. 1999; Shi et al. 2008). In addition, exposure via chronic contact of urban dust by children living within the vicinity of busy roads is also a significant pathway (Aelion et al. 2008). Numerous studies have been done on urban dusts specifically with respect to their PTE content, fractionation, source identification and contamination assessment, particle size and spatial distribution (e.g. Ahmed et al. 2007; Li et al. 2001; Duong and Lee 2011; Wei and Yang 2010; Han et al. 2008; Apeagyei et al. 2011; Wei et al. 2009; Charlesworth and Lees 1999; Mckenzie et al. 2008) as well as the contribution of lead-contaminated dust to children’s blood lead levels (Liu et al. 2011). On the other hand, there is little information on the human health risk assessment via the ingestion pathway (i.e. oral bioaccessibility) of urban dust. Previous work from this group (Okorie et al. 2012) has determined the total and oral bioaccessible fraction of six elements in urban street dust from one study site. The research highlighted a particular concern for the high lead content of urban street dust. This paper extends the research further by considering the lead content of urban street dust in five cities, all relatively close to each other and each with a known historical past.

Experimental

Fifteen urban street dusts were collected (typically between 2 and 5 g per site) from each of the five different cities in the UK: Durham, Edinburgh, Liverpool, Newcastle upon Tyne and Sunderland. The samples were all collected in the summer of 2010. Newcastle upon Tyne samples were collected on the 27th and 28th May, Sunderland samples on the 31st May, Liverpool samples on the 5th June, Edinburgh samples on the 12th June and Durham samples on the 15th June. The sampling days were selected due to the lack of precipitation, i.e. they were dry and sunny. All the sampled sites were selected randomly but with due regard to the volume of traffic and the location of pedestrians, i.e. central city locations were selected which featured, if possible, pedestrian walkways. Dust samples were collected using a plastic dustpan and brush (Robertson and Taylor 2007; Zhang and Wang 2009). Different dustpans and brushes were used at each site; in addition, gloves were worn to avoid cross contamination. Collected samples were transferred to self-sealing Kraft bags for transportation back to the laboratory. The sampling procedure was maintained for all sites to minimise sampling variability and maintain sample integrity. The samples were dried in a drying cabinet at a temperature of 35 °C for 48 h. The dust samples were then sieved using a <125 μm nylon sieve to remove extraneous matter such as small pieces of building material and other debris. The <125 μm dust samples collected after sieving were weighed (their mass recorded) and stored in sealed plastic containers. All procedures of handling were carried out without contact with metal objects/utensils to avoid potential cross contamination of the samples. All chemicals used in analyses were certified analytical grade. Details of their sources have been reported elsewhere (Elom et al. 2013).

Procedure for sample extraction using UBM and microwave digestion

The in vitro extraction test employed in this work is based on the unified bioaccessibility method (UBM); the preparation of the reagents has been described elsewhere (Wragg et al. 2009) as well as the generic extraction protocols and microwave digestion procedure (Okorie et al. 2012).

Procedure for ICP-MS determination

Samples to be analysed by ICP-MS were prepared in triplicate by measuring 1 mL of either the filtrate, CRM/guidance material or blank into a 10-mL Sarstedt tube; this was followed by 30 μL of an internal standard (Terbium) and 9 mL of water (1 % HNO3). The use of the CRM/guidance material was to assess the precision and accuracy of the methodology, whilst reagent blanks were included to check contamination. Calibration standards in the range 0–400 ppb (7 data points) were prepared, and internal standards were added; this was used to calibrate the instrument and also to construct the calibration graph. The instrument was tuned to verify mass resolution and maximise sensitivity; 208Pb was used to determine the content of samples and standards. A calibration curve based on a concentration range of 0–400 ppb, with 7 calibration data points, was done, and the regression coefficient (R 2) was obtained (0.999).

Results and discussion

The total concentration of Pb in urban street dust, from the five cities, was determined (Table 1). The appropriate accuracy of the total determination was established by the analysis of CRM BCR 143R (Table 1). A box plot of the data (Fig. 1a) shows the variations been different sampling locations in all the studied cities. In Newcastle upon Tyne, the Pb concentration varied between 94 and 1,636 mg/kg with a mean concentration of 558 mg/kg. The highest Pb level in this city was obtained from site 3 (i.e. 1,636 mg/kg), i.e. a central focus point and pedestrianised area in the city centre (near Grey’s Monument). In Durham, the concentration varied between 109 and 2,119 mg/kg with a mean concentration of 446 mg/kg; the highest concentration was determined on a popular street (Saddler Street) with both vehicles and pedestrians. In Liverpool, the concentration varied between 109 and 915 mg/kg with a mean concentration of 362 mg/kg; the highest concentration was found in a busy residential area (Brownlow Hill). In Edinburgh, the concentration varied from 76 to 1,273 mg/kg with a mean concentration of 443 mg/kg; the highest concentration was determined in the pedestrainized area close to the Palace of Holyrood house. Finally, in Sunderland, the concentration varied between 88 and 2,228 mg/kg with a mean concentration of 407 mg/kg; the highest was observed on a busy road (High Street West) close to the main shopping area.

Table 1 Summary of total, stage-related bioaccessibility and residual fraction of lead in urban street dusts and CRMs
Full size table
Fig. 1
figure 1

Box plots of a total lead and b oral bioaccessibility (gastric only) of lead in urban street dusts showing median, mean, box boundary (25th and 75th) percentile and whiskers (10th and 90th) percentile

Full size image

A key characteristic of all the sites investigated, and especially those found to contain the highest concentrations, was their proximity to a high population density, i.e. their central location within or close to pedestrianised areas. Previous work from this group (Okorie et al. 2012) identified mean Pb levels in Newcastle upon Tyne of 992 mg/kg (range 70.2–4,261 mg/kg), with a mean Pb concentration across 26 cities worldwide of 463 mg/kg. In addition, previous data (Ross et al. 2007) from the UK Soil and Herbage Pollutant Survey identified mean Pb levels of 110 ± 90 mg/kg (n = 87) (with a range of 8.6–387 mg/kg) in soils from urban locations.

Although direct comparison of results from different studies is complicated due to the variation in sampling methods, the different sample preparation methodologies, the variety of particle size fractions adopted as well as the digestion and sample analysis protocols, the mean results obtained in this study compared favourably with results obtained from other studies from different cities worldwide (Table 2). The mean Pb concentrations obtained in this study were found to be higher than the mean concentration from 32 cities, whilst results from Lancaster (UK), London (UK) and Madrid (Spain) were ≥2× higher. The general similarity of results across all the cities suggests that the sources of Pb in the urban street dust could be traced to common urban sources.

Table 2 Examples of Pb concentrations in street dust from across the world
Full size table

Oral bioaccessibility data can be used to estimate the amount of Pb, which could be absorbed into the body through the oral ingestion pathway; this is fundamental in assessing risks to humans particularly with respect to elevated Pb levels found in all the cities. The oral bioaccessibility of Pb from the 90 urban street dusts collected (from 5 cities) was assessed using the in vitro gastrointestinal extraction procedure (UBM). The results are shown in Table 1 as the minimum, median and maximum values.

However, the stage-related bioaccessibility (% BAF) is shown only for the maximum determined Pb concentration for each extraction stage thereby representing the worst-case scenario. The results show that the maximum % BAF was always highest in the gastric digest stage with an average of 43 ± 9 % (across all 5 cities). This is in line with many current findings where higher bioaccessibility values have been reported in the gastric phase (Lu et al. 2011; Poggio et al. 2009; Turner 2011). A box plot, showing the mean, median, box boundary (25th and 75th) percentile and whiskers (10th and 90th) percentile, summarises each individual %BAF result (Fig. 1b). In addition, the two certified reference materials (BGS Guidance Material 102 and BCR 143R) were also subjected to the same procedure. The results (Table 1) show reasonable agreement for the certified oral bioaccessibility guidance material (BGS 102); data are also reported for BCR143R showing a total recovery of 83.2 % based on stages II and II. This indicates that the analytical procedures in place were able to be replicated appropriately.

In order to estimate the human health risk associated with exposure to urban street dust, the concentration of Pb from a particular sample that a child (as the most sensitive receptor) might possibly ingest to reach the estimated tolerable daily intake (TDI, for oral ingestion) can be calculated (Pouschat and Zagury 2006; Swartjes 2011; Okorie et al. 2012) (Table 3).

Table 3 Estimated lead daily intake (orally ingested) from urban dust
Full size table

Comparing the data in Table 3 with the TDIoral (Baars et al. 2001) for Pb (3.6 μg/kgbw/day), it is observed that all locations exceed the calculated maximum daily intake, irrespective of city. Typically, a child (aged between 1 and <6 years) would need to consume 32 mg (Durham), 53 mg (Edinburgh), 73 mg (Liverpool), 41 mg (Newcastle) and 30 mg (Sunderland) per day in order to exceed the TDIoral guidelines. Minimum daily intake values are also included in Table 3; in each of these cases, the TDIoral is never exceeded. However, in any risk assessment, it is appropriate to always consider the worst-case scenario, i.e. based on maximum TDIoral. In addition, it is also possible to consider the actual exposure frequency that a child may be exposed to urban dust as a result of anticipated visits to the city centres over the period of 12 months (Table 3). By including exposure frequency, it is noted that the maximum estimated daily intake is considerably reduced (i.e. <0.50 μg/kgbw/day) and over 7 times lower than the TDIoral.

Also, based on the determined oral bioaccessibility (gastric phase only), it is possible to modify the equation to additionally include the fractional bioavailability (Okorie et al. 2012) as a numerator. The results, Table 3, for the maximum daily intake are marginally in excess of the TDIoral in three locations only (Durham, Newcastle and Sunderland); note that, minimum daily intakes are also included for comparison. The gastric phase was used to compute bioaccessible TDI’s because it yielded higher bioaccessibility in all cases and so represents the higher risk. However, it is not expected that a child would ingest (even one with pica behaviour) 100 mg of urban street dust during a 1 h/day visit (throughout the year) to these city locations.

Whilst the data have been based on a daily ingestion rate of 100 mg/day (U.S. EPA 2008), it is believed that this is an overestimate in the context of urban street dust ingestion. We therefore suggest (Okorie et al. 2012) that a realistic assessment of human health risks should be based on the actual determined airborne particulate matter. The process proposed is also limited by some assumptions: most notably that the data are specific to the actual date of sample collection per site; exposure duration of 5 h, i.e. between 10.00 and 15.00 h; that a child would occupy an active area of 1 m3 (based on an estimate of hand-to-mouth distances); and that the urban dust particle size fraction is based on airborne dust (10 μm) [the samples analysed in this study were actually settled street dust (<125 μm)]. [Note: these particular particle size fractions (<10 and 125 μm) have the potential to easily enter the human body either through oral (mouth) or inhalation (nose/mouth) pathways]. It has been reported (Plumlee et al. 2010) that the relationship between the two environmental matrices is basically in terms of sources, physical and chemical characteristics as well as toxicological mechanisms in exposed populations. A recent study (Sipos et al. 2012) that investigated the concentration of Pb and Zn in airborne dust (PM10) and settled dust described both environmental matrices as ‘mobile toxic components’ with the potential to harbour PTEs. The study highlighted that studies on sources, composition, distribution and health impacts of airborne dust and settled dust are necessary for their risk assessment to atmospheric quality, ecology and human health particularly in a populated urban environment.

In the UK, daily air quality data are published (http://uk-air.defra.gov.uk/data) providing experimental values for selected cities including Edinburgh, Liverpool and Newcastle upon Tyne; no data exist for Durham and Sunderland. The determined average experimental data (10-μm particle size, PM10) available on the sampling dates were 8 μg/m3 (Newcastle upon Tyne), 38 μg/m3 (Liverpool) and 9 μg/m3 (Edinburgh). It is then possible to reapply the equation using a revised dust ingestion rate of the following: 0.008 mg/day (Newcastle upon Tyne); 0.038 mg/day (Liverpool) and 0.009 mg/day (Edinburgh). It is now noted that the maximum Pb daily intake is considerably reduced (Table 3) compared to the TDIoral; it is concluded that the risk from Pb in urban dust is minimal when including the actual air quality data. However, it is to be noted that TDIoral only calculated from street dust can be used to assess health, but it was considered necessary to also carry out air quality calculations as a conservative and a holistic approach to human health risk assessment.

Conclusions

By considering the worst-case scenario based on maximum determined Pb concentrations in urban dust across five cities has indicated the low risk associated with exposure. Whilst the maximum estimated daily intake from Pb in urban dust exceeds the TDIoral in all cases, it is an overestimation of risk to the child. By including the oral bioaccessibility into the data analysis, the results indicate that the amount of Pb which is available for absorption into the body, through the oral ingestion pathway, is approximately 60 % lower. The use of oral bioaccessibility data is therefore fundamental in assessing risks to humans particularly with respect to elevated Pb levels found in this study and previously (Okorie et al. 2012).

Given the complexity of modelling exposure and intake pathway (ingestion/inhalation/dermal sorption), particularly in an urban environment, our risk assessments based on worst-case scenarios (i.e. maximum concentrations) must be treated with caution. Furthermore, whilst it is not expected that a child would ingest 100 mg (accepted soil + dust ingestion rate, U.S. EPA 2008) of urban dust during a daily visit to the city centres of these cities, our study shows that a child only needs to ingest <73 mg dust/day in order to exceed the Pb TDIoral. This coupled with the 43 % bioaccessibility highlights the need for regular monitoring of Pb levels in the urban environment and of robust environmental management practices, including regular street sweeping. With the gradual move from leaded to unleaded petrol, a number of studies have observed a reduction in Pb concentration in urban street dusts over recent decades (e.g. Charlesworth et al. 2003; De Miguel et al. 1997; Duzgoren-Aydin 2007; Laidlaw and Filippelli 2008); however, despite this decline monitoring of Pb as part of atmospheric monitoring programs remains warranted along with environmental policies (i.e. air quality management zones) to target a reduction in human exposure to Pb in the urban environment.

Alternate approaches for assessing the maximum daily Pb intake from urban dust have also been considered including those based on the likely annual exposure frequency and the actual air quality data (from three cities). In all cases, using these revised approaches, the environmental health risk was considerably reduced.