Child Exposure to Lead in the Vicinities of Informal Used Lead-Acid Battery Recycling Operations in Nairobi Slums, Kenya

Background. Child exposure to lead from informal used lead-acid battery (ULAB) recycling operations is a serious environmental health problem, particularly in developing countries. Objectives. We investigated child exposure to lead in the vicinities of ULAB recycling operations in the Dandora, Kariobangi and Mukuru slums in Nairobi between January and August 2015. Methods. Top soil (n = 232) and floor dust (n = 322) samples were collected from dwelling units (n = 120) and preparatory schools (n = 44) and analyzed using an inductively coupled plasma-optical emission spectrometer at the Mines and Geological Department Laboratory in the Ministry of Mining, Nairobi. From the obtained lead levels in soil and house dust, child blood lead levels were subsequently predicted using the Integrated Exposure Uptake Biokinetic Model for Lead in Children (IEUBK), Windows version. Results. Lead loadings in all the floor dust samples from the Dandora, Kariobangi and Mukuru slums exceeded the United States Environmental Protection Agency (USEPA) guidance value for lead on floors with a range of 65.2 – 58,194 μg/ft2. Control floor dust samples recorded lower lead loadings compared to the Dandora, Kariobangi and Mukuru slums. Lead concentration in 70.7% of the soil samples collected from waste dumps, industrial sites, residential areas, playgrounds and preparatory schools in Dandora, Kariobangi and Mukuru exceeded the respective USEPA guidance values for lead in soils. Lead concentration in 100% of control soil samples were below the respective USEPA limits. The IEUBK model predicted that nearly 99.9% of children ≤ 7 years old living near informal ULAB recycling operations in Dandora, Kariobangi and Mukuru were at risk of being lead poisoned, with predicted blood lead levels (BLL) above the Centers for Disease Control (CDC) reference value for blood lead. A total of 99.9% of exposed children living in the Mukuru slums are likely to have BLL above 34 μg/dL. Conclusions. There is a need for coordinated efforts to decrease lead emissions from informal battery recycling in Nairobi slums and to remediate existing soils, particularly around battery workplaces and dumpsites. The BLL of local children should be clinically tested and appropriate intervention measures taken.


Introduction
Lead exposure from informal used lead-acid battery (ULAB) recycling operations is a serious environmental health problem, especially in developing countries. 1-3 Research shows that young children living near informal ULAB recycling operations have elevated blood lead levels (BLL) and fatalities have also been reported. [4][5][6][7] The operations release large amounts of lead dust and wastes containing lead into surrounding soils, air, buildings and waterways. 1,4,6,8 The released lead ultimately finds its way into human bodies through various pathways. 2,5, 6 In young children, ingestion of lead-contaminated soil and dust is the major pathway of exposure. Young children are usually the most exposed and susceptible to the toxic effects of lead due to their behavioral, physiological and Used lead-acid battery United States Environmental Protection Agency exposure to lead in the formal battery recycling sectors and found significant levels of lead in human (blood, hair, nails) and environmental (soil, air and water) samples. 16, 17 To date, no studies have been carried out in the informal battery recycling sector in Nairobi. The present study therefore assessed child exposure to lead in the vicinities of informal ULAB recycling operations in Dandora, Kariobangi and Mukuru slums in Nairobi in an effort to characterize childhood lead exposure in these communities. The study focused on young children between 0-7 years (0-84 months), and had the following specific objectives:

1.
To determine the levels of lead in floor dust in dwelling units and preparatory schools in the vicinities of informal ULAB recycling operations in Nairobi slums.

2.
To determine the concentration of lead in soil in residential areas, preparatory schools, children's playgrounds and dumpsites in the vicinities of informal ULAB recycling activities and workplaces.
3. To predict BLL in children under seven years of age living within the study area using the Integrated Exposure Uptake Biokinetic Model for Lead in Children (IEUBK), Windows version. 18,19 The overall goal was to gather information that could be used by policy makers and relevant stakeholders to protect children from the risks associated with ULAB recycling in Nairobi.

Methods
This cross sectional study was carried out from January to August, 2015. The study area was purposively selected based on the presence of informal ULAB recycling activities that were suspected to be potential sources of childhood exposure to lead. A control area with similar traffic and industrial activity was chosen in Ruiru as shown in Figure 1.
In this study, a dwelling unit refers to a non-conventional, slum-type lowcost housing unit usually constructed with non-conventional materials which are obtained through informal means.

House Dust
After homogenizing each house dust sample, a 2.5 g representative was weighed using an electronic analytical balance (Kern & Sohn, GMbH, Germany) and placed into a 50 mL glass beaker then 20 mL de-ionized water added and the sample placed in the fume extraction hood. Next, 20 mL of 69% concentrated nitric acid was added. After 8 hours, 2 mL of 37% concentrated hydrochloric acid and 3 mL of 30% hydrogen peroxide were added and the contents allowed to react for approximately 5 minutes, then heated on a hot plate to 180°C-200°C for 5 minutes. The temperature was maintained for 10 minutes, then the beaker contents allowed to cool. On cooling, the beaker contents were filtered, diluted with de-ionized water and brought up to volume in a 50 mL volumetric flask prior to analysis. 22 This is a typical digestion process to release lead from the dust wipe matrix and the soil matrix.

Soil Sample
For each soil sample, 0.5 g was weighed and placed into an inert polymeric microwave digestion vessel (Multiwave 3000, Anton Paar GmbH, Germany). The vessels were placed in the fume hood, then 5.0 mL of double distilled water, 9.0 mL of concentrated nitric acid (65%), 1.0 mL of concentrated hydrochloric acid (30%), and 2.0 mL hydrogen peroxide (30%) were added. Double distilled water was added to improve mineral solubility and prevent temperature spikes due to exothermic reactions. In order to allow gases to escape, each sample was allowed to react for approximately 5 minutes prior to sealing. The vessels were then placed on the rotor and placed in the microwave and heated between 180°C and 210°C for 5.5 minutes, then maintained at the same temperature for another 15 minutes. After cooling, the vessel contents were filtered and diluted with double distilled water to a volume of 100 mL in a volumetric flask prior to analysis.

Analysis of Lead
Standard working solutions were prepared from commercial stock solutions in order to calibrate the inductively coupled plasma-optical emission spectrometer (ICP OES). Calibration curves were produced with a standards concentration range of 0.0 mg/kg -20.0 mg/kg. Lead point sources of lead in the control community (Ruiru).

Floor Dust
The US Department of Housing and Urban Development dust wipe method was followed during collection and preparation of floor dust samples. 20 The method determines lead loadings on surfaces, allows for comparisons, and samples the surface that most likely is the source of children's exposure to lead laden dust. Results from dust wipe analyses correlate with children's BLL and can therefore be suitably used in the model to make predictions. 20, 21 The midpoint or largest area in the room was selected for floor dust sampling unless the children had a specific play area in the room, in this case the play area was considered. A one square foot plastic template was carefully placed on the sampling area on the floor without disturbing the dust.

Research
Analytical Chemistry Laboratories in Nairobi using similar procedures. Lead was detected using atomic absorption spectrometry and graphite furnace atomic absorption spectrometry in KIRDI and KEPHIS, respectively. The correlation coefficient between the sets of soil and house dust lead levels was 0.94 and 0.99, respectively. The study results were therefore deemed to be reliable. Calibration curves are shown in the Supplemental Material.

Estimation of Child Blood Lead Levels
From the measured floor dust lead loadings and soil lead concentrations, IEUBK version 1.1 Build 11 (Syracuse Research Corporation, North Syracuse, New York) was used to predict BLLs in children age ≤ 7 years. 18 The IEUBK model has been widely used to predict blood lead concentrations in young children exposed to lead. The model mathematically and statistically links environmental lead exposure to blood lead concentrations for one child or a population of children between the ages 0-7 years. The model uses exposure, uptake, biokinetic, and probability distribution to estimate blood lead levels in children exposed to lead contaminated media. The geometric mean blood lead is predicted from available information about children's exposure to lead such as soil and dust data. 18,19 The amount of lead in residential dust is quantified by lead loading (measured in micrograms per square meter (μg/m 2 ) or micrograms per square foot (μg/ft 2 )), and lead concentration measured in micrograms per gram (μg/g or ppm). Dust lead concentration is calculated from lead loading and dust loading as follows: dust lead concentration = Lead loading/Dust loading. 24 From this distribution, the model estimates the risk (probability) that a particular child or a population of children will have their blood lead concentrations exceed the Centers for Disease Control and Prevention reference values for lead in blood. 18 For this study, the model assumed that soil and dust are the only major means by which children come in contact with lead. Therefore, all other lead pathway (maternal, air, water and diet) values were set to zero and only dust and soil values were inputted. The default soil ingestion value was set as 500 mg/day for the dust and dirt environments studied.
Currently, there are no reference levels for pediatric blood lead or house dust lead in Kenya. Therefore, 10 μg/dL and 5 µg/dL were used for comparison, while 40 μg/ft² was used as a reference value for house dust lead loading on floors. 25-27 United States Environmental Protection Agency (USEPA) guidance values of 400 mg/kg lead in soils in residential areas, schools and playgrounds, and 1,200 mg/kg lead in waste dumps and industrial soils were used. 28

Statistical Methods
Statistical analyses were performed using Minitab version 17.0 (Minitab Inc.) The Ryan-Joiner test was used to test the normal/log-normal distribution of the data for soil and floor dust lead values in the areas studied. All data were log-transformed. The geometric means and medians for soil and house dust were calculated. Comparison of lead concentrations in house dust in different sampling sites was done using one-way analysis of variance. Statistical significance was set at p<0.05, unless otherwise stated.

Results
Lead was detected in 100% of floor dust samples. Each floor dust lead loading measurement from dwelling units and preparatory schools in the Dandora, Kariobangi and Mukuru concentrations in soil and floor dust digests were determined using the calibrated ICP OES (Spectro Arcos ICP Model FHS12, Germany). The method detection limit for lead was calculated as three times the standard deviation for the digestion blanks (n=5). 23

Quality Control and Assurance
Detailed standard procedures for collection, transport, and storage of samples were followed. 20 Analytical grade chemicals (Sigma-Aldrich Co, Germany) were used throughout the analyses. Deionized water was used throughout the analytical procedures.
Floor dust field blanks were prepared following similar procedures for collection of floor dust. 20 Lead concentrations in the floor dust field blanks ranged from not detected to 2.1 μg/ft 2 . Laboratory wipe sample blanks were also prepared and digested. Lead was not detected in any of the laboratory blank wipe samples.
Reagent blanks were similarly digested and analyzed with the samples. Lead was not detected in any of the reagent blanks.

Table 2 -Percentage Distribution of Floor Lead Loadings in the Study Areas Abbreviations: N, Number of Samples; bn, Number of Sampling Sites; c40 µg/ft 2 , US EPA guidance value for lead loading in indoor floor dust. 27
Study Area

Study Area
Research Control (CDC) reference value of 5 µg/dL for lead in blood. 26 Children living near informal ULAB activities in the Mukuru slums were predicted to have the highest geometric mean blood lead level followed by those living in Dandora, Kariobangi and Ruiru, consecutively. Accordingly, children living in the control area (Ruiru) were predicted to have a low mean blood lead level below the CDC recommended value ( Table 5).

Discussion
We found lead contamination exceeding USEPA reference values in outdoor soils and interior floor dust in children's environments located within two kilometers radius from informal ULAB recycling operations in the Dandora, Kariobangi and Mukuru slums. We hypothesize that this contamination is a result of the informal ULAB recycling activities that we observed being carried out, with no environmental or human exposure controls, in residential areas, near preparatory schools and near children's playgrounds in the study area ( Figure 2)

Study Area
Research prevailing winds (Figure 2) as well as ULAB recycling wastes dumped in the open near preparatory schools and children's playgrounds within residential areas (Figure 3). We also observed dust from the ULAB operations settling on surrounding soils and buildings. The lead contaminated soils and dust could become airborne when disturbed and blown by wind, causing widespread indoor and outdoor contamination. 32,33 Family members can also take home lead-contaminated soil and dust from outside when they enter the house without removing their shoes and/or with contaminated work clothing. 32 Additionally, we anecdotally observed that many floors in the dwelling units and preparatory schools in the Dandora, Kariobangi and Mukuru slums were dirty, with walls constructed using broken iron sheets with openings that may have allowed lead-containing dust to penetrate indoors. The walls and floor surfaces also had crevices into which lead containing dust could have embedded. We also anecdotally observed residents dry sweeping their dirty floors with children nearby, dust settled on toys, hands, pacifiers and similar objects, and children eating with dirty hands. Studies have shown that children's hand-tomouth activity and pica behavior in lead contaminated environments is associated with elevated BLL. 9, 21 The USEPA estimates that the typical 1-6 year old American child ingests between 100-400 mg of soil and house dust every 24 hours, with the highest ingestion rate at the age of two years. 34 In the dusty environments of our study communities, it may be reasonable to assume that children's soil and dust ingestion rates are even higher.
The mean lead loading in 73% of floor dust samples from dwelling units and preparatory schools in the control area (Ruiru) exceeded the USEPA 40 µg/ ft 2 guidance value for lead on floors. 27 Only 27% of the control floor dust samples recorded lead loading values that were below the regulatory limit ( Table 1). This could be attributed to the fact that lead particles are readily transported by air and wind, resulting in contamination of further away places. 35 Studies have shown the correlation of soil and house dust lead to blood lead in support of pica and handto-mouth routes of lead ingestion in children.

Conclusions
We found lead contamination in outdoor soil and interior floor dust in the vicinities of informal ULAB recycling operations in the Dandora,

Study Area
Research currently at risk of lead poisoning due to the unprecedented growth of informal ULAB recycling.
The soil and house dust results illustrate the need for coordinated action to decrease lead emissions from informal battery recycling in Nairobi slums and to remediate existing soils, particularly around battery workplaces and waste dumps. Child BLLs in the communities studied need to be tested in order to establish poisoning. Parents, guardians, teachers, and workers in institutions dealing with children should be encouraged to perform activities that have been shown to reduce children's lead exposures in lead-contaminated areas, including regularly wetmopping floors and other housing unit components, regularly washing children's hands, pacifiers and toys, preventing children from playing in bare or contaminated soil, moving children's play areas, daycares and preparatory schools away from waste dumps and areas where informal ULAB recycling activities are carried out, encouraging children and other adults to remove shoes and lead-dust contaminated clothing when entering the house to prevent tracking lead into houses, properly feeding children and improving children's environments to reduce exposure to lead. 7,30,32,44 Pronczuk J, Filipe Junior AP, Bertollini R, Neira M.

Table 5 -Predicted Blood Lead Levels in Children ≤ 7 Years Old Across Study Areas IEUBK was used to predict the BLL; 10 µg/dL (Centers for Disease Control, 2002) and 5 µg/dL (Centers for Disease Control, 2012) were used as the cut off/reference for child BLL. Abbreviations: GM, Geometric mean blood
lead level in children; *Geometric standard deviation = 1.60. **Child soil ingestion rate = 500 mg/day. 18,19,25,26