Science of the Total Environment 424 (2012) 63–73

Contents lists available at SciVerse ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Multi-trace element levels and speciation in urine of e-waste recycling workers from Agbogbloshie, in

Kwadwo Ansong Asante a,b, Tetsuro Agusa a, Charles Augustus Biney c, William Atuobi Agyekum b, Mohammed Bello b, Masanari Otsuka a,d, Takaaki Itai a, Shin Takahashi a, Shinsuke Tanabe a,⁎ a Center for Marine Environmental Studies (CMES), Ehime University, Bunkyo-cho 2-5, Matsuyama 790-8577, Japan b CSIR Water Research Institute, P. O. Box AH 38, Achimota, Accra, Ghana c Volta Basin Authority (VBA), 10 P. O. Box 13621, Ouagadougou 10, Burkina Faso d Ehime Prefectural Institute of Public Health and Environmental Science, 8-234 Sanban-cho, Matsuyama 790-0003, Japan article info abstract

Article history: To understand human contamination by multi-trace elements (TEs) in electrical and Received 31 January 2012 (e-waste) recycling site at Agbogbloshie, Accra in Ghana, this study analyzed TEs and As speciation in Received in revised form 25 February 2012 urine of e-waste recycling workers. Concentrations of Fe, Sb, and Pb in urine of e-waste recycling workers Accepted 27 February 2012 were significantly higher than those of reference sites after consideration of interaction by age, indicating that Available online 24 March 2012 the recycling workers are exposed to these TEs through the recycling activity. Urinary As concentration was rel- atively high, although the level in drinking water was quite low. Speciation analysis of As in human urine Keywords: Trace elements revealed that arsenobetaine and dimethylarsinic acid were the predominant As species and concentrations of E-waste both species were positively correlated with total As concentration as well as between each other. These results Urine suggest that such compounds may be derived from the same source, probably fish and shellfish and greatly in- As speciation fluence As exposure levels. To our knowledge, this is the first study on human contamination resulting from the Agbogbloshie primitive recycling of e-waste in Ghana. This study will contribute to the knowledge about human exposure to Ghana trace elements from an e-waste site in a less industrialized region so far scantly covered in the literature. © 2012 Elsevier B.V. All rights reserved.

1. Introduction one major destination for e-waste worldwide. In these developing countries, there are few infrastructure and protocols to safely recycle In the 21st century, electrical and electronic waste (e-waste) has and dispose of hazardous e-waste and legislation dealing specifically become an emerging environmental and human health problem in with their flow and regulations for maintaining the environment and the world (Schmidt, 2002, 2006). E-waste refers to the end-of-life human health are not always effective. For instance, recyclers use electronic products including televisions, monitors, computers, primitive methods (mechanical shredding and open burning) to audio and stereo equipment, video cameras, fax/photocopy machines remove insulation from cables. This technique can re- and printers, telephones, motherboards, mobile phones, chips, wire- lease highly toxic chemicals and severely affect the environment and less devices, cathode ray tubes, and other peripheral items (Frazzoli human health if improperly managed. Indeed, heavy contamination et al., 2010). These contain harmful chemicals such as brominated by TEs and persistent organic pollutants (POPs) such as polychlori- flame retardants (BFRs) and classical toxic trace elements (TEs), Pb, nated dibenzo-p-dioxins (PCDDs) and furans (PCDFs), and polybro- Cd, Hg, As, and Cr. Furthermore, rare TEs including Sb, In, and Tl, minated diphenyl ethers (PBDEs) in soil, water, air, and humans which have recently been a matter of concern because of their envi- from the recycling town of Guiyu in South China, which is one of ronmental behavior and toxic effects (Ha et al., 2009), are also present the largest recycling centers for e-waste in the world, were reported in e-waste. According to the United Nations Environment Programme by several studies (Schmidt, 2002; Wang et al., 2005; Deng et al., (UNEP) (2005),20–50 million tons of e-waste are generated annually 2006; Leung et al., 2006, 2007; Huo et al., 2007; Bi et al., 2007; Li in the world. Of these quantities, significant amounts of e-waste have et al., 2007; Wong et al., 2007a,b,c). been exported to developing countries such as China, India, Pakistan, Our research group investigated informal e-waste recycling sites in Vietnam, and the Philippines for recycling (UNEP, 2005). Recently, de- Bangalore, India (Ha et al., 2009) and the Red River Delta in Vietnam veloping nations of West Africa (e.g. Ghana and Nigeria) have become (Tue et al., 2010a, 2010b) and found significant contamination by TEs such as Pb, Sb, Cu, In, and Bi, PBDEs, and dioxin-like chemicals in soil, air, house dust, and humans. However, until now there is no available ⁎ Corresponding author. Tel./fax: +81 89 927 8171. study on the chemical contamination status at e-waste recycling sites E-mail address: [email protected] (S. Tanabe). in Africa. The United Nations' Basel Convention on the control of

0048-9697/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2012.02.072 64 K.A. Asante et al. / Science of the Total Environment 424 (2012) 63–73 transboundary movements of hazardous wastes and their disposal them in 1994 (Amoyaw-Osei et al., 2011). The scrap yard has grown governs how nations that have ratified it handle e-waste containing steadily into a popular recycling area and is reputed to be the dump- hazardous materials. Although Ghana ratified the Basel Convention ing ground for disused computers, TVs, and other electronic and elec- in 2005, its provisions are yet to be incorporated into a national legisla- trical devices as well as household waste (Fig. 1). Thus to date, tion and so the dangers posed to humans and the environment from Agbogbloshie has become the hub of informal ‘recycling’ industry in such e-waste are not properly addressed. For instance, there are no spe- Ghana. People trying to make ends meet manually disassemble com- cific regulations that restrict e-waste recycling and as such, e-waste re- ponents and heavily pollute the area through the burning of cables cyclers work in appalling conditions, exposing themselves and the and smashing of computer monitors and other electronic devices to environs to serious . Agbogbloshie in Accra is the graveyard of retrieve and Cu from in which they are encased to e-waste in Ghana. Heaps of e-waste are continually dumped there with- sell. While they busily and incessantly burn the cables, the immediate out any regard for the that they pose to the environment and environment is engulfed in thick black smoke, which takes hours to people living in the vicinity. clear. Because it is a continuous act done daily, there is no respite To explicate the exposure status of TEs in e-waste recycling for people living in the environs or those who move in and out of workers from Accra, Ghana, this study analyzed TE concentrations the area. Much of this activity is carried out by young men, mostly in urine from the workers. In addition, the quality of drinking water using rudimentary tools and with no protective equipment. Ghana sources was assessed as the potential exposure source of TEs in has an unregulated and unrestricted import regime for second hand local people. electrical and electronic equipment (EEE). Therefore, any e-waste could enter the country under the guise of second hand EEE without 2. Materials and methods detection. The demand for EEE in Ghana continues to grow by the day and in 2009, the EEE imports into Ghana added up to 2.1. Study areas 215,000 tons and a per capita import of 9 kg (Amoyaw-Osei et al., 2011). Most of the e-waste comes from Europe and North America Accra with a population of about 4 million is the capital of Ghana (Amoyaw-Osei et al., 2011). (population of about 25 million) and the largest city in terms of in- Obuasi, a district in the Ashanti Region of Ghana and 200 km dustrial establishment and infrastructural development. Over 70% of northwest of Accra, has been the center of large-scale gold mining ac- Ghana's manufacturing capacity is located in Accra. Agbogbloshie tivity since the late 19th century through AngloGold Ashanti Limited scrap market located in Accra is the main center for the recovery of (formerly Ashanti Goldfields Corporation, AGC). In 2004, AGC merged materials from e-waste. Situated on the bank of the Odaw River and with AngloGold of South Africa to create the world's second-largest in the upper reaches of the Korle Lagoon, the Agbogbloshie site gold producer, AngloGold Ashanti Limited. The main gold-bearing started as a stuff market for onions and yam. Over the years, it ore is arsenopyrite. Mining activity in Obuasi is known to have grew into a slum with people dealing in all kinds of scrap on a large given rise to substantial airborne As (Amasa, 1975). Thus, scale. The scrap dealers discovering the place as a good location for As contamination as well as other toxic elements in the environment business later registered with the National Youth Council as the of Obuasi arise as a result of the mining activity, in addition to natural Scrap Dealers' Association of Ghana, and the land was leased to processes of water–rock interaction and sulfide oxidation.

A B C

D E F

Fig. 1. Overview of an e-waste recycling site in Accra and procedure for recovering valuable metals from e-waste. A, Collected TVs. B, Working place. C, Discarded waste after recovery. D and E, Burning of e-waste to salvage Cu and other metals. Many people are working without mask, glove, and shoes. F, Collected Cu wire after burning. Workers pouring water onto heated wire to cool. K.A. Asante et al. / Science of the Total Environment 424 (2012) 63–73 65

2.2. Samples instrumental drift in the ICP-MS measurements were corrected by the internal standard method with Y as the internal standard. Concen- Thirty-eight drinking water samples including boreholes, wells, tration of Hg was determined by an atomic absorption spectrometer spring, stream, and tap were collected from Obuasi. The water sources (Shimadzu AA680, Shimadzu, Kyoto, Japan) connected with a cold were selected based on the water demand of the surrounding commu- vapor system (Model HG-3000, Sanso, Tsukuba, Japan)(CV-AAS) nities as well as the scarcity of water from the public supplies. All well (Asante et al., 2007). water samples were taken with a water sampler. The stream/spring To confirm the accuracy of our analytical methods, a certified samples were collected under base flow conditions about 20–30 cm reference material, SLRS-4 River Water from the National Research depth from the surface. Ten tap water samples were randomly taken Council Canada (NRCC) was analyzed. Results of TE concentrations from the environment of Accra, a distance of 200 km from Obuasi. The in the reference sample were in good agreement (79–130%) with pH and conductivity values of the water samples were measured in situ. the certified values. Spot urine samples were randomly obtained from 20 workers (all males) at the e-waste recycling site (ERW) in Agbogbloshie, Accra, 2.4. Urine analysis Ghana in August 2008. The ages of the workers varied between 15 and 42 years. As the reference site (Ref-1), spot urine samples were A closed vessel microwave system (Ethos D, Milestone S.r.l., Sorisole, obtained from 25 donors in Obuasi (22 females and 3 males) with BG, Italy) was employed for the digestion of human urine samples using ages ranging from 19 to 81 years. All the donors in Obuasi were not nitric acid. The digestion program applied was: 2 min at 250 W, 3 min engaged in mining activity and thus are defined as reference group at 0 W, 5 min at 250 W, 5 min at 400 W, 5 min at 500 W, 10 min at (Ref-1) of ERW. Also, urine samples were collected from two males 400 W, and 5 min for ventilation. Concentrations of 23 TEs (V, Cr, Mn, and a female living but not working as e-waste recyclers in Accra Fe, Co, Cu, Zn, Ga, As, Se, Rb, Sr, Mo, Ag, Cd, In, Sn, Sb, Cs, Ba, Tl, Pb, (Ref-2). The subjects for this study have lived at their respective sam- and Bi) and Hg were determined by ICP-MSs and CV-AAS, respectively pling sites for over 10 years, and for some since infancy. The informed (Asante et al., 2007). consent was obtained from all the donors in an ethical manner. De- Analysis of As speciation was performed according to the method tailed information (age, height, weight, body mass index (BMI), occu- of Agusa et al. (2009) and Tu et al. (2011). Urine filtered through a pation (Obuasi subjects), alcohol drinking and smoking habits, and 0.45 μm syringe-filter was diluted five times with Milli-Q water. Arse- frequency of fish consumption) were obtained using a questionnaire. nic compounds (arsenate (AsV), arsenite (AsIII) monomethylarsonic Mean age, body height and weight, BMI, alcohol drinking and smoking acid (MMA), dimethylarsinic acid (DMA), and arsenobetaine (AB)) habits, and frequency of fish consumption are shown in Table 1. About were indentified and quantified with ICP-MS connected to a high- 36% of all Obuasi subjects (Ref-1) were traders (n=9), and 32% were performance liquid chromatography (HPLC; Shimadzu, LC10A Series, retired workers (n=5). Other occupations included farmer (n=3), ca- Kyoto, Japan). To separate As compounds, an Inertsil AS column shier (n=2), teacher (n=2), driver (n=1), nurse (n=1), hairdresser (15 cm×2.1 mm i.d.; GL Sciences Inc., Japan) was used with the mo- (n=1),andseamstress(n=1). 13.3% and 11.1% of all the donors from bile phase (10 mM sodium 1-butanesulfonate, 4 mM tetramethylam- both Obuasi and Accra had smoking and alcohol habits, respectively. monium hydroxide, 4 mM malonic acid and 0.5% methanol; pH 3.0 All the water and urine samples were collected in polypropylene was adjusted with nitric acid). Flow rate of mobile phase, tempera- bottles pre-cleaned at the Center for Marine Environmental Studies ture of column oven, and injection volume were 0.5 ml/min, 45 °C, (CMES), Ehime University, Japan. The samples were not subjected to and 10 μl, respectively. Rhodium was added to the mobile phase as any chemical treatment. They were stored in cold boxes on the field an internal standard to monitor analytical interference. and on return to the laboratory in Ghana, kept in a freezer. The frozen A certified reference sample (NIES No. 18 human urine) provided samples were airlifted to CMES in Japan on dry ice and kept in the by the National Institute for Environmental Studies (NIES), Japan was Environmental Specimen Bank (es-BANK) of Ehime University at analyzed for confirmation of the methodological accuracy. Analyzed −25 °C (Tanabe, 2006) until chemical analyses. concentrations of AB and DMA were in good agreement with the certified values with the recoveries being 90–106%. 2.3. Water analysis 2.5. Statistical analysis The water samples were acidified with concentrated nitric acid. Concentrations of V, Cr, Mn, Co, Cu, Zn, Rb, Sr, Mo, Ag, Cd, In, Sn, Sb, Statistical analysis was performed using PASW Statistics (version Cs, Ba, Tl, Pb, and Bi were determined by an inductively coupled plas- 18.0, SPSS Inc., Chicago, IL, USA) and StatView (version 5.0, SAS Insti- ma mass spectrometer (ICP-MS; Hewlett-Packard, HP-4500, Avon- tute, Cary, NC, USA). One-half of the value of the respective limits of dale, PA, USA) as explained in our previous work (Asante et al., detection was substituted for those values below the limit of detec- 2007). For the analyses of Fe, Ga, As, and Se, ICP-MS of Agilent tion and applied in statistical analysis. Because most of the concentra- 7500cx (Avondale, PA, USA), which utilizes an octopole ion guide tions of TEs and As compounds in water and human urine samples did enclosed in a collision/reaction cell, was used. Matrix effects and not show normal distribution, the data was log-transformed to

Table 1 Information on subjects from Accra and Obuasi in Ghana.

Group Group Location Arithmetic mean and range Numbers Frequency ID Age Height Weight BMI Sex Smoking Alcohol Fish (year) (cm) (kg) (kg/m2) Male/female Yes/no Yes/no (Times/week)

E-waste recycling worker ERW Accra 27 163 67 25 20/0 5/15 1/19 5.0 (n=20) (15–42) (149–175) (48–98) (20–32) General population Ref-1 Obuasi 45 161 62 24 3/22 1/24 4/21 5.0 (n=25) (19–81) (145–171) (50–82) (19–33) General population Ref-2 Accra 30 160 62 24 2/1 0/3 0/3 5.3 (n=3) (16–40) (145–170) (40–85) (19–29) 66

Table 2 pH (pH units), conductivity (μS/cm), and concentrations (μg/l) of trace elements in drinking water from Accra and Obuasi in Ghana.

Location Source pH Conductivity V Cr Mn Fe Co Cu Zn Ga As Se Rb Sr Mo Ag Cd In Sn Sb Cs Ba Hg Tl Pb Bi

Accra Tap n 10/10 10/10 10/ 10/ 10/ 10/ 10/ 10/ 10/ 7/10 10/ 4/10 10/ 10/ 10/ 7/10 9/10 7/10 8/10 5/ 1/ 10/ 0/ 4/10 10/ 7/10 (Ref-2) 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 AM 7.77A 143 0.63 0.38 5.19 13 0.099 20.8 94.2 0.04 0.34 0.1 3.85 124 1.85 0.45 0.07 0.02 0.45 b0.1 b0.1 140 b0.5 0.02 0.40 0.02 SD 0.31 37 0.34 0.42 3.09 4.5 0.056 14.9 217.7 0.03 0.21 0.1 1.93 56 3.15 1.26 0.06 0.01 0.75 0.05 230 0.01 0.34 0.01 Min 6.98 81 0.28 0.04 1.18 5.3 0.049 5.89 7.85 b0.01 0.10 b0.1 1.09 56.8 0.22 b0.01 b0.01 b0.01 b0.01 b0.1 b0.1 28 b0.5 b0.01 0.09 b0.01 Max 8.11 185 1.2 1.5 10.0 21 0.22 52.0 711 0.10 0.76 0.4 6.75 197 10.7 4.0 0.19 0.05 2.40 0.16 0.1 760 b0.5 0.04 1.23 0.05 63 (2012) 424 Environment Total the of Science / al. et Asante K.A. GM 7.77 138 0.55 0.24 4.11 12 0.089 16.5A 28.7 0.03B 0.29B b0.1 3.40 112A 0.89A 0.04 0.05 0.01 0.11 b0.1 b0.1 75A b0.5 0.01 0.31 0.01 Obuasi Tap n 8/8 8/8 8/8 7/8 8/8 8/8 8/8 8/8 8/8 8/8 8/8 0/8 8/8 8/8 8/8 5/8 8/8 1/8 0/8 0/8 1/8 8/8 0/8 1/8 8/8 3/8 (Ref-1) AM 5.80B 156 0.29 0.18 19.5 140 0.45 2.72 47.2 0.14 1.44 b0.1 4.43 53.9 0.27 0.01 0.04 b0.01 b0.01 b0.1 b0.1 29 b0.5 0.01 0.23 b0.01 SD 0.81 40 0.21 0.11 30.8 300 0.97 2.92 47.7 0.18 2.12 2.98 26.5 0.16 0.01 0.03 26 0.02 0.19 Min 4.45 112 0.07 b0.01 2.91 6.1 0.03 0.88 4.10 0.02 0.28 b0.1 1.50 22.9 0.05 b0.01 0.01 b0.01 b0.01 b0.1 b0.1 13 b0.5 b0.01 0.05 b0.01 Max 6.55 214 0.59 0.36 94.7 870 2.8 9.71 155 0.58 6.63 b0.1 7.85 91.9 0.45 0.03 0.10 0.01 b0.01 b0.1 0.4 91 b0.5 0.06 0.57 0.01 GM 5.75 152 0.22a 0.12 10.2b 33 0.14b 1.98B 29.0 0.09a,A 0.84A b0.1b 3.46 48.5B 0.22B 0.01a 0.04 b0.01 b0.01 b0.1 b0.1 23B b0.5 b0.01b 0.17 b0.01 Borehole n 12/12 12/12 11/ 12/ 12/ 12/ 12/ 12/ 12/ 6/12 12/ 5/12 12/ 12/ 12/ 1/12 11/ 1/12 1/12 2/ 9/ 12/ 0/ 3/12 12/ 1/12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 AM 5.46 185 0.09 0.28 54.4 25 1.0 5.17 24.1 0.03 2.73 0.3 2.17 69.4 0.18 b0.01 0.04 b0.01 0.03 b0.1 0.2 38 b0.5 b0.01 1.86 b0.01 SD 0.17 115 0.10 0.19 39.6 18 0.82 4.29 13.3 0.04 4.83 0.7 1.48 47.3 0.10 0.02 0.09 0.2 25 5.90 Min 5.23 90 b0.01 0.08 3.64 4.5 0.22 1.26 8.31 b0.01 0.16 b0.1 0.55 21.3 0.08 b0.01 b0.01 b0.01 b0.01 b0.1 b0.1 12 b0.5 b0.01 0.01 b0.01 Max 5.88 419 0.40 0.76 142 58 2.7 17.4 47.7 0.10 16.6 2.4 5.88 162 0.40 0.01 0.09 0.03 0.31 0.12 0.6 85 b0.5 0.03 20.6 0.02 GM 5.46 161 0.06b 0.23 37.9a 19 0.79a 4.08 20.8 0.01b 1.03 0.1ab 1.76 56.0 0.16 b0.01ab 0.04 b0.01 b0.01 b0.1 0.2 31 b0.5 b0.01b 0.18 b0.01 Spring n 1/1 1/1 1/1 1/1 1/1 1/1 1/1 1/1 1/1 0/1 1/1 1/1 1/1 1/1 1/1 1/1 1/1 0/1 1/1 0/1 0/1 1/1 0/1 0/1 1/1 0/1 Concentration 5.08 138 0.10 0.04 112 11 1.6 1.81 25.3 b0.01 1.20 0.2 2.10 33.4 0.11 0.02 0.04 b0.01 0.36 b0.1 b0.1 33 b0.5 b0.01 0.11 b0.01 Stream n 1/1 1/1 1/1 0/1 1/1 1/1 1/1 1/1 1/1 0/1 1/1 0/1 1/1 1/1 1/1 0/1 0/1 0/1 0/1 0/1 0/1 1/1 0/1 0/1 1/1 0/1 Concentration 5.28 171 0.49 b0.01 7.49 42 0.05 0.82 0.80 b0.01 17.3 b0.1 0.29 15.2 0.09 b0.01 b0.01 b0.01 b0.01 b0.1 b0.1 5.5 b0.5 b0.01 0.01 b0.01 Well n 16/16 16/16 16/ 16/ 16/ 16/ 16/ 16/ 16/ 9/16 14/ 10/ 16/ 16/ 16/ 7/16 14/ 3/16 4/16 5/ 9/ 16/ 0/ 10/16 16/ 3/16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 AM 5.24 261 0.27 0.31 60.9 21 1.9 3.71 19.2 0.04 0.77 0.4 3.52 64.6 0.25 0.01 0.05 0.02 0.04 b0.1 0.2 49 b0.5 0.03 0.20 0.02 –

SD 0.51 188 0.23 0.30 47.5 16 1.7 2.33 14.5 0.06 0.80 0.7 2.31 64.8 0.19 0.02 0.05 0.04 0.11 0.2 37 0.04 0.13 0.04 73 Min 4.46 45 0.02 0.07 5.10 5.6 0.12 0.88 0.42 b0.01 b0.1 b0.1 0.71 6.92 0.05 b0.01 b0.01 b0.01 b0.01 b0.1 b0.1 5.1 b0.5 b0.01 0.05 b0.01 Max 6.20 784 0.69 1.3 171 53 6.5 9.10 51.5 0.25 3.04 2.9 8.18 273 0.75 0.07 0.19 0.16 0.44 0.28 0.5 140 b0.5 0.17 0.44 0.15 GM 5.22 205 0.16a 0.23 41.2a 16 1.1a 3.01 12.7 0.02b 0.47 0.2a 2.75 44.1 0.19 b0.01b 0.03 b0.01 0.01 b0.1 0.1 35 b0.5 0.02a 0.16 b0.01 n; Number of detected samples/number of analyzed samples. AM; Arithmetic mean. SD; Standard deviation. GM; Geometric mean. a, ab, and b; Different letters indicate significant difference in TE concentrations among borehole, tap, and well waters from Obuasi (pb0.05). A and B; Different letters indicate significant difference in TE concentrations in tap water between Accra and Obuasi (pb0.05). K.A. Asante et al. / Science of the Total Environment 424 (2012) 63–73 67

perform the parametric analysis. Differences in TE concentrations 0.01 0.01 among tap, well, and borehole were assessed by the analysis of variance b b (ANOVA), followed by the Tukey–Kramer method. The Student-t test was used to determine differences in concentrations of TEs and As compounds in tap water between Accra and Obuasi, or human urine between ERW and Ref-1 and between alcohol drinkers and non- 0.01 0.86 0.01 0.31 drinkers, as well as between smokers and non-smokers. When the com- 0.11 2.32 0.049 0.05b 4.24 0.008 b parison was targeted on male donors, the Mann–Whitney U-test was 0.5 0.18 4.27 0.106 0.5 0.32 7.27 0.159 0.5 0.11 2.98 0.062 0.5 0.090.5 2.34 0.20 0.014 4.61 0.114 0.5 0.50.5 0.18 0.04 18.3 6.06 0.035 0.013 0.5 0.5 0.06 7.30 0.015 conducted because of the small sample size of males in Ref-1 (n=3). b b b b b b b b b b The Pearson's correlation coefficient was applied to understand rela- tionships between age, BMI, frequency of fish consumption and concen- trations of TEs and As compounds in urine. Analysis of covariance (ANCOVA) was used for the evaluation of the regional and age interac- tion on TE and As species concentrations in urine. The difference in ratios of male to female among locations was checked by the χ2 test. A probability value of pb0.05 was considered as statistically significant. 0.1 0.67 0.24 b 3. Results and discussion

3.1. Concentrations of TEs, pH, and conductivity values in water samples 0.01 2.17 0.12 0.52 1.0 0.01 1.09 Concentrations of TEs, pH and conductivity values in drinking water b b are summarized in Table 2. Water pH and conductivity values ranged from 4.45 to 8.11 and from 81 to 784 μS/cm, respectively. Generally, 0.07 0.11 8.04 0.11 4.7 0.8 0.17 0.02 1.10 0.6 2.5 5.0 concentration of Sr (GM; 56.2 μg/l) was the highest in water, followed by Ba (GM; 36 μg/l), Mn (GM; 19.5 μg/l), Zn (GM; 18.7 μg/l), and Fe μ 0.01 0.16 0.14 4.50 0.19 12 9.1 0.01 0.12 0.08 2.07 0.13 8.3 8.3 0.01 0.17 0.16 7.03 0.2 13.0 9.2 0.010.01 0.02 0.01 0.77 0.37 0.08 0.02 5.71 3.19 3.0 0.89 11 5.3 20 4.2 0.01 0.43 0.03 3.34 1.10.01 6.0 0.01 5.9 (GM; 18 g/l), as our previous study (Asante et al., 2007) observed. b b b b b b b b Lower concentrations in water were observed for Sn (GM; 0.01 μg/l), Ag (GM; 0.01 μg/l), and Tl (GM; 0.01 μg/l). For In (GM; b0.01 μg/l), Sb (GM; b0.1 μg/l), Cs (GM; b0.1 μg/l), and Bi (GM; b0.01 μg/l), concentra- tions in almost all the water samples were below detection limits. Remarkably, Hg (GM; b0.5 μg/l) was not detected in all the water sam- ples. Therefore, these elements were not included in further discussion. Despite the tap water samples coming from the same water treat- ment source, large variations in TE concentrations were observed in each city (Table 2). Corrosion of household plumbing systems is an important source of TEs found in tap water (Tamasi and Cini, 2004). Metal (galvanized iron) pipe is sometimes used in water distribution system in Ghana, probably explaining such variations in TE levels in tap water. 157 31 561 107 50.7 Among tap, well, and borehole waters in Obuasi, significant concen- tration differences were observed for V, Mn, Co, Ga, Se, Ag, and Tl 0.01 147 78 2090 189 53.0 0.01 359 110 2760 309 122 0.01 0.25 0.28 16.3 0.34 18 9.8 0.01 43.7 50 1670 96.8 24.6 0.01 76.40.01 33 201 82 1370 96.7 2140 38.9 211 0.01 65.5 0.35 0.04 4.48 0.24 4.2 3.2 0.01 15.1 5.8 324 19.8 6.12 0.01 28.2 7.8 87 24.8 7.87 b b b b b b b (pb0.05) (Table 2). Concentrations of V in tap and well waters were higher than those in boreholes (pb0.05). For Mn (pb0.05) and Co (pb0.01), well and borehole showed significantly higher concentra- tions than tap water. Concentrations of Ga (pb0.01) and Ag (p b0.05) were the highest in tap water, while Se (pb0.05) and Tl (pb0.05) con- centrations in well water were higher than tap water or borehole. There

ERW

AsV AsIII Ref-1 MMA DMA 0.01 0.544 13 0.13 5.25 54.7

b AB Ref-2 3/3 3/3 3/3 3/3 3/3 3/3 3/3 0/3 3/3 3/3 3/3 3/3 3/3 1/3 3/3 3/3 3/3 3/3 3/3 3/3 0/3 3/3 3/3 3/3 25/25 24/25 25/25 25/25 25/25 25/25 25/25 5/25 25/25 25/25 25/25 25/25 25/25 6/25 25/25 20/25 25/25 21/25 25/25 25/25 5/25 23/25 25/25 9/25 20/20 20/20 20/20 20/20 20/20 20/20 20/20 7/20 20/20 20/20 20/20 20/20 20/20 0/20 20/20 18/20 20/20 20/20 20/20 20/20 0/20 17/20 20/20 16/20 g/l) of trace elements in urine of e-waste recycling workers and reference population from Accra and Obuasi. μ

0 20406080100 GM 3.1 18 4.05 57 0.85 278 675 Max 4.9 26 7.90 190 1.9 442 964 SDMin 1.3 2.5 7 14 2.89 2.65 94 29 0.8 0.38 219 128 446 259 MaxGM 170 5.2 29 15 11.1 3.47 540 130 8.7 1.6 490 254 2483 0.15 614 222 0.01 64 43.4 29 3980 389 1760 99.3 463 83.7 MaxGM 340n 16AM 20 3.3 2.2 8.44 19 2.54 260 44 2.1 4.57 85 0.61 481 77 3363 1.0 0.29 713 295 174 709 100 6030 354 129 0.24 1.62 0.44 14.1 1.4 15 9.4 34 0.59 33.7 0.369 SDMin 37 2.6n 7 0.18AMSD 0.76 80 2.39Min 110 19 1.7 150 8.3 7.5 0.041 2.2 2.99 7.12 1.78 117 59 178 54 520 0.79 0.56 0.04 171 163 45.6 1017 751 15 0.03 0.07 85.8 1070 40.15 41 108 27 112 1920 1470 113 70 50.5 33.7 0.03 0.06 0.51 0.40 0.09 0.11 5.28 3.06 0.32 0.27 5.4 3.6 3.8 2.1 2.5 7.0 0.16 0.15 3.84 6.40 0.049 0.092 n AM 13 19 4.08 180 2.4 305 752 0.03 54.4 33 2090 142 121 Composition of As compounds in urine (%)

Fig. 2. Composition (arithmetic mean) of arsenicals in human urine from ERW, Ref-1, Ref-2 Ref-1 Group IDERW V Cr Mn Fe Co Cu Zn Ga As Se Rb Sr Mo Ag Cd In Sn Sb Cs Ba Hg Tl Pb Bi ; Number of detected samples/number of analyzed samples. n AM; Arithmetic mean. SD; Standard deviation. GM; Geometric mean. Table 3 Concentrations ( and Ref-2. 68

Table 4 Results of ANCOVA.

VCr MnFeCoCuZnAsSeRbSrMoCd

FpF pFpFpF pFpFpFpFpFpFpFpFp

ANCOVA Location 0.001 0.981 0.084 0.774 1.131 0.294 1.041 0.314 9.321 0.004 0.203 0.655 3.442 0.071 9.392 0.004 0.050 0.824 0.558 0.459 0.112 0.740 3.426 0.071 0.118 0.733 Age 4.156 0.048 4.301 0.044 4.618 0.038 1.347 0.253 17.745 0.000 4.525 0.040 0.215 0.645 4.946 0.032 0.004 0.952 1.678 0.203 0.024 0.878 2.686 0.109 0.202 0.655 Location×Age 0.030 0.863 0.073 0.789 0.952 0.335 0.000 0.996 7.067 0.011 0.000 0.995 2.914 0.095 5.335 0.026 0.003 0.959 0.497 0.485 0.084 0.773 1.458 0.234 0.220 0.642

Error pooled ..Aat ta./Sineo h oa niomn 2 21)63 (2012) 424 Environment Total the of Science / al. et Asante K.A. Location 0.247 0.622 2.103 0.154 0.189 0.674 7.603 0.009 2.148 0.150 1.506 0.227 0.507 0.481 5.475 0.024 0.222 0.640 0.063 0.804 0.032 0.860 3.744 0.060 0.059 0.809 Age 9.598 0.004 10.525 0.002 4.269 0.045 2.785 0.103 9.515 0.004 9.340 0.004 5.369 0.025 0.629 0.432 0.001 0.974 1.286 0.263 0.265 0.609 1.218 0.276 0.028 0.867 Numbers in bold indicate statistical significance at pb0.05.

ResultsTable 4 of(continued ANCOVA.) – 73 In Sn Sb Cs Ba Tl Pb Bi AsV AsIII MMA DMA AB

FpFpF p FpFpFpF pFpFpFpFpFpF p

ANCOVA Location 1.706 0.199 1.698 0.200 9.874 0.003 0.207 0.652 0.063 0.803 5.581 0.023 5.298 0.027 1.032 0.316 1.032 0.316 0.397 0.532 0.091 0.764 2.141 0.151 14.538 0.005 Age 0.020 0.890 0.007 0.933 1.636 0.208 1.351 0.252 2.604 0.114 0.003 0.957 10.768 0.002 2.603 0.114 2.603 0.114 1.468 0.233 1.078 0.305 0.958 0.333 7.888 0.008 Location×Age 0.344 0.561 0.289 0.594 2.437 0.126 0.160 0.691 0.079 0.780 2.019 0.163 2.706 0.108 0.206 0.652 0.206 0.652 0.000 0.985 0.036 0.851 1.664 0.204 8.839 0.005

Error pooled Location 4.175 0.047 4.654 0.037 19.648 b0.001 0.051 0.822 0.001 0.980 7.545 0.009 4.105 0.049 2.549 0.118 0.551 0.462 3.036 0.089 1.652 0.206 0.499 0.484 6.667 0.013 Age 0.628 0.433 0.180 0.674 0.058 0.810 1.582 0.215 4.102 0.049 1.774 0.190 8.636 0.005 7.658 0.008 9.034 0.005 3.036 0.086 2.824 0.100 0.008 0.927 0.835 0.366 K.A. Asante et al. / Science of the Total Environment 424 (2012) 63–73 69 were no significant differences in pH and conductivity among water were comparable to those reported in highly As-contaminated ground- source types (p>0.05). water areas in the world, suggesting the presence of other sources (ex- cept water) of As contamination in Ghana (Asante et al., 2007). In the 3.2. Regional difference of TEs in tap water present study, drinking water samples also had low As concentrations (up to 17 μg/l) in Accra and Obuasi (Table 2), but urinary As concentra- Regional differences in TE concentrations between Accra and tions in donors were relatively high (Table 3). Obuasi were assessed by using tap water sample. Concentrations of To identify the exposure source of As in humans, analysis of As Cu (pb0.001), Sr (p=0.002), Mo (p=0.009), and Ba (p=0.010) in compounds can be a useful tool, because As speciation is different tap water from Accra were significantly higher than those from among materials. In the present study, As speciation analysis revealed Obuasi, while the opposite results were found for Ga (p=0.047) that all As compounds were detected in all urine samples. AB (AM; and As (p=0.012) (Table 2). 36.2%) and DMA (AM; 35.1%) were the predominant As compounds Compared to previous studies from Obuasi (Amonoo-Neizer and in urine of all the subjects, followed by AsV (AM; 14.7%), AsIII (AM; Amekor, 1993; Smedley, 1996) and our study from Tarkwa, another 8.9%), and MMA (AM; 5.2%) (Fig. 2). Furthermore, AB (r=0.876, mining town in Ghana (Asante et al., 2007), As concentrations in the pb0.001) and DMA (r=0.798, pb0.001) concentrations were posi- water types in this study were relatively low. Concentration of Mn was tively correlated with total As concentration as well as each other significantly lower than our previous study in Tarkwa (Asante et al., (AB and DMA; r=0.561, pb0.001), suggesting that these compounds 2007). This is expected and relates to the difference in geology between may be derived from the same source and greatly influenced As expo- Obuasi and Tarkwa. Large deposit of Mn in Ghana is found in the West- sure levels in the residents. ern Region of which Tarkwa is a district. contamination of wa- It is considered that AB is derived from the consumption of fish ters, fish and crops through amalgamation by small-scale miners has and shellfish, which have high concentration of AB. People who con- been reported in Au mining areas of Ghana (Golow and Mingle, 2003; sume seafood, e.g., crab and lobster, excrete AB as the dominating As Adimado and Baah, 2002; Golow and Adzei, 2002), but Hg was not species in urine within 3 days (Chen et al., 2010). Smedley (1996) detected in any of the water samples. Probably because our detection also found low contribution of inorganic As to total As in human limit (0.5 μg/l) for Hg was high compared with other studies (Babut et urine from Obuasi, Ghana and thus suggested that high concentration al., 2003; Golow and Mingle, 2003), we could not detect Hg in this study. of total As in urine and low concentration of total As in drinking water Generally, the pH values of the Obuasi samples were acidic (4.5–6.5) reflect high consumption rate of fish. In Ghana, fish is recognized as and fell below the WHO guideline of 6.5–8.5. pH was more acidic in tap the most important source (over 60%) of animal protein in all regions water from Obuasi compared with Accra (pb0.001) (Table 2). Further- and mean consumption of fish, seafood, and seafood products is 78 g/ more, all water samples from Obuasi had lower pH (4.5–6.5) compared day (Aggrey-Fynn, 2001; FAO, 2010). Hata et al. (2007) reported that with Accra tap water. Smedley (1996) also reported that pH of ground- elevated concentration of AB (173.3 μg/l) in urine observed in Japanese water samples in Obuasi was low (in the range of 3.9 and 6.8) of which may be due in part to high seafood consumption (182 g/day). the acidity relates to the high rainfall and resultant short residence In a recent study involving the general population, Navas-Acien times of many of the groundwaters in the aquifer. et al. (2011) indicated that recent seafood intake contributed to in- Because local people use these waters as drinking among other creased urine DMA concentration which is consistent with human uses, human health risk through the exposure to TEs from drinking water is of great concern. The drinking water guideline values estab- 103 lished by the World Health Organization (WHO) are available for Cr (50 μg/l), Mn (400 μg/l), Fe (300 μg/l), Cu (2000 μg/l), As (10 μg/l), 102 Se (10 μg/l), Mo (70 μg/l), Cd (3 μg/l), Sb (20 μg/l), Ba (700 μg/l), Hg (1 μg/l), and Pb (10 μg/l) (WHO, 2004). Water concentrations of Cr, 1 Mn, Cu, Se, Mo, Cd, Sb, and Hg did not exceed the guideline values 10 (WHO, 2004). On the other hand, Fe in a tap water, As in a stream and a borehole, Pb in a borehole from Obuasi and Ba in a tap water 100 from Accra had levels over the WHO guideline values (WHO, 2004), suggesting that these waters are not suitable for drinking. However, -1 because the ratios of samples which exceeded the benchmarks of 10 Fe, As, Ba, and Pb to all samples (n=48) were less than 4%, in general, drinking water in Accra and Obuasi is not seriously polluted by TEs. 10-2 Fe*** Pb* Sb***

3.3. Concentrations of TEs in human urine 102

Concentrations of TEs in urine of e-waste recycling workers and its reference populations are shown in Table 3.Ingeneral,Rb(GM; 101 Max 1560 μg/l) was the highest concentration in the human urine, followed GM μ μ μ

by Zn (GM; 668 g/l), Cu (GM; 137 g/l), Sr (GM; 102 g/l), and Fe (GM; Concentration of TEs in urine (µg/l) Min μ μ μ 95 g/l). On the other hand, Tl (GM; 0.065 g/l), In (GM; 0.034 g/l), Bi 100 (GM; 0.015 μg/l), Ga (GM; 0.010 μg/l), and Ag (GM; 0.007 μg/l) concen- trations in urine were relatively low. Almost all donors had urinary Hg, ERW Ga and Ag concentrations below detection limits and thus, we excluded -1 10 Ref-1 these elements in further discussion. This trend was concordant with results of our previous study in Ghana (Asante et al., 2007). 10-2 3.4. Concentrations and composition of arsenicals in human urine Sn* Tl***

Fig. 3. Comparison of TE concentrations in human urine between ERW and Ref-1. TEs In our previous study, although As concentrations in drinking water with significant regional difference confirmed by the ANCOVA and the Student's in Tarkwa were much low, urinary As levels of non-mining workers t-test are shown in this figure. 70 K.A. Asante et al. / Science of the Total Environment 424 (2012) 63–73 experimental studies reporting increased urine concentrations of AsV (Table 4). Concentrations of Cr, Mn, Cu, Ba, Pb, Bi, and AsV in DMA and other As species after seafood (seaweed, fish, and shellfish) urine showed negative correlations with age regardless of location intake (Choi et al., 2010), most likely due to human metabolism of difference (pb0.05), suggesting that these TEs are low in older people. arsenosugars and arsenolipids (Francesconi et al., 2002; Schmeisser Positive correlations observed between age and V or Zn indicate age- et al., 2006). Other studies also found increased urinary excretion of dependent accumulation. DMA after seafood consumption (Arbouine and Wilson, 1992; Concentrations of Fe, In, Sb, Tl, and Pb significantly differed be- Heinrich-Ramm et al., 2001, 2002). tween locations regardless of age (pb0.05) (Table 4); ERW>Ref-1 However, fish consumption frequency obtained by the questionary for Fe, Sb, and Pb concentrations; ERWbRef-1 for Sn and Tl concen- investigation was not clearly associated with urinary AB and DMA con- trations by the Student's t-test (Fig. 3). For In, the concentration dif- centrations in this study (p>0.05) as observed in a Taiwanese study by fered between locations by the error pooled ANCOVA (Table 4), but Hsueh et al. (2002). More quantitative data on fish consumption is not by the Student's t-test (p>0.05). However, it should be empha- therefore required to validate human exposure to As compounds. sized that the ratio of male to female in ERW group was significantly different from those in Ref-1 group (pb0.001) (Table 1); all subjects 3.5. Age dependency of TEs and arsenicals in human urine in ERW were male workers, but females dominated in Ref-1. Here, we also compared concentrations of TEs and arsenicals in urine be- Among the factors including age, BMI, habits of alcohol intake and tween ERW and Ref-1 using only male subjects by the Mann–Whitney smoking, and fish consumption frequency, only age was significantly U-test to remove the interaction of sex with the regional difference. associated with concentrations of TEs in human urine in the present From the result, higher concentrations in urine of ERW subjects than study. Age dependency on urinary TE concentrations was assessed Ref-1 were observed for Fe, Sb, and Pb (pb0.05). These were consistent by the Pearson's correlation coefficient. Positive correlations were ob- with the results (Table 4 and Fig. 3) shown above, indicating that, at served between age and concentrations of V (r=0.513, pb0.001), Zn least, ERW subjects are exposed to Fe, Sb, and Pb through the incessant (r=0.314, p=0.030), As (r=0.300, p=0.039), and AB (r=0.347, recycling activity at the e-waste recycling site. On the other hand, p=0.016). On the contrary, Cr (r=−0.582, pb0.001), Mn (r=−0.409, human urinary concentrations of As, AB, and Tl in Ref-1, were significant- p=0.004), Fe (r=−0.471, pb0.001), Co (r=−0.553, pb0.001), Cu ly high compared with those in ERW. Because significant interaction of (r=−0.558, pb0.001), Mo (r=−0.362, p=0.011), Sb (r =−0.321, age with regional difference was considered for As and AB (Table 4), re- p=0.026), Ba (r=−0.353, p=0.014), Pb (r=−0.578, pb0.001), Bi sults for only Tl in males conformed with results in whole donors (r=−0.338, p=0.019), AsV (r=−0.526, p=0.004), AsIII (r=−0.406, (Table 4 and Fig. 3). The results are suggestive that there are some p=0.004), and MMA (r=−0.379, p=0.008)concentrationsinurine sources of Tl in Ref-1. These regional differences observed in humans were negatively correlated with age. (Table 4 and Fig. 3) were not consistent with those in tap water However, age was significantly different among locations (pb0.001); (Table 2), indicating other exposure sources except drinking water in subjects of Ref-1 were older than ERW and Ref-2 groups (Table 1). There- these populations. In case of Pb at the e-waste recycling site, our recent fore, regional differences in urinary TE concentrations may be overlapped study revealed high concentrations (up to 14,000 μg/g) in soil/ash mix- by the results of relationships between age and TEs. The possibility of tures (Otsuka et al., 2012), indicating that such an environmental mate- apparent results on age dependency on TE concentrations is examined rial is a potential exposure source to e-waste recycling workers. below. 3.7. Comparison of concentrations of TEs and arsenicals in human urine 3.6. Regional difference of TEs and arsenicals in human urine with other studies

Significant differences in urinary TE concentrations were observed To understand the contamination status by TEs in workers of the between ERW and Ref-1 (Table 3). The Student's t-test revealed that e-waste recycling site in Ghana, urinary TE concentrations in the pre- concentrations of Cr (p=0.002), Fe (pb0.001), Co (p=0.002), Cu sent study were compared with other studies in Ghana (Asante et al., (p=0.004), Mo (p=0.005), Sb (pb0.001), Pb (p=0.022), AsV 2007; Basu et al., 2011; Kwansaa-Ansah et al., 2010; Paruchuri et al., (p=0.016), AsIII (p=0.004), and MMA (p=0.016) in urine of sub- 2010; Smedley, 1996) and an e-waste recycling site in China (Wang jects from ERW were significantly higher than those in Ref-1, while et al., 2011)(Table 5). the opposite results were found for V (p=0.027), As (p=0.002), In general, concentrations of Co, Se, Pb, Cu and Zn in the present AB (pb0.001), Se (p=0.003), Sn (p=0.027), and Tl (p=0.004). study were in the range of previous studies except for Cu and Zn con- The former results indicate that ERW subjects may be exposed to centrations reported by Wang et al. (2011) from China. For Mn, the these elements in the e-waste recycling site through the recycling pro- concentration was similar to other reports except for results in cess. On the other hand, specific exposure sources in Ref-1 (Obuasi) are Tarkwa which showed high concentration of Mn in human urine suggested by the later findings. (Asante et al., 2007). Although there were no significant differences Since age of donors in Ref-1 was significantly high compared with in urinary concentrations of V, Cr, and Mo between ERW and Ref-1 that in ERW (Table 1) and significant age dependent accumulation of by ANCOVA followed by the Student's t-test, level of V from Ref-1 TEs was observed as we have mentioned above, whether or not there are population was high among the studies, but low for Cr and Mo. Urinary interactions with age dependency and regional difference on concentra- Sn and Ba concentrations in this study were higher than other studies, tions of urinary TEs and As compounds needs to be validated. Here, to while Rb, Ag, Hg, and Tl showed relatively low levels in this study. assess the regional difference in more detail, we used ANCOVA with Concentration of Sr in the present study except for Ref-2 was also low age as covariance (Table 4). Significant interactions between location compared with other studies. For In, Cs, and Bi, the concentrations and age were detected for Co, As, and AB (Table 4); In only ERW, urinary were relatively high in Ref-2 among our results. Co concentration was negatively associated with age (r=−0.610, Wang et al. (2011) reported that Cd level in both occupational dis- p=0.003), while positive correlations with age were observed for As mantling and non-occupational dismantling people from e-waste (r=0.477, p= 0.034) and AB (r=0.607, p=0.004). These results sug- recycling site in China was higher than that in the control group, sug- gest site-specific age dependency of TEs and arsenical concentrations gesting Cd contamination at the e-waste site. Our previous study an- in human, although the reason is unclear. alyzed hair samples from informal e-waste recycling workers in The error pooled ANCOVA revealed each association of location or Bangalore, India and suggested Cd exposure (Ha et al., 2009). However, age with levels of TEs and As compounds (Table 4). Significant rela- in the present study, there was no significant difference in Cd concen- tions with age were observed for V, Cr, Mn, Cu, Zn, Ba, Pb, Bi, and tration between ERW and Ref-1 and the level was low compared with Table 5 Comparison of trace element concentrations (μg/l) in human urine of this study with those of other studies.

Location Remarks V Cr Mn Fe Co Cu Zn Ga As Se Rb Sr Mo Ag Cd In Sn Sb Cs Ba Hg Tl Pb Bi References

E-waste recycling ERW GM 5.2 15 3.47 130 1.6 254 614 0.01 43.4 29 1760 99.3 83.7 b0.01 0.37 0.02 3.19 0.89 5.3 4.2 b0.5 0.04 6.06 0.013 This study site in Accra, (19.3)a Ghana Obuasi, Ghana Ref 1 GM 16 2.2 2.54 44 0.61 77 713 b0.01 76.4 33 1370 96.7 38.9 0.01 0.35 0.04 4.48 0.24 4.2 3.2 b0.5 0.09 2.34 0.014 This study (13.5)a Accra, Ghana Ref 2 GM 3.1 18 4.05 57 0.85 278 675 b0.01 147 78 2090 189 53.0 b0.01 0.16 0.14 4.50 0.19 12 9.1 b0.5 0.18 4.27 0.106 This study (16.2)a Taizhou, Zhejiang Occupational GM 2.04 0.72 3.46 0.3 1.62 Wang et al. province in dismantling (2011) southeast China Taizhou, Zhejiang Non- GM 2.37 0.5 4.11 0.33 1.79 Wang et al. province in occupational (2011) ..Aat ta./Sineo h oa niomn 2 21)63 (2012) 424 Environment Total the of Science / al. et Asante K.A. southeast China dismantling Taizhou, Zhejiang Control group GM 1.64 0.27 2.07 0.28 1.9 Wang et al. province in (2011) southeast China Tarkwa, Ghana Mine workers Median 4.1 22 7.4 1.7 421 620 260 68 2270 229 60.8 0.31 0.62 0.03 2.15 0.2 6.3 1.5 3.0 0.41 5.13 0.01 Asante et al. (2007) Tarkwa, Ghana Non-mine Median 3.5 22 7.33 2.1 377 695 240 62 3580 163 57.8 0.51 0.61 0.02 1.43 0.1 7.2 2.8 3.6 0.31 3.08 0.02 Asante et al. workers (2007) Accra, Ghana Non-mine Median 4.9 24 4.16 1.8 614 598 200 73 2290 263 89.1 b0.01 0.36 b0.01 1.5 0.1 5.7 1.7 2.8 0.21 1.55 b0.01 Asante et al. workers (2007) Mining community Miners, males GM 24.17 1.16 1.79 37.33 437.21 94.33 31.29 0.38 1.12 Basu et al. in the Talensi– (2011) Nabdam District in the Upper East region of Ghana World Bank, Ghana Miners Median 1.83 Paruchuri et al. (2010) Obuasi Ghana Miners Median 15.8 Paruchuri et al. (2010) Kejitia, Ghana Miners Median 38.9 Paruchuri et al. (2010) Upper Denkyira Farmers, males Median 0.51 Kwansaa-Ansah –

District, Ghana et al. (2010) 73 Upper Denkyira Farmers, Median 0.66 Kwansaa-Ansah District, Ghana females et al. (2010) Upper Denkyira Farmers, males Median 1.23 Kwansaa-Ansah District, Ghana and females et al. (2010) Upper Denkyira Miners, males Median 0.58 Kwansaa-Ansah District, Ghana et al. (2010) Upper Denkyira Miners, Median 0.86 Kwansaa-Ansah District, Ghana females et al. (2010) Upper Denkyira Miners, males Median Kwansaa-Ansah District, Ghana and females et al. (2010) Wumase, Obuasi, Water source; Median 297 Smedley Ghana stream (42)a (1996) Kolfkurom, Obuasi, Water source; Median 224 Smedley Ghana borehole (18)a (1996)

GM; Geometric mean. a Inorganic As concentration in bracket. 71 72 K.A. Asante et al. / Science of the Total Environment 424 (2012) 63–73 the study results by Wang et al. (2011) as well as Tarkwa, a mining in sampling. We gratefully acknowledge the assistance of the late Dr. town in Ghana (Asante et al., 2007)(Table 5). Stephen Dapaah-Siakwan, former director of the CSIR Water research On the other hand, higher concentration of Sb in human urine Institute, medical doctors and staff of Obuasi Government Hospital from ERW in this study than those from other locations in Ghana and St. Jude Hospital, Obuasi as well as workers at the Agbogbloshie (Asante et al., 2007) was observed (Table 5), suggesting specific ex- e-waste dumping site for their co-operation during the sampling cam- posure to Sb in e-waste recycling workers in Accra. Antimony is a paign. Our appreciation also goes to Dr. Jiro Ogawa, in-charge of the human and causes diseases of liver, skin, and respiratory Environmental Specimen Bank (es-BANK), CMES, Ehime University, and cardiovascular systems (Kuroda et al., 1991; Schnorr et al., Japan for his support in sample management. We thank Dr. Annamalai 2007). The main sources of Sb in the environment are human indus- Subramanian, CMES, Ehime University, Japan for critical evaluation of trial activities, including coal combustion, mining and smelting of the manuscript. This study was supported by the Global COE Program mineral ores, and the use of Sb products such as flame retardants from the Ministry of Education, Culture, Sports, Science and Technolo- and alloys (Wu et al., 2011). In cases of occupational exposure, Sb gy (MEXT), Japan and Steel Foundation for Environmental Protection mainly enters the human body by inhalation and skin contact (De Technology, Japan. Boeck et al., 2003). Because e-waste contains solder or flame retar- dants, Sb exposure to the e-waste recycling workers and its related References human health effects are of concern. Concentration of As in urine from ERW in Accra was lower than the Adimado AA, Baah DA. Mercury in human blood, urine, hair, nail and fish from the levels from other areas in Ghana, although there was no significant dif- Ankobra and Tano River Basins in Southwestern Ghana. Bull Environ Contam – ference in the level between ERW and Ref-1 in this study (Table 5). Toxicol 2002;68:339 46. Aggrey-Fynn E. The contribution of the fisheries sector to Ghana's economy. A paper Smedley (1996) reported not only total As but also inorganic As concen- prepared on behalf of the FAO as an input into the sustainable fisheries livelihoods trations in human urine from Wumase and Kofikurom villages near study; 2001. Obuasi in Ghana. Compared with the concentrations of urinary inorgan- Agusa T, Kunito T, Minh TB, Trang PTK, Iwata H, Viet PH, et al. Relationship of urinary arsenic metabolites to intake estimates in residents of the Red River Delta, Vietnam. ic As in the study by Smedley (1996), the level in the present study was Environ Pollut 2009;157:396–403. comparable to that from Kofikurom, but lower than that from Wumase Amasa SK. Arsenic pollution at Obuasi goldmine, town and surrounding countryside. (Table 5). For comparison of organic As compounds, there are no avail- Environ Health Perspect 1975;12:131–5. Amonoo-Neizer EH, Amekor EMK. Determination of total arsenic in environmental able data in Ghana as well as other e-waste recycling sites till now. samples from Kumasi and Obuasi, Ghana. Environ Health Perspect 1993;101:44–9. Amoyaw-Osei Y, Agyekum OO, Pwamang JA, Mueller E, Fasko R, Schluep M. Ghana 4. Conclusions e-waste country assessment. SBC e-waste Africa project; 2011. 111 pp. Arbouine MW, Wilson HK. The effect of seafood consumption on the assessment of occupational exposure to arsenic by urinary arsenic speciation measurements. J This study is the first to investigate human contamination resulting Trace Elem Electrolytes Health Dis 1992;6:153–60. from the primitive recycling of e-waste in Ghana and arguably from Af- Asante KA, Agusa T, Subramanian A, Ansa-Asare OD, Biney CA, Tanabe S. Contamination rica. We found elevated Fe, Sb, and Pb concentrations in urine of status of arsenic and other trace elements in drinking water and residents from Tarkwa, a historic mining township in Ghana. Chemosphere 2007;66:1513–22. e-waste recycling workers in Accra compared with the reference site Asante KA, Adu-Kumi S, Nakahiro K, Takahashi S, Isobe T, Sudaryanto A, et al. Human after consideration of the confounding factor, age. Results of TEs con- exposure to PCBs, PBDEs and HBCDs in Ghana: temporal variation, sources of expo- – centrations in drinking water did not support the levels in humans. sure and estimation of daily intakes by infants. Environ Int 2011;37:921 8. Babut M, Sekyi R, Rambaud A, Potin-Gautier M, Tellier S, Bannerman W, et al. Improving These indicate exposure to these metals by the workers through the the environmental management of small-scale gold mining in Ghana: a case study e-waste recycling process but not drinking water. of Dumasi. J Cleaner Prod 2003;11:215–21. However, this study has some limitations. For instance, sample Basu N, Nam D-H, Kwansaa-Ansah E, Renne EP, Nriagu JO. Multiple metals exposure in a small-scale artisanal gold mining community. Environ Res 2011;111:463–7. bias regarding sex between ERW and Ref-1 should be addressed in Bi X, Thomas GO, Jones KC, Qu W, Sheng G, Martin FL, et al. Exposure of electronics future research. To characterize significant contamination source of dismantling workers to polybrominated diphenyl ethers, polychlorinated biphenyls, TEs in e-waste recycling workers, chemical analyses of dust or soil sam- and organochlorine pesticides in South China. Environ Sci Technol 2007;41:5647–53. Chen LWL, Lu X, Le XC. Complementary chromatography separation combined with ples are necessary. Indeed, high concentrations of TEs were detected in hydride generation-inductively coupled plasma mass spectrometry for arsenic soil/ash mixtures from this e-waste recycling area (Otsuka et al., 2012). speciation in human urine. Anal Chim Acta 2010;675:71–5. Furthermore, as shown in Fig. 1, not only adults but also children are Choi BS, Choi SJ, Kim DW, Huang M, Kim NY, Park KS, et al. Effects of repeated seafood working to collect reusable resources at the e-waste recycling site. consumption on urinary excretion of arsenic species by volunteers. Arch Environ Contam Toxicol 2010;58:222–9. Because it is well known that children are much more sensitive to De Boeck M, Kirsch-Volders M, Lison D. Cobalt and antimony: genotoxicity and carci- chemical hazard than adults, risk assessment of humans, especially for nogenicity. Mutat Res Fundam Mol Mech Mutagen 2003;533:135–52. children at the recycling site is required to regulate and manage the Deng WJ, Louie PKK, Liu WK, Bi XH, Fu JM, Wong MH. Atmospheric levels and cytotoxicity of PAHs and heavy metals in TSP and PM2.5 at an electronic waste recycling site in recycling activity. southeast China. Atmos Environ 2006;40:6945–55. On the other hand, contamination by other chemicals in the FAO (Food and Agriculture Organization of the United Nations). Food consumption in e-waste recycling site is of concern. Recent findings have also indicated food security data and definitions. http://www.fao.org/economic/ess/ess-fs/fs- data/ess-fadata/en/2010. that e-waste and urban areas are the main sources of BFRs in the rural Francesconi KA, Tanggaar R, McKenzie CJ, Goessler W. Arsenic metabolites in human areas (Asante et al., 2011). Hence, in the future, BFRs as well as dioxins urine after ingestion of an arsenosugar. Clin Chem 2002;48:92-101. and furans should be monitored at the Agbogbloshie e-waste site since Frazzoli C, Orisakwe OE, Dragone R, Mantovani A. Diagnostic health risk assessment of electronic waste on the general population in developing countries' scenarios. the primitive and hazardous techniques used for recycling e-waste Environ Impact Assess Rev 2010;30:388–99. could to elevated levels in the ambient environment and higher Golow AA, Adzei EA. Mercury in surface soil and cassava crop near an alluvial goldmine body burdens of these contaminants in residents and workers at the at Dunkwa-on-Offin, Ghana. Bull Environ Contam Toxicol 2002;69:228–35. e-waste site. This study will contribute to the knowledge about Golow AA, Mingle LC. Mercury in river water and sediments in some rivers near Dunkwa-On-Offin, an Alluvial Goldmine, Ghana. Bull Environ Contam Toxicol human exposure to trace elements from an e-waste site in a less indus- 2003;70:379–84. trialized region so far scantly covered in the literature. Ha NN, Agusa T, Ramu K, Tu NPC, Murata S, Bulbule KA, et al. Contamination by trace elements at e-waste recycling sites in Bangalore, India. Chemosphere 2009;76: 9-15. Acknowledgments Hata A, Endo Y, Nakajima Y, Ikebe M, Ogawa M, Fujitani N, et al. HPLC– ICP-MS speciation analysis of arsenic in urine of Japanese subjects without occupational exposure. J – The authors wish to thank Mrs. Gifty Afua Asante (Provident Occup Health 2007;49:217 23. Heinrich-Ramm R, Mindt-Prufert S, Szadkowski D. Arsenic species excretion in a group Insurance Company Ltd.) and Mr. Henry Asabore Boham (Town and of persons in northern —contribution to the evaluation of reference Country Planning, Adenta Municipal Assembly, Accra) for their help values. Int J Hyg Environ Health 2001;203:475–7. K.A. Asante et al. / Science of the Total Environment 424 (2012) 63–73 73

Heinrich-Ramm R, Mindt-Prufert S, Szadkowski D. Arsenic species excretion after con- Tamasi G, Cini R. Heavy metals in drinking waters from Mount Amiata (Tuscany, Italy): trolled seafood consumption. J Chromatogr B 2002;778:263–73. possible risks from arsenic for public health in the province of Siena. Sci Total En- Hsueh YM, Hsu MK, Chiou HY, Yang MH, Huang CC, Chen CJ. Urinary arsenic speciation viron 2004;327:41–51. in subjects with or without restriction from seafood dietary intake. Toxicol Lett Tanabe S. Environmental specimen bank in Ehime University (es-BANK), Japan for 2002;133:83–91. global monitoring. J Environ Monit 2006;8:782–90. Huo X, Peng L, Xu X, Zheng L, Qiu B, Qi Z, et al. Elevated blood lead levels of children in Tu NPC, Agusa T, Ha NN, Tuyen BC, Tanabe S, Takeuchi I. Stable isotope-guided analysis Guiyu, an electronic waste recycling town in China. Environ Health Perspect of biomagnification profiles of arsenic species in a tropical mangrove ecosystem. 2007;115:1113–7. Mar Pollut Bull 2011;63:124–34. Kuroda K, Endo G, Okamoto A, Yoo Y, Horiguchi S. Genotoxicity of beryllium, gallium Tue NM, Sudaryanto A, Minh TB, Isobe T, Takahashi S, Viet PH, et al. Accumulation of and antimony in short-term assays. Mutat Res 1991;264:163–70. polychlorinated biphenyls and brominated flame retardants in breast milk from Kwansaa-Ansah EE, Basu N, Nriagu JO. Environmental and occupational exposures to women living in Vietnamese e-waste recycling sites. Sci Total Environ 2010a;408: mercury among indigenous people in Dunkwa-On-Offin, a small scale gold mining 2155–62. area in the south-west of Ghana. Bull Environ Contam Toxicol 2010;85:476–80. Tue NM, Suzuki G, Takahashi S, Isobe T, Trang PTK, Viet PH, et al. Evaluation of dioxin- Leung A, Cai ZW, Wong MH. Environmental contamination from electronic waste recy- like activities in settled house dust from Vietnamese e-waste recycling sites: rele- cling at Guiyu, southeast China. J Mater Cycles Waste Manage 2006;8:21–33. vance of polychlorinated/brominated dibenzo-p-dioxin/furans and dioxin-like Leung AOW, Luksemburg WJ, Wong AS, Wong MH. Spatial distribution of polybromi- PCBs. Environ Sci Technol 2010b;44:9195–200. nated diphenyl ethers and polychlorinated dibenzo-p-dioxins and dibenzofurans United Nations Environment Programme (UNEP). DEWA/GRID-Europe, e-waste, the hid- in soil and combusted residue at Guiyu, an electronic waste recycling site in southeast den side of IT equipment's manufacturing and use. Chapter 5—early warning on China. Environ Sci Technol 2007;41:2730–7. emerging environmental threats. http://www.grid.unep.ch/product/publication/ Li H, Yu L, Sheng G, Fu J, Peng P. Severe PCDD/F and PBDD/F pollution in air around an download/ew_ewaste.en.pdf, 2005. electronic waste dismantling area in China. Environ Sci Technol 2007;41:5641–6. Wang D, Cai Z, Jiang G, Leung A, Wong MH, Wai KW. Determination of polybrominated Navas-Acien A, Francesconi KA, Silbergeld EK, Guallar E. Seafood intake and urine con- diphenyl ethers in soil and sediment from an electronic waste recycling facility. centrations of total arsenic, dimethylarsinate and arsenobetaine in the US popula- Chemosphere 2005;60:810–6. tion. Environ Res 2011;111:110–8. Wang H, Han M, Yang S, Chen Y, Liu Q, Ke S. Urinary heavy metal levels and relevant Otsuka M, Itai T, Asante KA, Muto M, Tanabe S. Process Trace element contamination factors among people exposed to e-waste dismantling. Environ Int 2011;37:80–5. around the e-waste recycling site at Agbogbloshie, Accra City, Ghana. Interdisciplinary WHO (World Health Organization). Guidelines for drinking-water quality, 3rd ed., Vol. Studies on Environmental Chemistry, Vol. 6, Advanced Environmental Studies by 1. Geneva: World Health Organization; 2004. Young Scientist. Kwaguchi M, Misaki M, Sato H, Yokokawa T, Itai T, Tue NM, Ono Wong CSC, Duzgoren-Aydin NS, Aydin A, Wong MH. Evidence of excessive releases of J, Tanabe S. (Eds), TERRAPUB, Tokyo, Japan 2012:161–7. metals from primitive e-waste processing in Guiyu, China. Environ Pollut Paruchuri Y, Siuniak A, Johnson N, Levin E, Mitchell K, Goodrich JM. Occupational and 2007a;148:62–72. environmental mercury exposure among small-scale gold miners in the Talensi–Nab- Wong CSC, Wu SC, Duzgoren-Aydin NS, Aydin A, Wong MH. Trace metal contamination of dam District of Ghana's Upper East region. Sci Total Environ 2010;408:6079–85. sediments in an e-waste processing village in China. Environ Pollut 2007b;145:434–42. Schmeisser E, Goessler W, Francesconi KA. Human metabolism of arsenolipids present Wong MH, Wu SC, Deng WJ, Yu XZ, Luo Q, Leung AOW, et al. Export of toxic chemicals— in cod liver. Anal Bioanal Chem 2006;385:367–76. a review of the case of uncontrolled electronic-waste recycling. Environ Pollut Schmidt CW. E-junk explosion. Environ Health Perspect 2002;110:A188–94. 2007c;149:131–40. Schmidt CW. Unfair trade: e-waste in Africa. Environ Health Perspect 2006;114:A232–5. Wu F, Fu Z, Liu B, Mo C, Chen B, Corns W, et al. Health risk associated with dietary co- Schnorr T, Steenland K, Thun M, Rinsky R. Mortality in a cohort of antimony smelter exposure to high levels of antimony and arsenic in the world's largest antimony workers. Am J Ind Med 2007;27:759–70. mine area. Sci Total Environ 2011;409:3344–51. Smedley PL. Arsenic in rural groundwater in Ghana. J Afr Earth Sci 1996;22: 459–70.