Exploratory Health Assessment of Chemical Exposures at E-Waste Recycling and Scrapyard Facility in Ghana

It is estimated that 28 million metric tons of electronic and electrical waste, or e-waste, is generated across the globe each year with approximately 65% originating from Europe and North America.1 E-waste and its adverse effects on health and the environment has yet to be fully assessed and continues to be extensively studied.2,3 Manual processing of e-waste may result in exposure to a variety of toxins, including: heavy metals, acid gases, polycyclic aromatic hydrocarbons (PAHs) and dioxins. Experience from China and India indicate that activities at e-waste sites, including dismantling and burning, produce hazardous emissions that may have damaging health effects.4 Workers at these sites are exposed to dust via inhalation, ingestion and dermal contact, which may contain harmful levels of heavy/trace elements. For example, in Guiyu, a thriving area of illegal e-waste recycling in China, it is reported that 80% of children exposed to unsafe e-waste recycling practices suffer from respiratory diseases and are often exposed to harmful heavy elements such as lead. Computer waste disposed of in landfills are known to produce contaminated leachates which may pollute groundwater and water supplies. Guiyu is facing acute water shortages due to the contamination of water resources, in part attributable to e-waste. Other studies have shown that e-waste recycling sites pose major threats to waterways by contamination of nearby streams and rivers. Heavy elements and inorganic acids can leach into waterways through wastewater or ambient air emissions and run the risk of contaminating natural resources such as soil, crops, drinking water, fish and livestock.5 

In 2003, Ghana formulated its policy on information and communications technology (ICT) for accelerated development, which represents Ghana's vision in the Information Age.6 Among the specific objectives of Ghana's policy is the application of ICT to support the modernization of the civic, public, educational, social and health services. Given this initiative, the country has aggressively pursued ICT development. However, at present Ghana does not manufacture nor assemble the electronic equipment or computers to support realization of its ICT policy, and as such, all equipment is imported into the country. This new and high level of demand has led to an alarming influx of electric and electronic equipment (EEE) into Ghana. Unfortunately, only 30% of the imports are new, with the remaining being used equipment that was either donated or purchased from high-income countries. A considerable portion of this used EEE is inoperable and scavenged for useable components such as hard drives, power supplies and circuit boards. At present, Ghana has no formal facility for managing the processing and disposal of such e-waste but despite this, the volume of such imports is increasing.7 

In 2009, about 171,000 tons of e-waste from consumers, repair shops and communal collection reached the informal recycling sector of Agbogbloshie in the capital city of Accra.8 Initial surveys at the main dump sites in Agbogbloshie revealed that some institutions transport truckloads of e-waste for free or a nominal fee. In addition to this flow, other e-waste scavengers travel throughout Accra soliciting e-waste collection and disposal. Once at Agbogbloshie, a rather formal hierarchy exists of bidders, processors and copper wire recyclers. Operations at the site include dismantling of heavy equipment, engines, motors, refrigerators and copper recovery. The dominant exposure of concern at Agbogbloshie is the open burning of wires and equipment for copper recovery. In addition, cleaning and processing of used lead-acid vehicle batteries for export constitutes an on-going activity that has the potential to pose risks to human health and the environment.

The rate of e-waste dumping and processing at Agbogbloshie is alarming and increasing. Environmental health assessments at this site are limited, but in 2009 researchers quantified airborne and soil lead levels, revealing significantly elevated levels of lead in soil and in the downwind plume from the site where burning took place.8 Elevated urinary arsenic levels in e-waste workers at Agbogbloshie were also found with no conclusive source pinpointed.9 Additional exposure to toxic airborne agents from the burning of plastics, including dioxins, dibenzofurans and PAHs is inescapable but not reported.

In an effort to better understand the public health consequences of these exposures, the Ghana Health Service, with support from the health and environmental not-for-profit Green Advocacy Ghana (GreenAd), the Ghana Environmental Protection Agency and U.S.-based Hunter College of the City University of New York, School of Public Health, conducted an exploratory health assessment. The overall objective of the study was to assess and describe the health status and extent of exposure of handlers of e-waste to the chemicals associated with electronic waste. Specifically, the study aimed:

• —To determine level of knowledge, perceptions and practices in relation to health risks associated with exposure to e-waste recycling among e-waste handlers at Agbogbloshie;

• —To describe the general state of health of these workers; and

• —To determine their body burden of chemicals composed of heavy elements and trace elements.

This exploratory cross-sectional study was carried out at two markets in Accra: namely Agbogbloshie, where e-waste dumping and recycling activities occur, and the Makola market, which is free of e-waste dumping and recycling activities (i.e., the control population).

The geographical location of Ghana is shown in Figure 1. Images of Agbogbloshie are shown in Figure 2. E-waste is discarded in the scrap metal portion of the Agbogbloshie market, which is separated into two main areas. The forward area is located at the front of the market where numerous electronic/electrical items and car parts, such as batteries and engines, are hauled in, dismantled, sold and traded. In this area, women and children also cook and sell food to shoppers. The rear area, where processing occurs, is located on the edge of the market; here, materials considered by scrap dealers to be of no value can be found scattered over a large field. Workers also use a portion of the field to build small fires to burn plastic off electrical wires and coils in order to recover valuable metals, such as copper. Less than 1 kilometer south of the Agbogbloshie market and across the Odaw river, in the pathway of the toxic plume from the burning wires, lies an informal squatter settlement. The site was not included in this study but will be the focus of future investigation and assessment.

A non-random but representative sample of 87 participants was recruited from Agbogbloshie, and an equallysized comparison group similarly selected from Makola, for a total 174 participants. Each prospective participant was informed of the study objectives and and human subject procedures as mandated by the Ghanaian Health Service were followed.

For the administrative staff, a 2-day orientation was held to provide training for 10 persons (5 from Ghana Health Service, 2 from the Ghana EPA, and 3 from GreenAd). Participating staff included: 3 public health practitioners, 1 toxicologist, 3 environmental scientists, 1 social scientist, 2 laboratory technicians and a nurse. The training included questionnaire administration and, for the medical staff, procedures on collecting biological samples. Both didactic and hands-on field instructional exercises were included, terminating with a supervisory assessment. The various stages in data collection are given below:

Questionnaire Administration: This was done to assess knowledge, perceptions and practices regarding safe work practices and potential for poisoning from chemicals during e-waste handling. This was carried out over a 3-day period. The questionnaire was adapted to expedite and focus on the study's goals from a larger and more extensive health survey assessment used by the Ghana Health Service since 2003. Pre-testing on 15 individuals was implemented to ensure adequate comprehension and data quality collection. The final survey questionnaire was only administered to the exposed population.

Health Status Determination (E-waste Handlers): Participants were screened for general health indices as well as potential clinical symptoms and signs suggestive of exposure from heavy metals. This was done by taking a medical history and conducting a physical examination.

Biological Sampling: Determination was completed of serum and urine levels of 11 heavy metals and trace elements (barium, cadmium, chromium, cobalt, copper, iron, lead, manganese, mercury, selenium and zinc).

a. Blood Samples: Venous blood samples were collected by trained and experienced phlebotomists using approved Vacutainers and sterile methods under the supervision of the GHS. Ten ml of whole blood was collected per subject, which was then spun to isolate cells for serum analysis. Sealed blanks were incorporated into the analysis for quality control and to ensure no container-related cross-contamination.

b. Urine Samples: Participants were informed that urine samples would be collected at the beginning of the day, thereby representing a first-day collection of urine samples. Each participant was handed a sealed 120ml urine specimen container to fill. Upon completion, a medical assistant transferred 10ml to a sealed container for analysis. All containers were sterile disposables designed for single use and approved for medical collections.

In collecting blood and urine samples, specific precautions had to be taken to avoid contamination. In addition to investigators using aseptic techniques, the participants washed their hands thoroughly with soap and water before collecting urine. The samples were packed in a cool box containing ice packs to maintain the integrity of the samples. Samples were brought to the Ghana Standards Board Forensic Lab in Accra and spun to isolate cells to produce serum.

The lab implemented appropriate quality-assurance measures and trained staff to ensure that contamination of collected samples were minimized by avoiding sources of trace metal contamination such as rubber, wood, paper products, metal surfaces, skin and hair. In all instances, sample collection equipment and containers—including needles and caps—were pre-screened or soaked in trace metal-grade nitric acid (Fisher Scientific, Loughborough, UK) to ensure that trace/heavy metal contamination was minimized.

The Ghana Standards Board Forensic Laboratory has adopted the practice of analyzing serum, as opposed to whole blood, for heavy and trace metals. The analysis of serum was deemed appropriate in that the standards were also in blood serum. This practice, while not common in high-income countries, has been reported elsewhere.10,11 Plasma and serum may be more relevant indicators of human toxicity given easier distribution to organs.12,13 However, caution is urged in comparing these data to data from whole blood testing reported elsewhere or to normal clinical values.

Blanks were measured alongside urine and blood serum samples. The metallic contaminants laboratory is ISO 17025 certified to measure Cd, Hg and Pb in water and fish. The competence of the lab in measuring these metals gave us confidence in its ability to analyze the remaining elements. Throughout the analysis, quality assurance and control procedures were followed to demonstrate the accuracy and precision of the results. After every 10 sample runs, the working standards, spiked urine samples and certified reference materials—where appropriate—were run to ensure the accuracy of the results. All analyses were done on a Graphite Furnace AAS (Perkin Elmer).

The limits of detection of each of the metals were established by calculating the mean of 30 individual readings of the blank solutions. The limit of detection (LOD) was obtained by multiplying the value by 3. Certified reference materials in serum were available for cadmium, mercury and lead.14 For the other metals, a range of urine samples were spiked with known amounts of heavy/trace metals. These were spiked at 5, 10 and 20ug/g and the average recoveries calculated ranged between 88%–103%. The serum and urine samples were prepared for analysis by dilution with 10% (v/v) nitric acid (Fisher Scientific) in de-ionized water.

All data was entered into SPSS^® v.16 (IBM, Armonk, USA) then used for descriptive frequency analysis, as well as cross-tabulations, analysis of variance, chi-square and Student's t-test. This involved analysis of data pertaining to demographic characteristics, knowledge, perceptions and practices in relation to e-waste handling, health status and prevalence of elevated body levels of heavy metals and trace elements among participants. Excel v.14 (Microsoft, Redmond, USA) was used to produce tables and charts.

Results were reviewed in the light of results of soil and air sample analysis that had been collected in a related study by a team from the City University of New York School of Public Health. Air samples had been obtained by placing air monitor cassettes on workers and collecting breathing zone samples for a 3-hour period per subject.8 

Data from the Makola market group, who had not been exposed to e-waste, was compared to that obtained from the exposed group for a simple, descriptive comparison of similarities or differences.

Before contact with potential subjects, the survey was approved in writing by the Ethical Review Committee of the Ghana Health Service. The study was carried out in compliance with Ethical Review Committee conditions, which seek to protect the interest and rights of study subjects as far as practicable. This included: ensuring that no procedure caused harm or distress to the potential subjects; integrating the process of consenting subjects prior to enrolment with study procedures; and ensuring that the confidentiality of all subjects who agreed to take part in this study would be protected to the fullest extent possible.

The sample size of the study population was limited, which made it difficult to investigate significant associations between exposure and outcome variables. Furthermore, challenges in locating reference standards for some tests (using urine) meant that objective comparisons cannot be made between test results obtained from these and other studies. Finally, as stated earlier, the determination of heavy and trace metals is often done in whole blood (blood cells and the serum/plasma portion of blood). This study looked at heavy and trace metals in the serum/plasma portion of human blood and caution is urged in comparing the results with international standards and other studies.

The ages of the 87 e-waste exposed subjects from Agbogbloshie ranged from 15 to 73 years. The mean age was 32 years with a standard deviation of 5.6 (Table 1). The mean age for the Makola group was 25.5 years and the standard deviation 5.6. The age differences between each group was not statistically different (p>0.5).

In terms of gender distribution of participants, 83 (95.4%) from Agbogbloshie and all 87 (100%) from Makola were male.

With regard to educational level of the participants (Table 1), the 87 e-waste workers had generally more education: 54% of the Agbogbloshie participants had some formal education, as opposed to 30% for the Makola participants. Not only were the e-waste workers, in general, more educated, but they also completed higher academic levels (secondary and tertiary) as opposed to the non e-waste workers: 41.4% versus 4.6%.

The distribution of participants by their home region is also shown in Table 1. The significance here is confirmation that most workers at Agbogbloshie (e-waste site) come from the northern part of Ghana.

All participants in the exposed group from Agbogbloshie had been involved in e-waste handling for at least a year and were still involved in it at the time of the study. Additionally, no individuals from Makola worked at the e-waste site. The mean period of time the participants had been involved in e-waste recycling was 7.8 years, with a range of 1 to 33 years. The daily routine of 60 participants involved collection, while 58 were involved in dismantling of electronic wastes, 24 in burning, 5 in collection of copper from ashes after burning, 1 in lead smelting, and 12 in other activities in the same location, as shown in Figure 3. Individuals in the non-exposed group were not asked their perceptions on risk and therefore that information is not reported here.

Of Agbogbloshie respondents involved in burning, 11 have been involved with the process for a period of 3 to 5 years, 7 for a period of 6–10 years, 4 for 11–15 years, and 2 for 2 years or less. One respondent has been involved in the process for more than 20 years. Of Agbogbloshie respondents involved with ash/wire collection after burning, 2 have been involved with the process for a period of 3–5 years, 2 have been doing it for a period of 6–10 years and 1 respondent has been involved with the process for more than 11–15 years.

Sixty-two (or 71.3%) of Agbogbloshie respondents reported no knowledge of possible hazards they could come in contact with at work. The remaining had knowledge of hazards as follows: 6 (6.9%) mentioned acid, dirty oil and powder; 5 (5.7%) mentioned knowledge of smoke or fumes; 3 (3.4%) mentioned knowledge of airborne fibers or dust, while 3 (3.4%) mentioned other gases.

Fifty-four (or 62.2%) of all participants did not have knowledge of illnesses that could arise from their work. Seventeen (19.5%) participants mentioned bodily pains, while 12 (13.8%) mentioned headaches. Other commonly mentioned illnesses included: eye injury by 5 (5.7%); chest problems by 5 (5.7%); itching/rashes by 4 (4.6%); cough by 4 (4.6%); and joint pains, nausea and painful urination mentioned by 1 (1.1%) subject.

Among the Agbogbloshie participants, 44 (50.6%) reported they were sure that there are ways of protecting themselves during work to avoid being adversely affected by e-waste hazards. Thirty-nine (44.8%) did not think they could prevent effects of any e-waste hazards connected with their work, and 4 (4.6%) were not sure whether it is possible to prevent such effects or not.

For those who believed that symptoms caused by hazards related to e-waste work could be prevented, 20 (23.0%) mentioned hospital care, 5 (5.7%) answered protective clothing, 4 (4.6%) said education, 3 (3.4%) mentioned gas masks and prevention of contact, while 1 (1.1%) mentioned good nutrition.

Of all Agbogbloshie participants, 70 (80.5%) took no precautions with regard to work, 18 (20.7%) used protective clothing, including 2 who wore gloves and 1 who used goggles when working. Two (2.3%) took medication comprising orthodox drugs (Figure 4).

Table 2 summarizes the physical examination findings for the exposed and unexposed groups. Fifty-nine (67.8%) of the exposed group fell within normal weight range, with body mass indexes (BMI) ranging from 18.5 to 24.9. Twenty-two (25.3%) were overweight, with BMI of 25–29.9, 2 (2.3%) were obese with BMI greater than 29.9, and 1 was underweight with a BMI of less than 18.5. For the non-exposed group, 72 (82.8%) had normal weight and 15 (17.2%) were overweight. None were underweight or obese.

The mean BMI for the exposed and non-exposed group was 24.2 and 23.3, respectively. A Students t-test was done to test for any significant difference among these groups in terms of BMI. Without assuming equal variances among these two groups the calculated p value was 0.099 given α=0.05 (t0.025,109= −1.68). This means there was no significant difference between the two groups in terms of BMI at 0.05 significance level.

Out of the 87 exposed participants, 1 (1.1%) looked pale (a sign of possible anaemia) and 1 looked slightly jaundiced. Two (2.3%) had small scars in the neck region resulting from wounds, which they explained had been due to a laceration and septic spots, respectively. Three of them (3.4%) had lost 1 or 2 teeth, while 1 (1.1%) had prominent dental cavities. All participants had a normal gait.

Skin: Forty-six (52.8%) had no signs of abnormality of the skin while the remaining 41 (47.2%) had various abnormalities. The latter comprised of 22 (25.3%) with fungal rashes, while 2 (2.3%) had other kinds of rashes, 8 (9.2%) had patches of darkened skin and 5 had hypo-pigmented or lightened skin patches, 2 of which were a result of bleaching. Four (4.6%) had at least one scar on the chest, torso or limbs.

Chest Examinations: Eighty-three (95.4%) participants had no obvious abnormalities of the chest, both on physical appearance and on auscultation. Of the remaining 4, 2 had a few rhonchi (signs of wheezing suggestive of mild bronchitis, asthma or other obstructive airway pathology.) There was a slight reduction in air entry for the other 2.

Cardiovascular System: Blood pressure was normal for 77 (88.5%) and slightly elevated above the threshold diastolic value of 140mmHg in 10 (11.5%). The liver was just palpable in 2 (2.3%) of participants while there was no organ enlargement in the remaining 85 (97.7%) of them. Heart sounds were normal in all participants.

Musculoskeletal: Limb movements were normal. No tenderness or sensory deficits were noted in the upper or lower limbs.

Tables 3 and 4, respectively, presents the urine and blood serum levels of the 11 heavy and trace metals in exposed and unexposed participants together with confidence intervals and statistical results. An independent sample t-test (assuming unequal variance among the exposed and unexposed sample populations) was done and reported in Tables 3 and 4. Table 5 summarizes these findings, indicating up and down arrows for statistically significant results. An uparrow indicates an elevated level in the exposed group, while a down-arrow indicates a lower level.

The exposed group had elevated urine levels of barium, manganese, selenium and zinc compared to the unexposed group (Table 5). However, significantly lower levels of cadmium, chromium, iron and mercury were found among the e-waste workers.

For blood serum levels (Table 5), 7 of the 11 elements were significantly higher statistically in the exposed population as opposed to the unexposed. The remaining 4 (cadmium, manganese, mercury and lead) were all either undetected or not significantly different. In sum, heavy or trace elements in blood serum were either all elevated or inconclusive. None of the elements were elevated in the non-exposed group.

This study attempted to define and describe e-waste handlers at Ghana's Agbogbloshie market in terms of their demographic characteristics, as well as their knowledge, perceptions and practices in relation to health risks associated with exposure to e-waste recycling.

This cohort of workers involved in e-waste handling was composed of mostly young men from the northern region of Ghana, notably from the Tamale municipality, Tolon/Kumbungu and Savelugu/Nanton districts. Many have had no formal education and have been involved in the business of e-waste collection and dismantling for several years. Many lacked knowledge about health risks posed by their occupation and what they could do to protect themselves. A group of manual workers from Makola market who were used for comparison hailed mostly from the northern and upper-eastern regions.

Most of the participants who underwent the examinations had a healthy demeanor. They all had normal affect, understood questions asked, and cooperated well with the research assistants and examiners. None of them reported to be suffering from any chronic disease. The majority were of normal body weight; however, over 25% were overweight and 2 (2.3%) were obese. Blood pressure was normal in the majority (88.5%). The most frequently reported health problem identified was skin abnormalities (47.2%), the most common of which comprised fungal rashes (25.3%)

The results of the urine assay for heavy and trace metals showed mixed results, with several elements, namely cobalt, chromium, iron and mercury, found at higher levels in the unexposed group. Urine testing can produce variable results indicative of short-term exposure. Additionally, the analytes were measured as milligrams per liter of urine and not the more standard milligrams per gram of urinary creatinine. Thus, the lack of a standardized metric of urinary strength may lead to intermittent results. Caution should be taken in interpreting the urinary data presented in Table 3. However, given the exploratory nature of the study, it does offer guidance for future testing.

Blood serum levels are somewhat more consistent and indicative of worker exposure. Barium, cobalt, chromium, copper, iron, selenium and zinc were all significantly elevated in the e-waste cohort. All these elements are released at Agbogbloshie during the combustion of plastic-coated wires for copper recycling. Air sampling done earlier seems to confirm release of selected metals into the community.8 

However, these observations require careful interpretation considering the fact the estimations are bound to be very sensitive to reference standards used for the analysis. Additionally, the unavailability of universal limits for normal or acceptable physiologic concentrations for some of these heavy and trace metals in some body fluids puts a limitation on the toxicological interpretation of urinary and blood serum levels.

In the related environmental study carried out simultaneously, air samples were taken to assess chemical contamination of workers' breathing zones and ambient air in the environment. Soil samples collected around the perimeter of the area used by the scrap and e-waste dealers were also analyzed for heavy metals and trace elements. Due to sampling limitations, worker-breathing zone samples yielded insufficient data; however, ambient environmental air samples revealed that levels of aluminum, copper, iron, lead and zinc were above the U.S. Occupational Safety and Health Administration permissible exposure levels/standards.15 

The results for lead appear problematic yet can be understandable. Low levels of lead in blood serum are expected given lead resides predominantly in the erythrocyte and not the serum portion of human blood. The low levels observed revealed an analytical flaw in using blood serum. At the time of the study, Agbogbloshie was littered with thousands of cathode ray tube computer monitors, which contain abundant amounts of lead— several kilograms per monitor.16 The U.S. EPA's standard for lead in bare soil in children's play areas is 400 mg/kg (ppm) for ‘play areas' and 1200 mg/kg for non-play areas.17 Mean lead levels determined for 100 samples in 5 areas within the perimeter of e-waste handling grounds showed that 56% were above U.S. EPA standards. Of that 56%, the highest lead in soil content sample taken was 18,125 ppm, which is 15 times higher than the non-play areas standard. Results of the soil samples thus show that there is contamination at the site likely due to dismantling, burning and related activities. However, since the soil was only tested for lead, the extent of contamination from other heavy metals and trace elements is unknown.8 

This exploratory study was a first effort to ascertain the impact of e-waste handling on workers. The study was able to define and describe e-waste handlers at Agbogbloshie market in terms of their demographic characteristics, as well as their knowledge, perceptions and practices in relation to health risks associated with exposure to e-waste recycling.

The group of workers involved in e-waste handling comprised mostly young men from the northern region of Ghana, notably from the Tamale municipality, Tolon/Kumbungu and Savelugu/Nanton districts. The high prevalence of overweight individuals and fungal skin rashes is suggestive of underlying challenges of lifestyle health problems, including personal hygiene and nutrition.

The results revealed elevated urine levels of select heavy and trace metals as compared to a control population, as well as elevated levels of 7 elements in the exposed group's blood serum. However, a significant number among the non-exposed group also had elevations in some of the heavy/trace elements in their urine, which suggests there may be other environmental sources of exposure to these elements outside of e-waste processing that need to be investigated. Measurements from this study, however, require careful interpretation considering its exploratory nature. Also, the population sampled was essentially self-selected and may suffer from sampling bias.

The levels of select heavy metals and trace elements in ambient air and soil were also elevated at the scrap yard, which may explain the increased body burden of these elements among those operatives at the e-waste disposal sites involved in e-waste recycling activities.

On the basis of the above, it is clear a more in-depth assessment of environmental exposure and bio-markers would help further clarify the extent of public health risk among e-waste recycling workers. However, the current assessment does provide adequate data to initiate public health education interventions. This information and education on the increased risks to health and safety likely being posed by their methods of operation and practices should be provided to study participants, other e-waste operatives and the association of scrap dealers. Additionally, education of e-waste handlers should also extend to healthy lifestyle issues to reduce occurrences of fungal and other skin diseases, as well as limit problems of obesity and linked chronic diseases like diabetes and hypertension.

Changes should be planned and implemented with the participation of the e-waste handlers to ensure success. Should relocation of the recycling site become an option, this should be accompanied by a clean-up of the area in which operations are currently carried out in order to minimize health risks not only to operatives but also to persons working and shopping in the nearby food market. Finally, action on policies and legislation to minimize the influx of e-waste into the country should be explored.