Lead poisoning of children in Africa, II. Kwazulu/Natal, South Africa

ELSEVIER The Science of the Total Environment 197 (1997) I- 11

Lead poisoning of children in Africa, II. Kwazulu/Natal, South Africa

Jerome Nriagu a**, Champak C. Jinabhai b, Rajen Naidoo b, Anna Coutsoudis c

’ Department of Environmental and Industrial Health, School of Public Health, University of Michigan, Ann Arbor MI 48109. USA

’ Department of Community Health, Faculty of Medicine, University oj‘ Natal, Durban, South Africa

’ Department of Pediatrics, Faculty of Medicine, University of ,Vatal, Durban, South Africa

Received 16 August 1996; accepted 11 November 1996

Abstract

The prevalence of elevated blood lead (PbB) levels in rural and semi-urban areas of Kwazulu/Natal (KZN) as well as the risk factors for elevation of PbB among children in informal settlements are examined. The study involved over 1200 children in two age groups: 3-5 and S-10 years old. Average PbB level in Besters, an informal settlement in the Durban metropolitan region, was 10 pg/dl with 5% of the children showing PbB level of greater than 25 pgg/dl. By comparison, average PbB value in Vulamehlo, a rural area located 90-120 km from Durban, was 3.8 pg/dl and 2% of the PbB was greater than 10 ,~g/dl. The results show that many children in KZN are at risk of being lead poisoned. Household factors that were significantly associated with PbB levels in children included distance from tarred roads, overcrowding, hygienic habits in the household and the burning of solid wastes for heating or cooking. Lack of significant association with child risk behaviors is attributed to the over-riding influence of high levels of contami- nated dusts both indoor and outdoor. With recent introduction of unleaded gasoline in South Africa, a number of unique socio-ecological factors are expected to slow the change in,PbB levels of children in some townships. 0 1997 Elsevier Science B.V.

Kqwwds: Blood lead levels; Lead poisoning; Socioecological factors; South Africa

1. Introduction

Recent epidemiological studies show that the mental and behavioral development of children are adversely affected by exposure to very low

* Corresponding author.

levels of lead [I -31 and that lead pollution must be regarded as a threat to children’s health in every country. At present, information on preva- lence of childhood lead poisoning in South Africa (SA) is extremely limited, and the few studies, restricted almost exclusively to one (Cape) province, suggest that many children in the coun-

0048-9697/97/$17.00 0 1997 Elsevier Science B.V. All rights reserved PII SOO48-9697(96)05407-l

try are at risk of being poisoned by lead in their environment. For instance, a study of first-grade children in Woodstock, Cape Town found that about 700/o of the white children and 95% of the mixed-race children had PbB levels above 10 p g/dl, the current action level in the United States; about 13% of the mixed-race children had PbB levels greater than or equal to 25 @g/d1 [4]. Some in- dication that elevation of PbB is a country-wide problem comes from recent measurements showing ambient levels of lead in air that are high enough to constitute a hazard to the children in most urban areas. Average airborne lead concentrations in Durban range from 0.4 (residential) to 1.8 (indus- trial) ,ug/m3 compared to 0.03 pg/m3 in rural areas of the province [5]. Reported average concentra- tions of Pb in air in other cities of South Africa also range from 0.4 to 1.5 pug/m’ with the values in winter time often being greater than 1 .O @g/m’ (Albertyn, Personal Communication, 1995). The current aver- age concentrations in SA are similar to those of the urban areas of the United States in the 1970’s when childhood lead poisoning was known to be ram- pant [l].

This paper examines, for the first time, the prevalence of elevated blood lead levels in children in rural and urban communities in the province of Kwazulu/Natal (KZN). It discusses the household, child behavior and parental behavior risk factors for lead exposure and the unique socio-ecological features that predispose the children in an informal settlement to lead poisoning. Blood lead (PbB) levels obtained during the study provide a refer- ence database that can be used to follow changes in childhood PbB levels associated with the intro- duction of unleaded gasoline into South Africa early in 1996.

2. Study design

The study involved over 1200 children in age groups of 3-5 and 8- 10 years which were selected from communities with different levels of lead in their environments. One study group consisted of Grade I pupils from Vulamehlo, a rural area located about 90- 140 km from Durban. The other group was a random selection of children from

informal settlements in Inanda Township. In this study, we have used ‘Besters’ in a geographic sense to denote the settlement of Besters itself and the three adjacent and smaller settlements (namely Mzomusha, Nhlungwane and Ezimgwane) that participated in the study. It is done for convenience and no disrespect is intended to the residents of any of the settlements. The study protocols used in Besters and Vulamehlo were somewhat different and are described separately below.

2.1. Bes ters

Community support was an important element in the study. Consultations were made and permis- sion and authorization were obtained from the communities of Besters, Mzomusha, Nhlungwane and Ezimangwane, the Inanda Community Devel- opment Trust and the City of Durban Health Department. The protocol for the study was ap- proved by the Ethics Committee of the Faculty of Medicine, University of Natal, and sanctioned by Inanda Community Development Trust. Subse- quently, community-group facilitators were used to identify the housing units for the study, ensure that the children came to the clinic for blood sampling and assisted with collection of soil/dust samples around houses.

The following sub-communities participated in the study: Besters (Sections l-4), Mzomusha (Sec- tions A, C and D), Ezimangwani (Section 1) and Nhlungwane (Section 4). Survey maps showing the house numbers in each sub-community were ob- tained from the Inanda Community Development Trust. The house numbers were fed into a com- puter from which a randomized set of housing units were generated for the study. About 400 housing units were initially selected.

The household questionnaire was developed us- ing the previous health and welfare survey by Jinabhai et al. [6] as a guide. In addition to questions on household characteristics, specific questions to explore environmental risk factors for lead poisoning were included. The respondents were also asked to rank the environmental prob- lems in their communities. The questionnaire was tested in a pilot study involving 30 households, or about 7% the selected sample of housing units.

1. Nriagu et al. : The Sciencr of thr Total Environment 197 (1997) l-11 3

Experienced Zulu-speaking interviewers from the SA Medical Research Council (Durban office) were trained to administer the questionnaire. Each interviewer worked with a facilitator from the community who identified the housing units and served to reassure the residents. The inter- viewers were told to go to the unit with the next higher number where (a) the selected number could not be found, (b) there was nobody at home, or (c) the occupants were not willing to participate in the study. Less than 10% of the selected household units were replaced in the field for the reasons above. The interviewers averaged about eight households per day. The number of households that participated in the study was 384, and contained over 2000 individuals. The survey was completed during the period of 3-12 July 1995.

The head of the household or adult respondent was asked to sign a copy of the consent form (Zulu translation) and told when to bring the children 3-5 or 8- 10 years old to the community health clinic for further tests. At the clinic, the height, weight and knee-heel ratio of each child were measured. A physical examination was then given. About 2.5 ml of the venous blood was drawn, by a trained nurse, into a lead-free plastic vacuutainer containing dry potassium EDTA. The puncture point was cleansed with alcohol, soapy water and then with distilled water. The blood was drawn, carefully mixed with the anti-coagu- lant and then stored in a cooler box. All samples were frozen as soon as they got back to the laboratory. The blood samples were collected be- tween July 13 and 21, 1995. Of the 637 children (in the right age group) identified from the ques- tionnaire, 452 actually participated in the study. The high capture rate (71%) can be attributed to the enthusiastic support from the community. Blood samples could not be obtained from about 10% of the volunteers because the veins were too small or the child was too terrified to allow the sample to be drawn safely.

2.2. Vulamehlo

The study group was restricted to Standard 1 pupils (8- 10 years old) from the following ele-

mentary schools: Dududu, Hluthankungu, Vukaphi, KwaMaquza, Nomandla, Dumisa, Soul, St. Michaels, Tholimfundo and Zamafuthi. All the pupils in each class were asked to partici- pate and each child was given a ‘consent form’ which was signed by the parent or guardian and returned to the study team. Over 800 children were thus recruited.

Blood samples were collected in June 1995. For the students who were not available in late June (which was close to end of the term), a second visit was made in early August to collect the blood; the second set of samples constituted less than 10% of the sample population. The puncture point was properly cleaned as described above. About 2.5 ml of venous blood was drawn into lead-free plastic vacuutainers that had dry EDTA. The blood samples were stored in an ice box and when returned to the laboratory were quickly frozen.

2.3. Laborutory methods

A 1 ml blood sample was digested with 10 ml of trace metal-grade nitric acid in a sealed Teflon bomb of a microwave digestion system. The diges- tion resulted in clear solution which was diluted to 25 ml with Mill-Q water (Millipore Corp., Bedford, Masachussets). The digestion was done in batches of 12 samples and each batch included a reference blood samples (NIST 955a). The di- gestion of samples in any batch was repeated if the lead content of the reference blood sample deviated from the certified value by + 10%. Anal- ysis of blood samples was performed in the School of Public Health, University of Michigan and followed the ultra-clean laboratory procedure developed for water samples with very low metal levels [7,8]. In particular, all the labware that came into contact were decontaminated by the nine-step process described by Nriagu et al. [7].

Lead concentration in each sample solution was measured using a graphite furnace atomic absorp- tion spectrometer (GFAAS) equipped with a Zee- man background corrector. Replicate analysis of several samples show the range of error to be + 10% for all blood lead data reported.

4 .J. Nriagu et al. The Science elf the Total Encironnwnt 197 (1997) 1~ 1 I

3. Results

Informal settlements are everywhere in Kwazulu/Natal (KZN) and generally consist of dense shantytowns of self-constructed shelters un- der different conditions of informal or traditional tenure. The settlements contain over a quarter of the population of KZN province and are different in terms of shelter type, degree of official recogni- tion and location [9]. Congregation of large num- bers of people into informal habitation in urban areas entails several environmental health risks, childhood lead poisoning being just one example. The populations of Besters proper, Ezimanwane, Nhlugwane or Mzomusha are estimated to be 50000, 20 000, 8000 and 7800 respectively [9]. These settlements are functionally linked to Dur- ban through employment, migration, survival net- works, wage incomes and, most recently, administratively. In view of the limited informa- tion available on these settlements, it became nec- essary to obtain some basic demographic data for the communities studied.

Key data from the household survey are sum- marized in Table 1. The majority (77%) of the housing units in Besters were built of daub and wattle, 8% of discarded metal/wooden containers and only 6% of brick [6]. Most (60%) of the houses have two to three rooms and 47% of the housing units are located less than 100 m from a motorable road. The houses are poorly ventilated (70% of the rooms have one or no window), and 57% of respondents reported roof leaks. The ma- jority (92%) owned their housing units and 80% of the families had lived in Besters for over 3 years [6]. The duration of residence and large percentage of close family relations in the house- holds indicate a stable community.

The number of persons per household varied from one to 20 and averaged 5.4. About 51% of the heads of households attained an educational level of Standard 5 to 10 (equivalent to American Grade 7 or higher), and 54% of the households have an income in the range of I- 1000 Rand (US$l .OO 2 4.5 Rand late 1996 rate) derived pri- marily from formal (39%) and self employment (16%). Most of the households get their water supply from communal taps using plastic contain-

Table 1 General demographics of the study population in Besters

Characteristic feature Frequency Percentage

Housing units surveyed Besters Ezimangweni Nhlungwane Mzomusha

Age and sex of children in sur- vey 3-5: male 3 -5: female 8-10: male 8- 10: female

Education of household head <Std 2 Std 3-Std 4 Std 55Std 10 No schooling Unknown

Household income/month < R500 R501-RlOOO Over RIO01 Unknown

Source of income Formal employment Informal activity Self employment Unemployment Other

Distance of house from high- way or tarred road cl00 m 100&500 m r500 m

Number of rooms in house 1 2 3+

House sweepings/week l-2 times 3-4 times 15 times

Child playing outside <h/day 3-6 h/day >I h/day

215 56 55 14 42 11 72 19

183 29 181 28 146 23 127 20

40 10 54 15

185 48 25 6.5 80 1.5

34 76 4.0

28

39 16 1.3

30 8.3

39 47

9.1

29 45 20

1.6 4.0

93

11 (6.6)” 38 (34)” 29 (31)

a For 335 years old; percentage for the S-10 years old are given in brackets.

J. Nriugu et al. / The Science of’ the Total Enoiroment 197 (1997) 1-I 1

160 -

140 -I

120 i

100 i

60 -I

4 6

Fig. 1. Histogram of blood lead in children in Besters, a semi-urban area.

12 16 20

Lead Concentration (ug/dL)

24

ers in which the water is also stored until used. Litter, lack of sewer for wastewater and animal wastes are of particular concern to the residents. There are smokers in 30% of the households and about 49% of the children have respiratory prob- lems. The highest ranking environmental prob- lems from the residents’ perspective include overcrowding (97%), inadequate recreational fa- cilities for children (93%), wastewater (85%) and litter (83%). Enviromnental issues such as food poisoning (9%), acid rain (16%), noise (24%) and air pollution (47%) received very low ranking and may be symptomatic of what the people know about lead poisoning. It is evident that education of the community must be an important adjunct to any strategy for reducing the risk of childhood exposure to lead in their environment.

A histogram of blood lead (PbB) levels in pedi- atric population of Besters has a skewed distribu- tion (Fig. 1). About 50% of the children had PbB > 10 pgg/dl, a level for concern recommended

by the US Centers for Disease Control [lo]. About 5% of the children had PbB > 25 pg/dl. Average PbB value for all the children was 10 pg/dl. The blood lead values for Besters may be compared with the following recent (1991) aver- ages in various parts of Cape Town: 16 pggidl (n = 243) in Woodstock, 16 pg/dl (n = 48) in Schotcheskloof, 14 pg/dl (n = 115) in Hout Bay, and 15 pg/dl (n = 104) in Mitchell’s Plain. In all areas of Cape Town, over 90% of the children were reported to have PbB > 10 pggidl [l I]. Higher prevalence of elevated PbB levels in Cape Town is consistent with observed higher concentrations of atmospheric lead concentrations in that city com- pared to Durban [6].

The distribution of PbB in Vulamehlo children (Fig. 2) is surprisingly similar to the pattern for US children, l-5 years old, during the 1988-- 1991 period [12]. The PbB levels (Fig. 2) point to the fact that some children in rural areas of KZN are also at some risk of getting lead poisoning. Mean

J. Nriagu et al. 1 The Science of the Total Envirormenf 197 (1997) I I I

300

250

I I 2

T

Pb Concentration (ug/dL) Fig. 2. Blood lead levels in children of Vulamehlo, a rural area

PbB value for US children was 3.6 pg/dl com- pared to 3.8 pg/dl (n = 660) in Vulamhlo. About 8.9% (5.5% among white children) of the US children had PbB > 10 pg/dl compared to 2% in Vulamehlo. The similarity is not surprising since the ambient air concentrations of lead in Vu- lamehlo (0.03 pg/m’) are similar to current levels in many rural and even urban areas of the United States [13]. In their study of children from Wup- perthal, a rural area in northwestern Cape Province, Grobler et al. [14] found average PbB level of 3.4 pg/dl which was claimed to be “repre- sentative of background whole blood lead level of South Africans”. This myth was perpetrated by a number of studies that reported high PbB levels (often > 10 pgg/dl) in other rural areas of the Cape Province [l 11. We found many children in Vu- lamehlo with PbB in the range of 0.2-1.0 pggidl, well below the so-called ‘background’ value. Fur- thermore, the PbB concentrations in rural KZN

are many times higher than the 0.02 pg/dl esti- mated to be the pretechnological background value for the general population [15].

There are major differences in PbB levels in various sub-communities of Besters (Table 2). For all ages and both sexes, highest average concen- tration (17.4 pgg/dl) was found in Ezimangwane. The highest concentration observed (108 pg/dl) was in this neighborhood and over 10% of the children there had PbB > 30 pg/dl. The lowest concentrations (range, 2.6 to 15.6 pg/dl; average, 6.1 bgg/dl) were found in Nhlungwane. The neigh- borhood disparities can be attributed to differ- ences in miles of tarred roads, hence vehicular volume, proximity to the major expressway and to the volume and nature of cottage industries.

In general, PbB levels for children 3-5 years old in Besters are higher than those for 8- IO-year olds, both for males and females (Table 2). The difference between the two age groups may be

J. Nriagu et al. / The Sciewe oj’ the Total Environment 197 (1997) 1 -1 I

Table 2 Average blood lead levels (pg/dl) in children from different communities inKwazulu/Natal

Location” 3-5 years old 8 - 10 years old Both ages

Mzomusha” Male Female Both sexes

Ezimangwane” Male Female Both sexes

Nhlungwane” Male Female Both sexes

Besters” Male Female Both sexes

Dududu’. both sexes Tholimfundo’. both sexes Nomandla’, both sexes Zamafuthi’. both sexes Hluthankungu & Soul”, both sexes St. Michaels & KwaMaquza’, both sexes

11.6

35.P 18.6 24.8

7.1 6.9 7.0

7.9 8.9 8.7

‘I Semi-urban communities in Durban Metropolitan area. b This value is biased by a few outliers. c Schools in rural areas of Vulamehio.

related to the fact that younger children are apt to play dirty in contaminated local environments. For the S-lo-year olds, the PbB levels of males are commonly higher than those of females, but there is little difference in PbB between the sexes in the 3%5-year age group. It is not clear why the younger children in Ezimangwane seem to be particularly predisposed to excessive exposed to lead.

Average PbB levels in school age children (8- 10 years old) in rural areas of KZN are well below those in Besters (Table 2). The diff- erence can be attributed to higher environmental lead levels in Besters. Average atmospheric lead level in Besters was 0.26 pug/m3 compared to 0.03 /lg/rn’ in Vulamehlo [5]. Average concentra- tion of lead in soils was also higher in Besters (58 pg/g) compared to Vulamehlo (14 pg/g) (Nriagu, Unpublished Results, 1996). To a large extent, automotive tail pipe accounts for most of

9.2 10.3 9.5 9.5 9.3 10.1

II.6 21.3 6.6 12.3 8.6 17.4

S.8 6.4 4.3 5.8 5.3 6.1

8.4 8.1 6.6 8.1 7.3 8.1 3.6 3.0 3.1 2.3 3.1 4.2

the difference in ambient lead levels of urban versus rural areas and most likely for the differ- ence in PbB levels between Besters and Vu- lamehlo. Our results contradict the claim by the South African Department of National Health and Population Development [16] that the con- tribution to the body of lead from gasoline is extremely small.

The questionnaire (used in Besters only) in- cluded several items on child and parental risk behaviors as well as household risk factors that can result in undue exposure to lead. Bivariate analysis was applied to the variables using the statistical methods in Epi Info software. Since the distribution of PbB is skewed (Fig. l), analy- sis of variance was performed by Mann-Whit- ney or Kruskal-Wallis test for two groups with significance set at the 90% confidence interval for this study.

Because of the unique socio-ecological features

of informal settlements, traditional risks factors identffied in many studies may not be valid in Besters. Since most of the households get their water from communal taps, lead in the distribu- tion system is of little concern. No significant association was found between PbB and the type of containers (plastic, metal or earthenware) used to collect or store the water. The daub and wattle structures were rarely painted so that lead paint was not a factor of concern. Among the household risk factors, there was a significant association between PbB and distance of hous- ing unit from tarred roads (P = 0.0673; x2 = 5.4) number of rooms in the house (P = 0.0571; x2 = 12) number of rooms used for sleeping (P = 0.0857; x2 = 8.2) and roof leaking during rainfall (P = 0.0627; 1’ = 3.4). No correlation was found with the number of times the rooms were swept or the presence of pests in the house. Strong association was found with the type of fuel used in cooking (P = < 0.0001; x’ = 17) or heating (P = 0.0024; ,yz = 9.3) (wood, gas, paraffin or electricity) and lighting (candles, kerosine or electricity; P = 0.0022; 1' = 9.4)). Whether the child was present while the food was being cooked or slept in the same room where the food was prepared did not appear to affect the PbB values. The number of smokers in the house was also significantly associated with PbB (P= 0.0003; x2 = 21).

There was no significant association between children’s PbB and the educational attainment or income of the head of the household. This should not be surprising; other risk factors should over-ride the minor dffterences in socio- economic conditions in the settlement. In fact, the community was selected for study so as to minimize the socio-economic determinants of childhood lead exposure. Blood lead levels were not significantly associated with a number of child risk behaviors, such where they played (around the house, on the street or the park), number of hours played outside per day, sucking of thumbs or frequency of hand washings per day. The significance of the associations, or lack thereof, are discussed below.

4. Discussion

This study demonstrates that many children in semi-urban areas of KZN have PbB above 10 jlg/dl. A small number of children in rural areas also have PbB above this level of concern. The habitation pattern for Durban and other cities of South Africa (SA) is rather unique in that the affluent whites live in inner cities and the econom- ically deprived blacks are confined to native (semi-urban) townships. This has implications in terms of the exposure of different racial groups to lead. A recent study found average airborne lead levels in Durban to be 1.8. 0.86, 0.56 and 0.56 pg/m3 in industrial, commercial, park/beach and residential areas respectively, compared to 0.26 lug/m3 in Besters, a semi-urban community [5] These elevated atmospheric levels point to the fact that large numbers of children in the inner city of Durban, regardless of the race, are at risk of being lead poisoned. On top of that, houses in older parts of the city may contain lead pipe and lead paint. The prevalence of chronic lead poisoning observed in Besters is believed to be symptomatic of a widespread public health problem throughout the Durban area. A cross-sectional study to ascer- tain the cultural differences in susceptibility to lead poisoning in Durban is urgently needed.

Lead pipe (water contamination) and lead paint are not important risk factors in Besters so that the major sources of lead exposure are believed to be food, air/dust/soil and traditional medicines. Dietary exposure to lead in SA is basically un- known. Since few households in Besters or Vu- lamehlo possess refrigerators, the people presumably depend mostly on fresh produce or canned foods. Although a number of food brands use lead solder (from a brief inspection of shelves of a local food store), the practice is far from being common. It is conceivable that some of the children in the high end of the PbB spectrum have derived their excess lead from habitual consump- tion of cheap cans of food with lead solder.

Traditional medicines are still widely used among the Zulu population of KZN. Our survey, however, found that local clinics and hospitals were preferred by 82”/;, of parents and care givers in Besters. Other potential sources of lead expo-

sure in local townships include local pottery glazed with lead and rainwater collected from contaminated roof tops. Regardless of the exact pathways, our study shows that the combined routes result in lead exposures that are excessive, likely to produce disease and need to be reduced. The disparity in average PbB values in Besters and Vulamehlo helps to put the relative contribu- tions from various sources in perspective. With few roads, no electricity, no pipe-borne water, no hospitals, few clinics and an economy based on subsistence farming, children in Vulamehlo are more likely to be exposed to lead from traditional local foods, medicines and other causes. Average PbB in these children have become elevated (3.8 Llg/dl) but are small compared to those in Besters ( 10 /dg/dl). The difference (6 pg/dl) can be at- tributed to lead pollution in the air-dust/soil con- tinuum of the urban enviromnent. It is not surprising that there is a significant association of PbB with distance of housing unit from a mo- torable road in Besters.

The absence of significant relationships between PbB and child risk behaviors in Besters is con- trary to what is typically found in other countries [ 17 ~- 191 or even in formal settlements in the coun- try [20]. The heavy dustiness of the settlement would result in extensive contamination of both indoor and outdoor environments, and this could have obscured the effect due to how long or where the child plays outside. Correlation with heating and cooling is probably related to the fact that many types of lead-containing material (printed paper, cardboard boxes, painted woods and other solid wastes) are often used for these purposes. Sweeping of the floor with a broom would only serve to resuspend contaminated dusts in the air and should offer little protection to the child. Association of PbB with the number of rooms in the house, the size of the household or roof leak indirectly reflects deficiencies in hygienic habits of individual households.

The importance of cottage industries on child- hood PbB levels in local townships should be noted. A substantial fraction (25%) of the heads of households had informal jobs or were self employed. Common service/repair activities in homes in the settlements involved appliances, bat-

teries, electronics, welding and paint finishing. Lead solder and lead compounds used in such activities can result in extensive contamination of the home enviromnent. It is believed that expo- sure to lead from cottage industries may be re- sponsible for some of the high-end PbB values in both Besters and Vulamehlo.

Although lead poisoning (in its acute form with defined clinical syptoms) is a notifiable condition in South Africa, few cases have been reported during the past 5 years. A number of studies, however, have reported high prevalence of chronic lead poisoning in pediatric population of the Cape province [4,11,20-221. Our study shows, for the first time, that large numbers of children are also at risk of being lead poisoned in KZN province especially in urban areas. There is no information (published) on PbB levels in other provinces. Chronic plumbism in children should, in fact, be expected in most cities of SA in view of recent studies showing elevated levels of airborne lead in the urban areas, see [5]. Public awareness of the health hazards of lead poisoning remains very limited in the country and little attempt is being made to identify and control the sources of lead exposure. Continuing to disregard a serious public health problem by the medical establish- ment in the new SA is a disservice to the children whose mental development may be impaired.

The primary potential source of airborne lead in most urban areas is the automobile tail pipe. A risk reduction strategy that has been adopted by most developed countries involves removing lead from the gasoline, and reductions in lead content of gasoline in many countries have been accompa- nied by a sharp decline in PbB levels in every segment of their population [ 11.33 ~-351. Lead con- tent of gasoline sold in SA was reduced from 0.84 g/l in 1983 to 0.4 g/l in 1989 and early in 1996 unleaded gasoline was introduced [11.26]. It would, however, be a serious error to presume that the problem of childhood lead poisoning in SA has been solved with the introduction un- leaded gasoline. In the first instance, only a small fraction (30%) of atmospheric lead emission in the country comes from automobiles [26]. More im- portantly, the unique socio-ecological features of South African communities suggest that the tra-

I 0 J. jVriagu et al. The Science of the Total Enrironnzent 197 (1997) I 1 I

jectory of change in PbB will likely be different from anything that has been reported in Europe or North America. For instance, a study of chil- dren, 4-6 years old, in Cape Town found no difference whatsoever in average PbB levels be- tween 1982 and 1991 [2 11, and that was after the reduction of lead content of gasoline by half. Reference has already been made to the close similarity in PbB of rural children in Cape Province in 1982 [ 141 and Vulamehlo in 1995. Maresky and Grobler [27] found a small reduc- tion in PbB of Cape Town residents, from 9.7 pggidl in 1984 to 7.2 pg/dl in 1990. Table 3 shows the PbB content of blood samples ob- tained from children in Umlazi in 1985. Umlazi is a formal black settlement in southern part of Durban (both Umlazi and Besters are now part of the new Durban Metro system) and the study involved 637 children from birth to the age of 13 years. The remarkable thing is that the aver- age PbB in Umlazi (12 pgg/dl) and prevalence of values above 10 pggidl (65%) are only slightly higher than those of Besters even though the samples were taken 10 years apart. The studies cited above tend to suggest that the susceptibil- ity to lead in the environment may be different and that the rate of change of PbB in SA chil- dren following the introduction of unleaded gasoline will likely be much slower than re- ported rates in developed countries. Socio-eco- logical factors of native townships likely to

Table 3 Distribution of blood lead levels (pg/dl) in Umlazi children”

Ages (years) Blood lead levels (kgg/dl)

O-10 11-19 >20 Totalb

o-3 52 106 13 171 (27) 4-7 33 131 17 181 (28) 8-11 81 145 4 230 (36) >11 25 29 1 55 (9) Totalb 191 (30) 411 (65) 35 (5) 637

il These samples were analyzed by the National Center for Occupational Health (NCOH) in Johannesburg by graphite furnace atomic absorption method. b Percentage of each age or blood lead category is given brackets.

affect the exposure and rate of PbB change in- clude excessive dustiness. narrow streets and overcrowding, vibrant outdoor lifestyle, preva- lence of contaminated house dusts, consumption of contaminated rain water, malnutrition, poor hygienic conditions, cottage industries, use of lead-containing medicines and cosmetics and pan- demism of communicable diseases [28].

The South African Constitution has enshrined the right to an environment that is not harmful to health and rights to equity, justice and social development. In keeping with this fundamental right, the Department of Environmental Affairs and Tourism has adopted a policy of sustainable resource management, as part of a national envi- ronmental policy, which involves investigating the health implications of domestic fuel use, ve- hicle emissions of leaded gasoline with a view to developing an integrated pollution control strat- egy [29]. This study re-inforces the need and ur- gency for such investigations and provides a timely database that can be used to follow the decline in PbB levels of children in KZN in re- sponse to the removal of lead in gasoline sold in the province.

Acknowledgements

This study was partially funded by a grant from Fogarty International Training program awarded to the Center for Human Growth and Development, University of Michigan. We also thank the Joint Programme for Health and So- cial Development of the Human Sciences Re- search Council of South Africa for their support especially with the field work. Appreciation is extended to the following persons who helped with the field work and/or data processing: Dr A. Hoque, Dr A. Heeran, F. Rangongo, M. Cele, M.C. Shezi, F. Sheni, N. Mkhize, M. Tay- lor, V. Govender, C. Govender, N. Morar, P. Pillay, N. Thusi, J. Mshentshela, 0. Qula, N. Cele, Dr. T. Sebitloane and A. Mburu. The fol- lowing students worked on the project as part of their Fogarty training: R. Yearby, L. Chism and J. Gater. Analysis of samples in the laboratory was done by T.S. Lin and X. Wang.

.I. Nriagu et al. / The Science qf’ the Total Enoironment 197 (1997) I- 1 I 11

References

[‘I

PI

[31

[41

[51

[61

US Environmental Protection Agency, Air Quality Crite- ria for Lead, Report No. EPA-600/8-8-33/028aF. Re-

search Triangle Park, NC, Environmental Quality Office. 1986.

Needleman, H.L. The current status of childhood lead

toxicity. Neurotoxicity. 14 (1993) 161-166. In: C.P. Hawson, M. Hernandez-Avila and D.P. Raul

(Eds.), Lead in the Americas. Washington, DC, US Na-

tional Academy of Sciences Press, Institute of Medicine, Board of International Health. 1996.

van Schirnding, Y.E.R., D. Bradshaw, R. Fuggle and M. Stokol. Blood lead levels in South African inner-city

children. Environ. Health Perspect., 94 (1991) 125-130. Nriagu, J.O., C.C. Jinabhai, R. Naidoo and A. Cout-

soudis. Atmospheric lead pollution in KwazuluiNatal.

South Africa. Sci. Total Environ., 191 (1996) 69976. Jinabhai. C.C., L.D. Mametja and D.N. Pillay. The

Besters Health and Welfare Study. Durban, South Africa, Univesity of Natal. Center for Health and Social Studies,

1993. Nriagu, J.O., G. Lawson. H.K.T. Wong and J.M. Azcue.

A protocol for minimizing sample contamination in the

analysis of trace metals in the Great Lakes. J. Great Lakes Res., 19 (1993) 175-182.

Nriagu. J.O.. G. Lawson. H.K.T. Wong and V. Cheam. Dissolved trace metals in Lakes Superior, Erie and On-

tario. Environ. Sci. Technol., 30 (1995) 178-187.

Hindson, D. and J. McCarthy. Here to Stay: Informal Settlements in Kwazulu-Natal. Indicator Press, Dal-

bridge, Natal, South Africa, 1994. Centers for Disease Control. Preventing Lead Poisoning

in Children: A statement by the Centers for Disease Control. Atlanta, GA. US Dept. of Health and Human

Services. Public Health Service, 1991.

von Schirnding, Y.E.R. M.A. Kibel. R. Fuggle and A. Mathee. An overview of childhood lead exposure in

South Africa. City of Johannesburg, South Africa De-

partment of Enviromnental Health, 1995. [I21 Pirkle, J.L., D.J. Brody, E.W. Gunter, R.A. Kramer.

D.C. Paschal, K.M. Flegal and T.D. Matte. The decline in blood lead levels in the United States. JAMA, 272

(1994) 2844291. [13] US Environmental Protection Agency. National Air

Quality and Emission Trends Report. Report No. EPA-

450-R-92-001. Research Triangle Park, NC. Office of Air Quality Planning and Standards, 1992.

[14] Grobler. S.R.. R.J. Rossouw and L.S. Maresky. Blood

lead levels in a remote, unpolluted rural area in South

Africa. S. Afr. Med. J.. 68 (1985) 3233326.

t151

U61

[I71

[I81

v91

WI

PII

WI

~31

v41

~251

Flegal, A.R. and D.R. Smith. Lead levels in preindustrial humans. New Engl. J Med., 326 (1992) 129331294.

Department of National Health and Population Develop- ment. Effecr of lead in petrol on blood lead levels in children: Press release, Pretoria. South Africa, 1992. Gottheb, K. and J.R. Koehler. Blood lead levels in chil-

dren from lower socioeconomic communities in Denver. Colorado. Arch. Environ. Health, 49 (1994) 260- 265.

Norman. E.R. W.C. Bordley, I. Hertz-Picciotto and D.A. Newton DA. Rural-urban blood lead differences in North

Carolina children. Arch. Environ. Health, 94 (1994) 59- 63.

Olaiz, G.. T.L. Fortoul, R. Rojas, M. Doyer. E. Palazuelos and C.R. Tapia. Risk factors for high levels of lead in blood of school children in Mexico City. Arch.

Environ. Health, 51 (1996) 122-125. von Schirnding, Y.E.R. Enviromnental lead exposure

among inner-city Cape Town children: A study of associ- ated risk factors. PhD Thesis. University of Cape Town, South Africa, 1988.

von Schirnding. Y.E.R. A. Matte, P. Robertson. N. Strauss and N.U. Kibel. A study of the distribution of

blood lead levels in school children in selected suburbs in the Cape Peninsula. Department of Enviromnental Health, City of Johannesburg, South Africa, 19Y5.

Deveaux, P., M.A. Kibel. W.S. Dempster, F. Pocock and K. Formenti. Blood lead levels in pre-school children in

Cape Town. S. Afr. Med. J., 69 (1986) 421-424. Brody. D.J.. J.L. Pirkle, R.A. Kramer. K.M. Flegal. T.D.

Mate, E.W. Gunter and D.C. Paschal. Blood lead levels in the US population, JAMA. 272 (1994) 277- 283. Stromberg. U., A. Schutz and S. Skerfving. Substantial

decrease of blood lead in Swedish children, 1978894, associated with petrol lead. Occup. Environ. Med., 52

(1995) 764-769. Schuhmacher, M., M. Belles, A. Rico. J.L. Domingo and J. Corbella. Impact of reduction of lead in gasoline on the

blood and hair lead levels in the population of Tarrdgona Province, Spain, 1990&1995. Sci. Total Environ.. 184

(1996) 203-209. [26] Nriagu. J.O., M.L. Blankson and K. Ocran. Childhood

lead poisoning in Africa: A growing public health prob- lem. Sci. Total Environ., 181 (1996) 93- 100.

[27] Maresky. L.S. and S.R. Grobler. Effect of the reduction

of petrol lead on the blood lead levels of South Africans. Sci. Total Environ., 136 (1993) 43348.

[28] Nriagu, J.O. Toxic metal pollution in Africa. Sci. Total

Environ.. 121 (1992) I-37. [29] Dept. of Enviromnentd1 Affairs and Tourism. Consulta-

tive National Environmental Policy Process: Towards a New Environmental Policy in South Africa. Government Printer. Pretoria. SA. 1995.