TRANSACTIONS OF THE ROYAL SOCIETY OF TROPICAL. MEDICINE AND HYGIENE (1991) 85, 101-103

Serological markers of hepatitis B virus and certain other viruses in the population of eastern Caprivi, Namibia

101

J. J. Joubert’, C. A. van der Mewe*, J. H. Lourens3, G. Lecatsas4 and C. Siegriihn’ ‘Department of Medical Microbiology, University of Stellenbosch, South Af?ica; ‘Department of Quantitative Management, University of South AfLica, Pretoria, South Africa; 3Department of Virology, Faculty of Medicine, University of Pretmia, South Afica; 4Department of Microbiology, Medical University of South Africa (Medunsa), Pretoria, South Aftica

Abstract Serum samples from 475 male and 420 female

subjects from the population of eastern Caprivi, the north-easternmost territory in northern Namibia, were examined for the presence of hepatitis B surface antigen (HBsAg), antibodies to HBsAg (anti-HBs) and hepatitis B core antigen (anti-HBc). There appeared to be a rapid acquisition of antibodies to the different markers in childhood, reaching maximum values in the 20-24 year age group. The prevalence rate for all markers was lower in eastern Caprivi than in Kavang?, an area 180 km to the west. The reason for these hfferences is unclear. It is postulated that the hut tampan tick (Omithodoros moubatu), which has previously been shown to be implicated in the transmission of hepatitis B virus in Kavango, is more common in the latter territory than in eastern Caprivi. Huts in eastern Caprivi are usually better constructed than in Kavango, where reeds are commonly used as building material. Tampans lodge between the reeds and are more difficult to eradicate. West Nile virus was the most common of the arboviruses found in our sero-prevalence studies.

Introduction Surveys for the prevalence of hepatitis B virus

(HBV) markers in the northern territories of Narni- bia, excluding Caprivi, have been carried out by BERSOHN et al. (1974), BOTHA et al. (1984), and JOUBERT et al. (1985a). As far as can be ascertained, no survey for the prevalence of HBV markers has been conducted in eastern Caprivi-an area of almost 1 188 000 ha in extent between Zambia and Bots- wana. According to the 1981 census, the population of the territory was approximately 38 000, and consisted mainly of the Masubia and the Mafwe tribes (VAN DER VEGTE et al.. 19831. The Masubia tribe inhabit mainly the eastern pari. The Mafwe, a more heter- ogeneous group, are scattered over the central and western parts. The 2 groups do not seem to differ significantly in their customs and social structure (MURDOCK, 1959).

The average annual rainfall at Katima Mulilo is 706 mm and-most of the precipitation occurs between December and March. October is the hottest month with average daytime temperatures of 32*8”C (VAN DER VEGTE et al., 1983). The land is verv flat and, in seasons of heav$ rainfall, the eastern quarter’ of roughly 3480 km , situated between the Zambesi and the Chobe rivers, is flooded during the summer

Correspondence to Professor J. J. Joubert, Department of Medical Microbiology, Tygerberg Hospital, P.O. Box 1063, Tygerberg 7505, Republic of South Africa.

months and is habitable only on the higher areas. Consequently the inhabitants of this area are compel- led to migrate seasonally to ensure grazing for their cattle and to tend cultivated fields.

This study formed part of a general survey of disease profiles in the northern territories of Namibia.

Subjects and Methods Blood samples were obtained during April 1984

from patients and staff at Katima Mulilo hospital, black military staff at Mpacha, as well as patients and schoolchildren from 14 different clinics and schools throughout the territory. Consent to blood being taken was obtained from all adults; written consent was obtained from either parents or guardians df the children. The area sampled included all ethnological groups present in the region. Military staff and schoolchildren all came from the area under study and were representative of the population as a whole. Blood, drawn from the antecubital fossa, was centri- fuged within 24 h of collection. Serum samples were kept at 4°C until use. All samples were tested for hepatitis B antigen (HBsAg), antibodies to HBsAg (anti-HBs), and antibodies to hepatitis B core antigen (anti-HBc) by solid phase radio-immunoassay (Abbott, Chicago, USA). HBsAg-positive samples were also tested by radio-immunoassay for hepatitis B e-antigen (HBeAg). Antibodies to the following arbo- viruses were determined by the haemagglutination inhibition (HI) test: Sindbis (SIND), Chikungunya (CHIK), Wesselsbron (WSL), West Nile (WN) and Rift Valley fever (RVF). Antibodies to Marburg (MBG), Ebola (EBO), Lassa fever group (LAS), Hantaan group (HTN) and Crimean Congo (CON) haemorrhagic fever viruses were determined by in- direct immunofluorescence (IF) tests. Confidence intervals for the prevalence rates of the different markers were calculated according to the methods described by NETER et al. (1979).

Results and Discussion The results of the survey of hepatitis B markers are

shown in Table 1 and the prevalences of certain other viruses are shown in Table 2.

Table 1 shows that there was a rapid acquisition of antibodies to the different hepatitis B markers in childhood, reaching maximum values in the 20-24 year age group. The drop in antibody prevalence in older age groups may perhaps indicate a decline in (detectable) antibody levels in later life. A total of 41 persons of all ages (4.6% of the samples) were HBeAg-positive. This figure was comparable to that obtained in Kavango (JOUBERT et al., 1985a) and

102

Table 1. Frequency of HBV markers in 895 subjects in eastern Caprivi, Namibia

Marker Q-9 1614 Age group (years)

15-19 2&24 25-75 Total

95% confidence interval for population

Female subjects Anti-HBc positive only

Anti-HBs positive only

Anti-HBs and anti-HBc positive

HBs & anti- HBc positive

Negative for all markers

Total

Male subjects Anti-HBc positive only

Anti-HBs positive only

Anti-HBs and anti-HBc pos

HBs & anti- HBc positive

Negative for all markers

lY105 (14.29%)

13005 (12.38%)

28/105 (26.67%)

6/105 (5.71%)

431105 (40.95%)

105

91103 (8.74%)

191103 (18.45%)

231103 (22.33%)

131103 (12.62%)

391103 (37.86%)

103

111118 10197 (9.32%) (10.31%)

191118 13197 (16.10%) (13.40%)

45/l 18 49197 (38.14%) (50.52%)

14/l 18 13197 (11.86%) (13.40%)

291118 12197 (2458%) (12.37%)

118 97

20/l 14 (17.54%)

151114 (13.16%)

451114 (39.47%)

161114 (14.04%)

181114 (15.79%)

114

15173 (20.55%)

5173 (6.85%)

34173 (46.58%)

13173 (17.81%)

6173 (8.22%)

73

l/28 (3.57%)

5/28 (17.86%)

13128 (46.43%)

4128 (14.29%)

5/28 (17.86%)

28

4184 (4.76%)

3184 (3.57%)

49184 (58.33%)

26/84 (30.95%)

2/84 (2.38%)

84

9172 (12.50%)

13/72 (18.06%)

30/72 (41.67%)

8172 (11.11%)

12172 (16.67%)

72

lO/lOl (9.90%)

81101 (7.92%)

541101 (53.47%)

221101 (21.78%)

7/101 (6.93%)

101

(lO&)

(15&/O)

165 (39.29%)

(lo.::%)

101 (24.05%)

420 (100%)

(12G0)

(lo&o)

205 (43.16%)

(18&~)

(15.:&)

475 (100%)

(796-13.94)

(13.2616.74)

(34.6143.40)

(7.75-13.68)

(19.96-28.14)

(9.26-15.16)

(7.7613.29)

(38.70-47.62)

(15.42-22.48)

(11.93-18.39)

Tabke 2. Frequency of occurrence of antibodies to certain viruses in eastern Caprivi, Namibia

Virus Soldiers Civilians

(111 subjects)(SlO subjects) Total HA1 testsa West Nile 65’ 114’ 179 Wesselsbron

IF tertsb

AC 26 14’ :;

8 13 1 1

11 7’ :s

Lassa A

4 5 Congo Crimean Marburg 0 : :

No. positive for one or more of the above antibodies: 248 No. negative for all the above antibodies: 373 “Haemagglutination inhibition. %direct immunofluorescence. ‘Statistically significant difference between soldiers and civilians.

included 22 persons (18 males and 4 females) of 15 years and older.

The prevalence rates of most of the hepatitis markers obtained in this survey were lower than those obtained in Kavango (JOUBERT et al., 1985a). Conse- quently the prevalence of subjects negative for all markers (11*9%18*4% for males and 2&28.1% for females) &as considerably higher in eastern Caprivi than in Kavango, where a combined figure of 064.5% for males and females was obtained. The

reason for this difference was not clear; ecologically, the 2 regions were very similar so that it could possibly be ascribed to differences in cultural prac- tices. We have previously presented evidence that the hut tampan tick (Ornithodoros moubata) may play a role in the transmission of hepatitis B virus in Kavango, an area to the west of Caprivi (JOUBERT et al., 1985b; JUPP et al., 1987). These ticks live in the sand of hut floors and occurred commonly in Kavango and ivcapyivi. It seems reasonable to assume that the tampan could also be partly responsible for the transmission of the virus in eastern Caprivi.

Relapsing fever, caused by Borrelia dutconi, trans- mitted bv 0. moubata. was nrevalent in eastern Caprivi ahd in Kavangb. The ‘2 areas are 180 km apari. No accurate figure of the incidence of relapsing fever in the 2 territories is available. According to Mr A. P. Els, Chief of Public Hygiene, Namibia Depart- ment of National Health (personal communication), the incidence of this disease was considerably lower in eastern Canrivi than in Kavango. Walls of huts in eastern Caprivi were constructid of either reeds or clay. The latter was preferred, but due to the very sandy nature of the soil, clay suitable for building was not always available and reeds were often used. In huts with clay walls and floors, tampans could be controlled more easily by regular sweeping. It was difficult to control tampans in sandy floors and between reeds in the walls. In general, clay huts were more numerous in eastern Caprivi than in Kavango. Furthermore, it appeared that changing sleeping quarters frequently for fear of witchcraft was preva- lent in Kavango, but not in eastern Caprivi. The

103

alternating use of sleeping huts by carriers and susceptible individuals would ensure optimal condi- tions for transfer of HBV by the tampan population in the hut.

Various other factors may be conducive to the spread of HBV in eastern Caprivi. Most inhabitants of the region showed scarification of the skin produced by traditional healers for treatment of a variety of disorders. The virus might be transferred by unsteril- ized instruments used repeatedly on several patients. Circumcision was not practised in eastern Caprivi or in Kavango. Furthermore? promiscuity was common, as indicated by the high mcidence of sexually trans- mitted diseases.

In the non-hepatitis B survey (Table 2), antibodies against West Nile virus were the most frequently encountered, at titres considerably higher than those to other viruses. This indicated a nast enidemic of West Nile virus in this area which, according to Professor R. Swanewel of the National Institute for Virology, Sandringham, Johannesburg, South Africa (personal communication), probably formed part of a much more widespread epidemic which occurred during 1984 in the western Transvaal and Orange Free State in South Africa. A 13% prevalence rate of antibody to West Nile virus, determined at the National Institute for Virology in sera collected in Eastern Caprivi during 1983, had increased to 33.7% in the present study.

The antibody titres to Wesselsbron virus probably represented cross-reactivity with West Nile virus antibodies, as these 2 flaviviruses are known to be serologically closely related. No increase in prevalence to Sindbis virus antibodies was noted since 1983 and there was only a slight increase in the prevalence of antibodies to Rift Valley fever virus. Using the x2 test there was, however, a statistically significant higher orevalence of antibodies to Rift Vallev fever (P<O.Ol) and to West Nile (P<O*OOl) and Ebola viruses (P<O*Ol) in soldiers than in civilians (Table 2).

Acknowledgements We are indebted to Professor R. Swanepoel of the

National Institute for Virology, Sandringham, Johannesburg for doing the HAI and IF tests, Dr F. de Villiers, Pretoria, for help with the field work, General D. P. Knobel for arranging air transport to, and accommodation at, Katima Mulilo, and Dr A. Gunders for valued criticisms. This work was supported by a grant from the South African Medical Research Council to J.J.J.

References Bersohn, I., MacNab, G. M., Pyzikowska, J. & Kew, M. C.

(1974). The prevalence of hepatitis B’(Australia) antigen in southern Africa. South African Medical 3ourna1, 48, 941-944.

Botha, J. F., Dusheiko, G. M., Ritchie, M. J. J., Mouton, H. W. K. & Kew, M. C. (1984). Hepatitis B virus carrier state in black children in Ovamboland: role of perinatal and horizontal infection. Lancer, i, 1210-1212.

Joubert, J. J., Prozesky, 0. W., Lourens, J. G. H., Van Straten, A. M. S., Theron, J. W., Swanevelder, C., Meenehan, G. M. & van der Merwe, C. A. (1985a). Prevalence of hepatitis virus and some arbovirus infec- tions in Kavango, northern SWANamibia. South African Medical 30urna1, 67, 50&502.

Joubert, J. J., Jupp, P. G., Prozesky, 0. W. & Lourens, J. G. H. (198Sb). Infection of wild populations of the tampan, Ornithodoros moubata Murray, with hepatitis B virus in the Kavango, northern Namibia. South African 3ournal of Science, 81, 167-168.

Jupp, P. G., Joubert, J. J., Cornel, A. J., Swanevelder, C. & Prozesky, 0. W. (1987). An experimental assessment of the tampan tick Ornithodoros moubata Murray, as vector of hepatitis B virus. Medical and Veterinary Entomology, 1, 361-368.

Murdock, G. P. (1959). Africa; iu Peoples and their Culture History, New York: McGraw-Hill, pp. 365 et seq.

Neter, J,., Wasserman, W. & Whitmore, G. H. (1979). Applwd Statistics. Boston: Allyn & Bacon, p. 296.

Van der Vegte, J. H., Forster, C. W. & Forse, W. D. (1983). Epstern Caprivi Regional ?evelo?ment Strategy. fz;;marttzburg: Physical Planning Directorate, pp.

Received’ 20 June 1990; revised 3 September 1990; accepted for publication 4 September 1990