Objectives: To establish the prevalence of HIV infection in rural South Africa and to investigate demographic factors that influence this prevalence.
Design: An anonymous HIV seroprevalence survey was performed in conjunction with a population-based malaria surveillance programme.
Setting: The rural area of northern Natal/KwaZulu, South Africa.
Participants: A total of 5023 black African participants were recruited by malaria surveillance agents during house-to-house visits; each house in an endemic malaria area is visited approximately once every 6 weeks. Participants included 4044 healthy and 979 febrile individuals (i.e., suspected of having malaria).
Main outcome measures: HIV-1 and HIV-2 serological status, degree of mobility, age and sex.
Results: Sixty of the 5023 blood specimens were confirmed to be HIV-1-antibody-positive by Western blot, an overall prevalence of 1.2% (95% confidence interval, 0.9-1.5). None of the specimens was positive for HIV-2 antibodies. After adjusting for age, presence of fever and migrancy, women had a 3.2-fold higher prevalence of HIV-1 infection than men. HIV-1 infection was approximately three times more common among subjects who had changed their place of residence recently (2.9 versus 1.0%, P < 0.01).
Conclusions: The prevalence of HIV-1 infection is higher among women than men resident in rural Natal/KwaZulu, South Africa. This is at least in part the result of oscillatory migration, particularly of men who work in urban areas but have families and homes in rural areas. Migration is associated with a higher prevalence of HIV-1 infection, suggesting that improving social conditions so that families are not separated and become settled in their communities is one way to help reduce the spread of HIV-1.
PIP: As part of a population-based malaria surveillance program in late 1990, surveillance agents took blood samples from 979 people who had had a fever within the last 2 weeks and from 4044 healthy people during regular house-to-house visits in rural northern Natal/KwaZulu, South Africa, to determine HIV seroprevalence and risk factors of HIV infection. 60 (1.2%) people were HIV-1 seropositive. No one had HIV-2 infection. Febrile people had a 30% higher sex-adjusted relative risk (RR) of HIV-1 infection than healthy individuals, but this increase was insignificant. Women were at greater risk of HIV-1 infection than men (1.6% vs. 0.4%; age-adjusted RR = 3.8). In fact, this risk still existed when the researchers controlled for fever (RR = 3.75) and migrancy (RR = 3.2). The fall in the RR for women from 3.8 to 3.2 when controlled for migrancy suggested an underrepresentation of migrant male workers in the study sample. 2.3% of the women in their childbearing years (15-44) were HIV-1 seropositive, indicating an increased likelihood of transmission of HIV-1 to newborns. The youngest person afflicted with HIV-1 was a 12-year-old female and the oldest was a 66-year-old woman. No 10-to-19-year-old males tested HIV-1 positive, while 1.7% of the 10-to-19-year-old females did, suggesting that the young females had sex with older men. This may have indicated teenage prostitution and sexual abuse. 2.9% of the people who changed their place of residence within the last year (migrancy) had HIV-1 infection. For women it was linked to a 2.4 times higher RR (age-adjusted) of HIV-1 infection. For men, the age-adjusted RR was even greater (7.3). Even though HIV-1 seroprevalence was about 45% greater in areas crossed by the main national road than it was in other areas (1.3% vs. 0.9%), the difference was not significant. Since migrants were a key source of HIV-1 infection, improvement in social conditions, allowing families to live together and to settle in their communities, may reduce HIV-1 transmission.