Interactions between HIV and tuberculosis in a rural area of Malawi.

A case-control study of adult tuberculosis (TB) was conducted in Karonga District Malawi comparing confirmed incident TB cases diagnosed between 1990 and 1995 with age-sex-neighborhood matched controls selected from a computerized population database. Controls from that study and those from a parallel study of leprosy started in 1988 were used to estimate age/sex/period-specific HIV prevalences for the population. The age and sex structure of the population observed during a 1986-89 total population survey was used as the basis upon which TB and HIV rates were estimated. The risk factors for HIV positivity among TB cases were also examined. 3% of the population aged 15 years and over in the district in 1988/89 were infected with HIV. That prevalence is now over 10% however relatively low in men under age 25 and over 55 and women over age 45. Among TB cases HIV seropositivity is associated with more schooling and better housing. 101/225 cases and 59/558 controls were HIV positive yielding an odds ratio of 7.4. For smear-positive cases the odds ratio was 6.3. Odds ratios were similar for all age groups and males and females. Using a constant odds ratio of 6.3 the proportion of TB cases attributable to HIV infection adjusted for age and sex for smear-positive cases was 17% in 1988/89 and 39% in 1992/95. However the actual impact of HIV upon TB is even greater that 39% because increased case numbers increase transmission to both HIV-infected and noninfected segments of the population.