Epidemiological aspects of tuberculosis in Guinea-Bissau, West Africa

Tuberculosis (TB) has plagued the humanity for several thousands of years. The bacteria causing TB is mainly spread from person to person as an aerosol transmission. It is estimated that one third of the world’s population is infected with the disease; about 10% of these will develop active TB during their lifetime. In year 2000, 8 million TB-cases were detected, of which 2 million died. Except for studies of the correlation between TB and HIV/AIDS, the epidemiology of TB in developing countries has been little investigated. The aims of this thesis have been to evaluate risk factors for positive tuberculin skin test (TST), which is the test for diagnosing TB infection, in the population and among TB-case household members, to assess incidence and risk factors for active TB, and to determine clinical predictors for death in TB patients.

A study area in Bissau, the capital of Guinea-Bissau, with a population of about 43,000 has been followed through a demographic census system for over 25 years. In 1996, a TB surveillance system was set up in the area; all adults (age >15 years) with active TB within the chest were included. The background data from the census system allows analyses comparing TB-cases with the whole population.

From 1996 to 2001 a total of 811 TB-cases were included and investigated. The risk of positive TST was closely related to previous or recent exposure to TB, and among the family members of the TB cases the extent of contact was important. BCG vaccination status did not confound the interpretation of the test. TST performed during the early rainy season, from June to August, resulted in smaller reactions, and testing during these months may confound the results. The incidence of active TB in the area was high; 471/100,000 for the adult population. Both HIV-1 and HIV-2 infection greatly increased the risk for TB, although the effect was considerably higher for HIV-1. Other independent risk factors for active TB corresponded well with factors believed to increase the risk also during the epidemic situation in Europe and USA during the 19th and the beginning of the 20th century: male sex, older age, poverty, less schooling and more adults in the family (adult crowding). Interestingly it seemed as if children had a protective influence; children in the household reduced the risk for TB among the adults. The mortality was increased for HIV-1-positive TB-cases, but also co-infection with HIV-2 increased the risk of death compared to HIV-negative. Signs of weakened immune function, such as oral candida infection, and malnutrition increased the risk of dying in both HIV-positive and HIV-negative patients. Estimating the degree of malnutrition using the mid-upper-arm circumference provided a good tool in the evaluation of increased risk for TB death. A civil war 1998-1999 caused involuntary treatment interruption among the patients which considerably increased the mortality, an effect that was most marked in HIV-positive. The increased mortality in HIV-positive TB-cases continued also after the war had ended.