Resistance and Susceptibility to Mycobacterium tuberculosis Infection and Disease in Tuberculosis Households in Kampala, Uganda

Abstract

Tuberculosis (TB), caused by Mycobacterium tuberculosis (Mtb), remains a major public health problem. Household contact studies identify children and adults along the spectrum from Mtb exposure to disease. In the Kawempe Community Health Study (conducted in Kampala, Uganda), 872 culture-confirmed pulmonary TB cases and their 2,585 contacts were enrolled during 2002-2012 and followed for up to 2 years each. Risk factors identified by time-to-event analysis for secondary TB differed among children, women, and men. Younger age (P = 0.0061), human immunodeficiency virus (HIV) (P = 0.0002), thinness (P = 0.01), absent bacille Calmette-Guérin vaccination (P = 0.002), and epidemiologic risk score (P < 0.0001) were risks for children. For women, risks were HIV (P < 0.0001), thinness (World Health Organization criteria; P < 0.0001), and epidemiologic risk score (P = 0.003). For men, HIV (P = 0.0007) and low body mass index (P = 0.008) resulted in faster progression to TB. Tuberculin skin testing (TST) identified contacts with Mtb infection and those with persistently negative TST. Risks for faster time to Mtb infection were identified, and included age (P = 0.0007), baseline TST induration (P < 0.0001), and epidemiologic risk score (P < 0.0001) only in children. Those with persistently negative TST comprised 10% of contacts but had no unique epidemiologic characteristics among adults. The burden of Mtb infection and disease is high in TB households, and risk factors for progression from exposure to infection and disease differ among children, women, and men.

Figure 1.

Tuberculosis (TB) household contact (HHC) study enrollment and demographic factors, Kawempe Community Health Study, Kampala, Uganda, 2002–2012. Among the HHCs enrolled, the mean number of contacts per household was 3.0 (range, 1–29), the mean number of adults per household was 1.4 (range, 0–17), and the mean number of children/household was 1.6 (range, 0–12). Mtb, Mycobacterium tuberculosis; HIV, human immunodeficiency virus; TST, tuberculin skin test.

Figure 2.

Study of tuberculosis (TB) household contacts, Kawempe Community Health Study, Kampala, Uganda, 2002–2012. A) Secondary TB events grouped according to age and binned according to the number of months of follow-up. B) Plot of the cumulative cases over the 24-month follow-up period with 95% confidence interval. The number of secondary TB cases and number at risk are shown in Web Table 6.

Figure 3.

Study of tuberculosis (TB) household contacts, Kawempe Community Health Study, Kampala, Uganda, 2002–2012. A) Tuberculin skin test (TST) conversion events grouped according to age and binned according to the number of months to conversion. B) Plot of the cumulative proportion of converters over the 24-month follow-up period with 95% confidence interval. The number of conversions and number at risk are shown in Web Table 6.