Predictors of tuberculosis and human immunodeficiency virus co-infection: a case-control study

OBJECTIVES The human immunodeficiency virus (HIV) and Mycobacterium tuberculosis co-infection is a major global challenge. It is not clear why some HIV-positive people are co-infected with tuberculosis (TB) while others are not. This study addressed that question. METHODS This case-control study was conducted in Tehran, Iran in June 2004, enrolling 2,388 HIV-positive people. Cases were selected from those who were co-infected with TB and controls from those without TB. Multiple logistic regression analysis was performed to assess the association between M. tuberculosis/HIV co-infection and several predictors. Odds ratios (ORs) and their 95% confidence intervals (CIs) were calculated. RESULTS In this study, 241 cases were compared with 2,147 controls. Sex, age, marital status, educational level, imprisonment, smoking, narcotic use, route of HIV transmission, previous TB infection, isoniazid preventive therapy (IPT), antiretroviral therapy (ART), and low CD4 count (<350 cells/mm(3)) were independently associated with M. tuberculosis/HIV co-infection (p<0.001). However, after adjusting for all other variables in the model, only the association between M. tuberculosis/HIV co-infection and the following predictors remained statistically significant: imprisonment (odds ratio [OR], 3.82; 95% confidence interval [CI], 2.11-6.90), previous TB infection (OR, 5.54; 95% CI, 1.99-15.39), IPT (OR, 0.13; 95% CI, 0.06-0.31), ART (OR, 1.81; 95% CI, 1.26-2.61), and CD4 count <350 cells/mm(3) (OR, 2.34; 95% CI, 1.36-4.02). CONCLUSIONS Several predictors are associated with M. tuberculosis/HIV co-infection, but only a few indicators were significantly associated with M. tuberculosis/HIV co-infection. It is estimated that a number of predictors of M. tuberculosis/HIV co-infection remain unknown and require further investigation.