Objective: To determine whether a previously observed association between human immunodeficiency virus type 1 (HIV-1) infection and smoking in Haitian women could be explained by confounding high-risk behaviors.
Design and setting: A nested case-control study at a primary care health clinic in Cité Soleil, Haiti.
Participants: Women who reported having smoked (n = 89) and randomly selected nonsmokers (n = 329) who had participated in a survey 1 to 12 months earlier evaluating risk factors for HIV-1 infection.
Main outcome measures: Lifetime sexual practices, smoking, health beliefs and practices, and other factors potentially confounding the relationship between smoking and HIV-1 infection.
Results: Compared with nonsmokers, smokers reported higher rates of high-risk behaviors, including more lifetime sex partners (P less than .001), being less likely to be married (P less than .01), and being more likely to have visited folk healers (P less than .01). No intravenous drug use was reported, and no significant differences were noted between smokers and nonsmokers in numbers of past surgical or dental procedures or injections. Adjustment for all factors associated with HIV-1 infection and smoking in regression analyses revealed an independent association between smoking and HIV-1 infection (odds ratio [OR], 3.4; 95% confidence interval [Cl], 1.6 to 7.5). Other factors associated with HIV-1 infection included having more than two life-time sex partners (OR, 3.4; 95% Cl, 1.7 to 6.8) and lower socioeconomic status as reflected by a dirt floor in the home (OR, 8.6; 95% Cl, 3.3 to 22.0).
Conclusions: Smoking is a marker for high-risk sexual behavior and is associated with an increased risk of HIV-1 infection in this population. The persistent association between smoking and HIV-1 infection after adjustment for all known risk factors suggests the possibility of a biologic effect of smoking that warrants further evaluation in other populations.