The epidemic of human immunodeficiency virus (HIV)/AIDS has had a severe impact on the National Tuberculosis Control Program in Côte d'Ivoire, West Africa. In Abidjan, the principal city, the HIV seroprevalence among the ≈4,000 tuberculosis patients diagnosed annually from 1990-1994 was 46% (1); in contrast, the HIV seroprevalence among 18,099 delivering women in a large maternal clinic in Abidjan from 1990 to 1992 was 12.0% (9.4% HIV-1, 1.6% HIV-2, and 1.0% dually reactive) (2), and the HIV seroprevalence among 1,257 male blood donors in Abidjan in 1991 was 11.4% (3). Tuberculosis is the most common opportunistic illness among HIV-infected persons in Abidjan; an autopsy study conducted in Abidjan's largest hospital in 1991 revealed that tuberculosis was the leading cause of death (32%) among 247 HIV-infected patients (4).
To characterize HIV risk practices among men with tuberculosis in Abidjan, and to determine what factors are associated with HIV infection in this population, we analyzed data collected from patients enrolled in a cohort study between 1989 and 1992.
In Abidjan, all patients with newly diagnosed tuberculosis, regardless of socioeconomic status, are evaluated at two large ambulatory tuberculosis treatment centers. From July 1989 to February 1992, these patients were offered free, confidential HIV antibody testing that was performed according to a standardized serologic algorithm (5). During this period, a subset of HIV-infected and uninfected adult patients who were Abidjan residents and who had sputum smear-confirmed pulmonary tuberculosis were invited to participate in a prospective cohort study to compare their response to tuberculosis therapy. The results of this study have been described (6).
For each HIV-2-infected patient who was enrolled in the study, one HIV-1-infected patient and two HIV-negative patients were enrolled; all patients dually reactive to HIV-1 and HIV-2 were also invited to participate. Because 85% of the patients recruited into this cohort were men, the analyses that are presented in this article are restricted to men.
Standardized questionnaires were administered and physical examinations were performed in private examining rooms at the tuberculosis centers by trained Project RETRO-CI research physicians. The questionnaire included demographic characteristics such as age, sex, residence, country of birth, marital status, and level of education; and HIV risk factor information including condom use, circumcision, age of first sexual contact, homosexual contact, injecting drug use, blood transfusions, lifetime number of both female sex worker (FSW) and non-FSW partners, and history of sexually transmitted diseases (STDs) (genital ulcers and urethritis) in the past 5 years. The relative importance of various risk factors for HIV infection in this population was assessed in a multivariate analysis, which included each of the demographic and behavioral factors listed herein.
Information on numbers of sexual partners and STDs in the 1-year period before enrollment was also collected to assess recent HIV risk behavior. However, these variables were not included in the multivariate analysis because a large proportion of the HIV-seropositive men in this study were presumably already infected with HIV before this 1-year period and this their behaviors during the past year could not reasonably be assessed as risk factors for HIV infection, and because the numbers of lifetime sexual partners and STDs in the past 5 years already included this information.
For univariate analyses, odds ratios (OR) and 95% confidence intervals (CI) were calculated using Epi Info Version 5.1 (Centers for Disease Control and Prevention, Atlanta, GA, U.S.A.). For the multivariate analysis, the stepwise logistic regression procedure of SAS was used (SAS Institute Inc., Cary, NC, U.S.A.). In the multivariate analysis, when the value of an individual variable for a given patient was missing, the overall mean for all patients with known values for that variable was used; this procedure allowed all patients to be included in the model. The proportion of missing values did not exceed 5% for any variable other than duration of residence in Abidjan (15%). Because patients were partially matched on age during enrollment, analyses by age were not done.
Overall, 490 HIV-positive (HIV-P) and 239 HIV-negative (HIV-N) men were enrolled; the HIV-P men included 139 HIV-1, 127 HIV-2, and 224 dually reactive patients (Table 1). More than half of the men were unmarried, had received no formal education, were born outside of Côte d'Ivoire, and had lived in Abidjan for 10 or more years.
In the univariate analyses, no differences were found in the proportion of HIV-P and HIV-N men who were married (37 versus 36%, OR 1.1, 95% CI 0.8-1.5) or who had no formal education (54 versus 57%, OR 1.0, 95% CI 0.7-1.6). HIV-P men were significantly more likely to have been born in Côte d'Ivoire (41 versus 32%, OR 1.5, 95% CI 1.0-2.1) and to have lived 10 or more years in Abidjan (69 versus 57%, OR 1.6, 95% CI 1.1-2.4). Regarding HIV risk factors, HIV-P men were significantly more likely than HIV-N men to have never used condoms (76 versus 70%, OR 1.4, 95% CI 1.0-2.0), to be uncircumcised (15 versus 8%, OR 2.2, 95% CI 1.3-4.0), to have had their first sexual contact at ≤20 years old (84 versus 77%, OR 1.6, 95% CI 1.0-2.4), to have had sex with FSWs in their lifetime (83% versus 63%, OR 2.9, 95% CI 2.0-4.2), genital ulcer disease in the past 5 years (38 versus 15%, OR 3.4, 95% CI 2.2-5.2), and urethritis in the past 5 years (44 versus 23%, OR 2.6, 95% CI 1.8-3.8). No differences were found in the proportion of HIV-P and HIV-N men who had at least one lifetime non-FSW partner (98 versus 97%, OR 1.3, 95% CI 0.4-3.8), who had homosexual contact (0.2 versus 0.4%, OR 0.5, 95% CI 0.0-17.9), or who had injected drugs (0.2 versus 0%, OR undefined).
In the multivariate analysis, HIV-P men remained significantly more likely than HIV-N men to have lived in Abidjan for 10 or more years [adjusted odds ratio (AOR) 1.5, 95% CI 1.0-2.2], to be uncircumcised (AOR 2.0, 95% CI 1.3-3.3), to have had 10 or more (AOR 2.7, 95% CI 1.7-4.1) and one to nine (AOR 1.7, 95% CI 1.1-2.6) lifetime FSW partners, and to have had two or more (AOR 5.1, 95% CI 2.3-11.2) or one (AOR 1.9, 95% CI 1.2-3.1) episodes of genital ulcer disease in the past 5 years; two or more episodes of urethritis in the past 5 years also remained significant (AOR 1.7, 95% CI 1.1-2.4), although one episode of urethritis did not. No differences between HIV-P and HIV-N men were found by marital status, country of birth, educational level, condom use, age of first sexual contact, or number of lifetime non-FSW sexual partners.
Overall, 374 of 490 (76%) of HIV-P men and 166 of 239 (70%) of HIV-N men never used condoms, and only 6 of 490 (1.2%) of HIV-P men and 4 of 239 (1.7%) of HIV-N men always used condoms. Among men who reported sex with one or more FSWs in their lifetime, 327 of 393 (83%) of HIV-P men and 117 of 143 (82%) of HIV-N men never used condoms with FSWs, and only 5 of 393 (1.3%) and 3 of 143 (2.1%) always used condoms with FSWs. Among men who reported sex with one or more non-FSW partners in their lifetime, 445 of 476 (94%) of HIV-P men and 217 of 229 (95%) of HIV-N men never used condoms with their regular partners, and only 3 of 476 (0.6%) and 2 of 229 (0.9%) always used condoms with their regular partners. No statistically significant differences were found between HIV-P and HIV-N men with regard to condom use.
In the 1-year period before enrollment in the study (Table 2), HIV-P men were significantly more likely than HIV-N men to have had at least one FSW partner (43 versus 25%, OR 2.3, 95% CI 1.6-3.3), at least one episode of genital ulcer disease (24 versus 9%, OR 3.1, 95% CI 1.5-4.8), and at least one episode of urethritis (18 versus 8%, OR 2.7, 95% CI 1.5-4.8); no difference was found in the proportion with at least one non-FSW partner (84 versus 79%, OR 1.3, 95% CI 0.9-2.0).
When HIV-1-infected men were compared with HIV-2-infected and dually reactive men, no significant differences were found in the proportion who were married (38 versus 47 and 31%), born in Côte d'Ivoire (44 versus 43 and 37%), who had lived in Abidjan for 10 or more years (64 versus 73 and 70%), who had no education (55 versus 47 and 58%), who never used condoms (78 versus 81 and 73%), who were uncircumcised (15 versus 13 and 17%), who had their first sexual contact at ≤20 years old (80 versus 87 and 86%), who had at least one lifetime FSW partner (83 versus 82 and 83%), who had at least one lifetime non-FSW partner (99 versus 97 and 98%), and who had at least one episode of genital ulcer disease in the past 5 years (35 versus 27 and 45%). A similar proportion of HIV-1 and HIV-2 infected men (38 versus 39%), but a lower proportion of HIV-1 infected than dually reactive men (38 versus 51%, OR 0.6, 95% CI 0.4-0.9) had at least one episode of urethritis in the past 5 years.
This study encapsulates the behavioral dynamics that have fueled the epidemic of HIV/AIDS in Côte d'Ivoire, and highlights the important role of commercial sex in sustaining this epidemic. The majority of men described in this analysis were unmarried, lacked formal education, and were immigrants to Côte d'Ivoire. When these factors are considered in conjunction with the strikingly high numbers of sexual partners and STDs and the virtual absence of regular condom use in this population, both the susceptibility and the contribution of these men to the rapid spread of HIV is apparent.
Sexual contact with FSWs was found to be strongly associated with HIV infection in these men. This finding is not surprising in view of the 80% HIV seroprevalence that has been documented in FSWs in Abidjan (7), and is consistent with previous observations in male blood donors and in male STD clinic attenders in Abidjan (3,8). The strong association between HIV infection and genital ulcer disease has been described previously and can be explained either because genital ulcer disease may be a risk factor for HIV transmission (9-11) and/or because genital ulcers may represent an opportunistic illness in immunosuppressed HIV-infected persons (7,12). A previous study in FSWs in Zaire suggested that nonulcerative STDs may also be a risk factor for HIV infection (13); it is uncertain whether our data support this observation because the strength of the association between urethritis and HIV that was found in the univariate analysis decreased considerably in the multivariate analysis. Lack of circumcision was found to be associated with HIV infection, also consistent with previous reports (14).
Perhaps the most critical finding of this investigation was the large numbers of sexual partners and STDs of both HIV-P and HIV-N men in the past year. Among HIV-P men, many of whom were presumably already infected with HIV during this period, 43% had at least one FSW partner and 83% had at least one non-FSW partner in the past year. With the virtual absence of regular condom use in this population, these data demonstrate the high leve of HIV exposure of both FSWs and other women in Abidjan. In addition, the high prevalence of STDs in HIV-P men in the past year (23% with genital ulcer disease and 18% with urethritis) indicates that these men might be very efficient transmitters of HIV. Last, even among HIV-N men with tuberculosis, 25% had at least one FSW partner in the past year. Given the extremely high HIV seroprevalence of FSWs in Abidjan, it is clear that these men are at great risk of becoming infected with HIV themselves.
The finding that HIV-1, HIV-2, and dually reactive men with tuberculosis had strikingly similar HIV risk profiles is consistent with previous work demonstrating that, as with HIV-1 transmission, sexual contact is the primary mode of HIV-2 transmission in West Africa (15,16). The significantly higher rate of urethritis, and the higher (although not significantly) rate of genital ulcer disease in the past 5 years among dually reactive men compared with HIV-1-infected men (51 versus 38% and 45 versus 35%) could be related to increased sexual exposure both to STDs and to HIV among dually reactive men.
These data should be interpreted with caution because male tuberculosis patients are not necessarily representative of all men in Abidjan. However, patients from all socioeconomic and educational backgrounds are treated in the tuberculosis clinics, because tuberculosis therapy in Abidjan is restricted to the two centers where this study was conducted. In addition, the rates of FSW contact, non-FSW contact, and STDs that were demonstrated in men with tuberculosis in this study are very similar to those found among male blood donors in Abidjan during this same period (3). No data on HIV risk behaviors among women in Abidjan were analyzed during this investigation, and further studies are being conducted to address this issue.
These data present a clear public health imperative for the widespread education of men in Abidjan about the risks of commercial sex and the importance of using condoms to prevent HIV infection and other STDs. Last, this study also suggests that HIV counseling programs should be implemented in health care settings, such as tuberculosis clinics, which provide medical care to large numbers of sexually active adults.
Acknowledgment: We thank Dr. Meade Morgan of CDC for his assistance with the multivariate analysis.
1. Djomand G, Greenberg AE, Sassan-Morokro M, et al. The epidemic of HIV/AIDS in Abidjan
, Côte d'Ivoire: a review of data collected by Projet RETRO-CI from 1987 to 1993. JAIDS
2. Adjorlolo-Johnson G, DeCock KM, Ekpini E, et al. Prospective comparison of mother-to-child transmission of HIV-1 and HIV-2 in Abidjan
, Ivory Coast. JAMA
3. Schutz R, Savarit D, Kadjo J-C, et al. Excluding blood donors at high risk of HIV infection in a west African city. Br Med J
4. Lucas SB, Hounnou A, Peacock C, et al. The mortality and pathology of HIV infection in a West African city. AIDS
5. DeCock KM, Porter A, Kouadio J, et al. Rapid and specific diagnosis of HIV-1 and HIV-2 infections: an evaluation of testing strategies. AIDS
6. Kassim S, Ackah A, Abouya L, et al. Mortality and response to chemotherapy in HIV-positive and HIV-negative pulmonary tuberculosis
patients in Abidjan
, Côte d'Ivoire. Presented at Eighth International Conference on AIDS in Africa and Eighth African Conference on Sexually Transmitted Diseases. Marrakech, December 12-16, 1993. (Abstract ThPB054).
7. Ghys PD, Diallo MO, Traoré-Ettiègne V, et al. Genital ulcers associated with HIV-related immunosuppression in female sex workers in Abidjan
, Côte d'Ivoire. J Infect Dis
8. Diallo MO, Ackah AN, Lafontaine M-F, et al. HIV-1 and HIV-2 infections in men attending sexually transmitted disease clinics in Abidjan
, Côte d'Ivoire. AIDS
9. Cameron DW, D'Costa LJ, Maitha GM, et al. Female to male transmission of human immunodeficiency virus type 1: risk factors for seroconversion in men. Lancet
10. Greenblatt RM, Lukehart SA, Plummer FA, et al. Genital ulceration as a risk factor for human immunodeficiency virus infection. AIDS
11. Simonsen JN, Cameron DW, Gakinya MN, et al. Human immunodeficiency virus infection among men with sexually transmitted diseases: experience from a center in Africa. N Engl J Med
12. Cameron DW, Ngugi EN, Ronald AR, et al. Condom use prevents genital ulcers in women working as prostitutes: influence of human immunodeficiency virus infection. Sex Transm Dis
13. Laga M, Manoka A, Kivuvu M, et al. Non-ulcerative sexually transmitted diseases as risk factors for HIV-1 transmission in women: results from a cohort study. AIDS
14. Moses S, Plummer FA, Bradley JE, et al. The association between lack of male circumcision and risk for HIV infection: a review of the epidemiologic data. Sex Transm Dis
15. De Cock KM, Brun-Vézinet F, Soro B. HIV-1 and HIV-2 infections and AIDS in West Africa. AIDS
16. Poulsen A-G, Kvinesdal B, Aaby P, et al. Prevalence of and mortality from human immunodeficiency virus type 2 in Bissau, West Africa. Lancet
Keywords:© Lippincott-Raven Publishers.
HIV/Risk factors; STD; Tuberculosis; Abidjan