GENITAL ULCERS are a common sexually transmitted disease (STD) in sub-Saharan Africa, and chancroid has been shown to be the most common cause of genital ulceration in men in at least some countries in this region.1–3 Although Haemophilus ducreyi, the causative organism of chancroid, may be cultured from the base of ulcers, culture detection from patients with a clinical diagnosis of chancroid has a sensitivity of only approximately 75% under field conditions.4,5 The recent development of H. ducreyi serologic assays for detection of antibodies has permitted assessment of the distribution and epidemiology of the agent.6,7 An enzyme-linked immunosorbent assay (ELISA) developed from lipooligosaccharide has a sensitivity of 96% and a specificity of 97% for detecting H. ducreyi antibodies in patients with culture-proven chancroid.7 The duration of humoral immunity after infection is unknown, but antibodies of the IgG class have been shown to persist for at least 18 months (unpublished data).
Several epidemiologic studies that included assessment of H. ducreyi seroprevalence have been conducted. H. ducreyi seroprevalence rates ranged from 10.4% in a community-based survey of adults in rural Uganda,8 to 19.4% among women's clinic attendees in Addis Ababa, Ethiopia,9 27.0% among prostitutes in Lima, Peru,10 and 71.7% among prostitutes in the Gambia.11 Presence of H. ducreyi antibody has been associated with serologic evidence of other STDs, including syphilis,9,11Neisseria gonorrhoeae,9Chlamydia trachomatis,9 and herpes simplex virus type 2 (HSV-2).9 Among Gambian prostitutes, there was a significant association between antibodies to H. ducreyi and human immunodeficiency virus (HIV)-2.11 In Uganda, there was a trend for an association between H. ducreyi and HIV-1 seropositivity, but it was not statistically significant.8 In the only published study with prospective data, the 1-year seroincidence of H. ducreyi was 1.2%.8
In 1993, a prospective cohort study of HIV-1-seronegative trucking company employees was initiated in Kenya to characterize HIV-1 seroincidence and risk factors. A substudy was conducted to evaluate prevalence and correlates of H. ducreyi antibodies, and the incidence of seroconversion to H. ducreyi. We were particularly interested in determining whether H. ducreyi seropositivity was an objective marker of high-risk behavior and if H. ducreyi seroincidence would be sufficiently high to serve as an objective end point for behavioral intervention trials.
Study Participants and Procedures
A prospective cohort study of male trucking company employees was initiated in March, 1993 in Mombasa, Kenya as part of the Preparation for AIDS Vaccine Evaluation/HIV Network for Prevention Trials (PAVE/HIVNET) program, as previously described.12 A mobile research clinic was held on site at each of six trucking companies once a week. After pretest counseling, 10 ml of blood was obtained from participating men for HIV-1 serologic testing. The HIV seroprevalence in this group of employees was 17.7%. Men returned in 1 week for posttest counseling and enrollment in the prospective cohort study, if HIV-1 seronegative. Men underwent a standardized enrollment questionnaire regarding demographic, occupational, sexual, and medical history, and a physical examination. Sera from enrolled men were tested for antibodies to H. ducreyi, Treponema pallidum, HSV-2, and C. trachomatis.
The enrolled men were asked to return every 3 months for follow-up visits. A follow-up questionnaire was completed, a physical examination performed, and HIV-1 serologic testing done at each visit. Serologic testing for H. ducreyi was repeated after 1 year for those men who were seronegative at enrollment.
Antibodies to H. ducreyi were detected by lipooligosaccharide ELISA as described by Alfa et al.,7 using purified lipooligosaccharide from the H. ducreyi strains HD524 and HD034.7 Immulon 1 microwells (Dynatech Labs, Chantilly, VA) were coated with the lipooligosaccharide preparation overnight and blocked with bovine serum albumin. Human sera were added at a dilution of 1: 100 and goat anti-human HRP (horseradish peroxidase) conjugate (Jackson Labs, Biocan Scientific, Mississauga, Ontario) was used at 1: 5,000. The wells were developed with the substrate ABTS (Sigma Chemical Co., St. Louis, MO) and the plate read at 405 nm. Cutoff values were established using the methods described by Alfa et al.7 An absorbance value of at least 0.40 ELISA units was used to define a positive reaction. Positive controls consisted of sera from men with culture proven chancroid.
The serologic diagnosis of HIV-1 was made by serial ELISAs (Enzygnost; Behring, Marburg, Germany; and Recombigen; Cambridge Biotech, Worcester, MA). Serologic testing for syphilis was performed with rapid plasma reagin and T. pallidum hemagglutination assay (TPHA). Testing for antibodies to HSV-2 was by Western blot.13 Densitometric analyses were used to distinguish HSV-2 from HSV-1 antibodies. C. trachomatis antibodies were detected using a microimmunofluorescence test.14
Data management and analysis were conducted using SPSS software. Wilcoxon rank sum tests were used to compare medians and distributions for continuous variables. Odds ratios (OR) and 95% confidence intervals (CI) were calculated to assess categorical data. Potential correlates of H. ducreyi seropositivity were evaluated using multiple logistic regression to adjust for possible confounding.
Between March, 1993 and December, 1993, 501 HIV-1 seronegative men were enrolled. Table 1 shows the characteristics of the men in the cohort. Their ages ranged from 16 to 62 years, with a median of 29 years. Fifty-nine percent of the men had a history of sex with a prostitute. Only 51% gave a history of condom use. Sexual contact with a prostitute in the past year was reported by 33% of men, two thirds of whom had engaged in unprotected sex. Thirteen percent of the men were uncircumcised.
The seroprevalence of H. ducreyi antibodies in this cohort at enrollment was 26.5%. The seroprevalence of the other STD pathogens was also high, with 8% of the men having antibodies to T. pallidum, 49% to HSV-2, and 41% to C. trachomatis. The demographic, occupational, and sexual characteristics of H. ducreyi seropositive and seronegative men are compared in Table 2. Seropositive men were significantly older than seronegative men and more likely to give a history of occupational travel for more than 2 weeks per month. Age at first sex was similar in the two groups, but H. ducreyi seropositivity was associated with more years of active sex, history of sex with a prostitute, number of sex partners in the past year, and frequency of unprotected sex with a prostitute in the past year. A history of alcohol consumption was significantly related to H. ducreyi seropositivity. Presence of antibody to H. ducreyi was significantly associated with presence of antibody to T. pallidum (OR 4.3, 95% CI 2.2–8.3), HSV-2 (OR 4.9, 95% CI 2.0–11.5), and C. trachomatis (OR 3.2, 95% CI 1.5–6.9).
To evaluate the independent association between risk factors and H. ducreyi seropositivity, a multiple logistic regression analysis was performed. As shown in Table 3, age (P < 0.001), married status (adjusted OR 2.1, 95% CI 1.1–3.8), history of prostitute contact (OR 1.9, 95% CI 1.2–3.0), and history of alcohol intake (OR 1.7, 95% CI 1.1–2.6) were independently associated with H. ducreyi antibodies, after controlling for occupational travel. Separate models were created to evaluate the association between seropositivity to other STD pathogens and H. ducreyi antibodies after adjusting for the demographic and exposure variables listed in Table 3. In these models, the adjusted ORs were 2.6 for TPHA (95% CI 1.3–5.3), 3.4 for C. trachomatis (95% CI 1.4–8.3), and 5.7 for HSV-2 (95% CI 2.0–16.4). A final model was created to evaluate the effect of circumcision status on H. ducreyi seropositivity after controlling for the core variables listed previously. Uncircumcised status conferred no increased risk of H. ducreyi antibody detection (adjusted OR 1.1, 95% CI 0.5–2.1).
Of the 368 men who were seronegative at enrollment, 241 were reevaluated for H. ducreyi antibodies after 1 year of follow-up. There were 9 seroconversions in 251 person-years of follow-up, giving a seroincidence of 3.6 seroconversions per 100 person-years. Of these nine men, four had a documented or reported episode of genital ulceration during the follow-up interval. In contrast, only 23 of 232 men who remained H. ducreyi seronegative had genital ulcer disease during follow-up.
The seroprevalence of H. ducreyi in this group of trucking company workers was 26.5%, a higher rate than has been previously reported for any nonprostitute cohort. This may be a conservative estimate, because HIV-1 seropositive men were excluded from this analysis. Seroprevalence rates for syphilis (8%), HSV-2 (49%), and C. trachomatis antibody (41%) were also high, confirming previous observations that truck drivers in this region are a high-risk population for STDs.15,16
High-risk sexual behavior was reported by most of the men, with 59% having a history of sex with a prostitute. Despite extensive acquired immunodeficiency syndrome education programs in Kenya, two thirds of men reporting prostitute contact within the past year had had unprotected sex. The presence of H. ducreyi antibodies was significantly correlated with a variety of measures of high-risk sexual behavior. In multivariate analysis, history of sex with a prostitute was associated with a twofold increased risk of H. ducreyi seropositivity, consistent with studies that found commercial sex workers to play a central role in the propagation of chancroid epidemics.17,18
Alcohol use was another behavior reported by about half the men in the cohort and was also found to be significantly correlated with H. ducreyi seropositivity. The relationship of alcohol and high-risk sexual behavior has been observed elsewhere. In a study in Harare, Zimbabwe, alcohol use was associated with a sevenfold increased risk of STD.19 In a national survey of high-risk behaviors conducted between 1988 and 1990 in the United States, Anderson et al. found that alcohol consumption increased the odds of sexual risk behavior by a factor of 2 to 3.20 A recent study found that alcohol infusion was associated with enhanced replication of HIV-1 in vivo, thought to be due to inhibition of CD8+ T-lymphocyte function.21 It is unknown whether immunomodulatory effects of alcohol could increase susceptibility to bacterial infections such as H. ducreyi. If so, this could be an alternative mechanism for the relationship between alcohol consumption and H. ducreyi seropositivity. In East Africa, truck stop bars along highways are a common meeting ground for truck drivers and prostitutes. The regular provision of condoms to the bars at the truck stops, to encourage protected sexual intercourse, may be an effective intervention to reduce STDs among these men.
Seropositivity to H. ducreyi was correlated with seropositivity to other STDs. This may simply reflect a common sexual exposure history (e.g., a man with sexual behavior that results in one STD is also at risk for other STDs). However, the associations between H. ducreyi and seroreactivity to syphilis, HSV-2, and C. trachomatis persisted after controlling for demographic and sexual exposure variables. This raises the possibility that presence of a genital ulcer, with attendant breach of epithelial integrity, may increase the likelihood of infection with a second genital ulcerative pathogen given exposure. This would be analogous to the cofactor role of genital ulcers in enhancing HIV transmission.22,23
Circumcision status has been shown to influence the susceptibility of men to HIV-1 and certain genital ulcerative diseases.15,23 In cross-sectional studies, an association between uncircumcised status and certain genital ulcerative diseases has been observed,24,25 although we are aware of no prospective studies that have evaluated this interaction. To our surprise, we found no association between circumcision status and H. ducreyi seropositivity. Circumcision status among trucking company workers is associated with religion, occupational travel, and sexual behavior,15 and it is possible that negative confounding masked a true association despite the multivariate analysis that was performed. Alternatively, the power of this study may have been inadequate to address this relationship because of the low prevalence of uncircumcised status in the cohort (only 13% of the men were uncircumcised). To detect a twofold difference in H. ducreyi seroprevalence between circumcised and uncircumcised men, the power of our study was only 66%.
Our results suggest that in regions where chancroid is common, the seroprevalence of H. ducreyi may be used in epidemiologic and intervention research to provide an objective assessment of the degree of sexual risk of a population. H. ducreyi seroincidence may also be useful as an objective end point for efficacy of behavioral or other intervention trials, especially in communities that engage in high-risk sexual behavior but whose HIV-1 seroincidence is too low to permit evaluation of such interventions with HIV-1 as an end point.
1. Nsanze H, Fast MV, D'Costa LJ, Tukei P, Curran J, Ronald A. Genital ulcers in Kenya: Clinical and laboratory study. British Journal of Venereal Disease 1981; 57:378–381.
2. Dangor Y, Fehler G, Exposto FD, Koornhof HJ. Causes and treatment of sexually acquired genital ulceration in southern Africa. S Afr Med J 1989; 76:339–341.
3. Bogaerts J, Ricart CA, Van Dyck E, Piot P. The etiology of genital ulceration in Rwanda. Sex Transm Dis 1989; 16:123–126.
4. Dylewski J, Nsanze H, Maitha G, Ronald A. Laboratory diagnosis of Haemophilus ducreyi:
sensitivity of culture media. Diagn Microbiol Infect Dis 1986; 4:241–245.
5. Fast MV, D'Costa LJ, Nsanze H, et al. The clinical diagnosis of genital ulcer disease in men in the tropics. Sex Transm Dis 1984; 11:72–76.
6. Museyi K, Van Dyck E, Vervoort T, Taylor D, Hoge C, Piot P. Use of an enzyme immunoassay to detect serum IgG antibodies to Haemophilus ducreyi.
J Infect Dis 1988; 157:1039–1043.
7. Alfa MJ, Olson N, Degagne P, et al. Humoral immune response of humans to lipooligosaccharide and outer membrane proteins of Haemophilus ducreyi.
J Infect Dis 1993; 167:1206–1210.
8. Wagner HU, Van Dyck E, Roggen E, et al. Seroprevalence and incidence of sexually transmitted diseases in a rural Ugandan population. Int J STD AIDS 1994; 5:332–337.
9. Duncan ME, Roggen E, Tibaux G, Pelzer A, Mehari L, Piot P. Seroepidemiological studies of Haemophilus ducreyi
infection in Ethiopian women. Sex Transm Dis 1994; 21:280–288.
10. Gotuzzo E, Sanchez J, Escamilla J, et al. Human T cell lymphotropic virus type 1 infection among female sex workers in Peru. J Infect Dis 1994; 169:754–759.
11. Pépin J, Dunn D, Gaye I, et al. HIV-2 infection among prostitutes working in the Gambia: Association with serological evidence of genital ulcer diseases and with generalized lymphadenopathy. AIDS 1991; 5:69–75.
12. Martin HM, Jackson DJ, Mandaliya K, et al. Preparation for AIDS vaccine evaluation in Mombasa, Kenya: Establishment of seronegative cohorts of commercial sex workers and trucking company employees. AIDS Res Hum Retroviruses 1994; 10(Suppl 2):S235–S237.
13. Ashley RL, Militoni J. Use of densitometric analysis for interpreting HSV serologies based on Western blot. J Virol Meth 1987; 18:159–168.
14. Wang SP, Grayston JT, Alexander ER, Holmes KK. Simplified microimmunofluorescence test with trachoma-lymphogranuloma venereum (Chlamydia trachomatis)
antigens for use as a screening test for antibody. J Clin Microbiol 1975; 1:250–255.
15. Bwayo J, Plummer F, Omari M, et al. Human immunodeficiency virus infection in long-distance truck drivers in east Africa. Arch Intern Med 1994; 154:1391–1396.
16. Carswell JW, Lloyd G, Howells J. Prevalence of HIV-1 in east African lorry drivers. AIDS 1989; 3:759–761.
17. Plummer FA, D'Costa LJ, Nsanze H, Dylewski J, Karasira P, Ronald A. Epidemiology of chancroid and Haemophilus ducreyi
in Nairobi, Kenya. Lancet 1983; 2:1293–1295.
18. Asin J. Chancroid: A report of 1402 cases. Am J Syph Gon Ven Dis 1952; 36:483–487.
19. Gwati B, Guli A, Todd CH. Risk factors for sexually transmitted disease amongst men in Harare, Zimbabwe. Cent Afr J Med 1995; 41:179–181.
20. Anderson JE, Dahlberg LA. High-risk sexual behavior in the general population: Results from a national survey, 1988–1990. Sex Transm Dis 1992; 19:320–325.
21. Bagasra O, Bachman SE, Jew L, et al. Increased human immunodeficiency virus type 1 replication in human peripheral blood mononuclear cells induced by ethanol: Potential immunopathogenic mechanisms. J Infect Dis 1996; 173:550–558.
22. Plummer FA, Simonsen JN, Cameron DW, et al. Cofactors in male–female sexual transmission of human immunodeficiency virus type 1. J Infect Dis 1991; 163:233–239.
23. Cameron DW, Simonsen JN, D'Costa LJ, et al. Female to male transmission of human immunodeficiency virus type 1:; risk factors for seroconversion in men. Lancet 1989; 2:403–407.
24. Parker SW, Stewart AJ, Wren MN, Gollow MM, Straton JAY. Circumcision and sexually transmissible disease. Med J Aust 1983; 2:288–290.
25. Cook LS, Koutsky LA, Holmes KK. Circumcision and sexually transmitted diseases. Am J Public Health 1994; 84:197–201.