Female sex workers and their male clients play important roles as core and bridging populations in the epidemiology of sexually transmitted infections (STI) and HIV even in countries with generalized HIV epidemics.
Since the 1990s, studies have reported the impact of a variety of targeted HIV/STI prevention and care interventions on STI and HIV rates among sex workers in sub-Saharan Africa. Behavioural interventions provided access to free condoms, information on HIV/STI and group or individual counselling on risk reduction [1–7]. Interventions to enhance the early diagnosis and treatment of STI included improved access to STI therapy at special clinics for sex workers [4,5], STI screening and treatment at regular time intervals of 1–6 months [1,4–6] and/or monthly antibiotic chemoprophylaxis [3,7]. Some studies were able to demonstrate that behavioural interventions in combination with improved access to STI care were associated with a reduction not only in STI prevalence but also in HIV incidence [1,2,4,8], whereas others found reductions in STI through enhanced STI treatment [3,7] but no reductions in HIV incidence .
In Tanzania, high rates of HIV and STI have been reported among women working in bars, restaurants and guesthouses (female barworkers) in trading centres situated along major transit routes [9–11]. In late 2000, the Mbeya Medical Research Programme established an open cohort of female barworkers in Mbeya Region, Tanzania and, among other studies, changes in STI prevalence and HIV incidence were assessed and risk factors for incident HIV and herpes simplex virus type 2 (HSV-2) infections determined.
A total of 600 female barworkers, aged 18–35 years, were recruited at 14 trading centres and towns along the highways in Mbeya Region in late 2000 and followed up at 3-monthly intervals for 27 months. Recruitment procedures and eligibility criteria have been described previously . An additional 153 women were recruited at nine follow-up examination rounds to replace women who were lost to follow-up. At each examination participants underwent a health check-up including a vaginal speculum examination. They received syndromic treatment for symptomatic STI and aetiological treatment for asymptomatic laboratory-confirmed infections with Trichomonas vaginalis (metronidazole) and Candida albicans (clotrimazole). Venous blood samples and genital swabs were taken for laboratory screening for HIV/STI. Data on sociodemographic and behavioural characteristics were collected through structured interviews. Participants were offered voluntary HIV counselling and testing, health education and free condoms.
Trichomonas vaginalis (TV) and Candida albicans (CA) infections were diagnosed by culture (InPouch TV; Biomed Diagnostics, USA) at recruitment and by wet preparation microscopy at subsequent examinations. Bacterial vaginosis was diagnosed using the Nugent method . Ulcer swabs were examined by multiplex polymerase chain reaction (PCR) for Treponema pallidum (TP), Haemophilus ducreyi (HD) and Herpes simplex virus (HSV). All cervical swabs collected at baseline plus a 30% random sample at alternate follow-up examinations were tested for Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) by PCR (Roche Diagnostics, USA). HSV-2 serology was assessed by enzyme immunoassay (EIA) (Kalon, UK), and syphilis serology by the Serodia-Treponema pallidum particle agglutination assay (TPPA) (Fujirebio Inc., Japan) and the rapid plasma reagin (RPR) test (VD25; Murex Diagnostics, UK). All sera underwent dual testing for HIV (Enzygnost Anti HIV1/2 Plus; Behring, Marberg, Germany; Determine HIV 1/2; Abbott, Germany). Discordant results were resolved by western blot (Genelabs Diagnostics, Switzerland).
Time trends in the prevalence of curable STI/RTI were evaluated using random effects logistic regression taking account of correlations between repeated measurements on individual women to obtain odds ratios (ORtrend) corresponding to the change from one 3-monthly examination to the next. To assess whether trends were explained by changes in sociodemographic or behavioural factors, crude odds ratios were compared with odds ratios adjusted in turn for each of these factors. Data from rounds 7 and 9 were excluded from this analysis because different study instruments were used to collect behavioural data during these rounds.
The incidence of HIV and HSV-2 was determined using the time at risk from the date of entry into the cohort until seroconversion (assumed midpoint between examinations) or last examination. Associations of other STI/RTI and of sociodemographic, behavioural and biological characteristics (measured at the visit prior to seroconversion) with incident HIV and HSV-2 were assessed using conditional logistic regression, conditioning on examination round. The final models included all variables significantly associated with incident HIV or HSV-2 at the 10% significance level.
Changes in sociodemographic characteristics and sexual behaviours over time
The baseline characteristics of the 600 women recruited initially and the 153 women recruited thereafter did not differ. The mean age was 25.5 years (range, 16–39 years). The proportion of participants reporting sex with a permanent partner remained stable at around 85% over time. Reported sex with casual partners within the last 4 weeks fluctuated between 27 and 38% at different rounds. Consistent condom use with casual partners during the last 4 weeks was reported by 33% of women initially, rising to 75% after post-HIV-test counselling in the second and third examination round, and falling to around 45% at 6 months thereafter. Initially 26% of women reported condom use during most recent sex with their permanent partner but in subsequent rounds this proportion varied between 5 and 14%.
Trends in STI prevalence over time
The prevalence of clinical STI/RTI signs and laboratory confirmed bacterial infections decreased over time for all STI/RTI with the exception of bacterial vaginosis (Table 1). In contrast, the prevalence of outcomes associated with viral STI, including genital warts and genital herpes did not decline.
Although HIV status and several sociodemographic and behavioural characteristics were associated with current infection with one or more of these STI , there was little change in the odds ratio for trend when adjusted for each of these factors (data not shown), suggesting that time-trends in STIs were not due to behaviour change. STI prevalence at baseline did not differ between women who were lost to follow-up and women who remained in the cohort. Observed STI trends persisted after restricting to women who were seen throughout (data not shown).
Prevalence and sero-incidence of HIV, HSV-2 and syphilis
Baseline HIV prevalence was 67% and was higher among participants who were lost to follow-up: 84.3% of 153 women lost to follow-up were HIV infected compared with 62.5% of 600 participants remaining in the study throughout (P < 0.001).
Incidence was determined for the entire period of 27 months and three consecutive time intervals of 9 months. Intervals were defined: (1) according to calendar period; and (2) according to individual follow-up time. HIV incidence declined significantly from 13.9/100 person-years in the first 9-months calendar period to 5.0/100 person-years in the last period (Table 2). The decline in incidence was similar when periods of individual follow-up time were compared: HIV incidence decreased from 13.2/100 to 6.1/100 person-years.
HSV-2 incidence was 17.3/100 person-years and there was no recognizable trend over consecutive periods of calendar time (Table 2). TPPA seroconversions decreased from 2.3/100 person-years during the first 9 months to 0.9 per 100 person-years during the last 9 months, although this decrease was not statistically significant (data not shown).
Risk factors for HIV and/or HSV-2 acquisition
In multivariate analysis incident HIV was significantly associated with genital ulcers and with positive syphilis serology (Table 2). Neither the presence of herpetic lesions (data not shown), nor HSV-2 seropositivity, nor the presence of other STI was associated with a significantly increased risk of HIV acquisition. Two socio-behavioural variables were significantly associated with incident HIV and thus included in the final multivariate model: the time that participants had worked as barworkers (higher risk in those recently joining profession) and whether they reported exchange of sex for money during the past 3 months.
The risk of HSV-2 seroconversion was increased in HIV-infected participants in multivariate analysis. Incident HSV-2 was also associated with change of residence during the year before enrolment and work in a traditional bar (Table 2).
The prevalence of HIV at enrolment was very high. In comparison, rates among antenatal care attendees in Mbeya Region were significantly lower (21.7% in Mbeya town and 17.4% in roadside settlements ). Even though the HIV epidemic is generalized in Tanzania, female barworkers may function as a core group and their clients as a bridging population for HIV transmission in the general population [14,15].
The decrease in bacterial STI in the presence of regular STI screening and treatment in Mbeya is consistent with other studies evaluating similar interventions in sex workers in Ivory Coast, Senegal, South Africa and Kenya [3,6,7,16]. As in previous studies, it was not possible to separate out the effects on STI prevalence of the behavioural and medical components of the intervention in Mbeya. The possibility that selective dropout of high-risk women caused the decline in STI prevalence can be excluded, as there was no difference in prevalence of bacterial STI at baseline between drop-outs and participants who remained in the study, and the observed STI trends persisted after restricting to women who did not drop out. Furthermore, time-trends in STI prevalence showed little change when adjusted for sociodemographic and behavioural characteristics. This implies that trends were not explained by behaviour changes, suggesting the biomedical component of the intervention (particularly STI treatment) may have been more important. However, these findings should be interpreted cautiously in view of the well-recognized limitations of reported data on sexual behaviour.
HIV incidence was very high at 10/100 person-years but declined from 14 to 5/100 person-years during the study. Declines in HIV incidence among cohorts of high-risk populations have been reported previously and possible reasons included greater loss to follow-up in subgroups at high HIV risk , a cohort attrition effect whereby women at highest risk seroconvert early during follow-up leaving a lower-risk population in later periods , the adoption of protective behaviours  and a decline in STI among cohort participants . In our study, among HIV-uninfected women, those lost to follow-up reported more risky behaviours than those followed to the end of the study, which may have led to an accumulation of HIV-uninfected participants with lower levels of risk behaviours during later follow-up rounds. There was no decrease in the reported number of casual sexual partners. However, data indicate that consistent condom use with casual partners may have increased. Both GUS and syphilis were shown to be significantly associated with HIV incidence and both were shown to decline significantly over time. However, it is not possible to determine how much each of the above factors contributed to the observed decline.
Few biological and behavioural characteristics were identified as risk factors for incident HIV. Having had a genital ulcer around the probable time of HIV infection increased the risk four-fold and positive syphilis serology increased it two-fold. This is in line with observations from several other studies examining the association between genital ulcers and HIV .
Bacterial vaginosis was associated with prevalent but not with incident HIV in our study suggesting that HIV infection may predispose to bacterial vaginosis (rather than vice versa), as previously concluded in a study of HIV-positive women in Thailand .
In contrast to findings from several other studies , in Mbeya neither symptomatic genital herpes nor HSV-2-seropositivity was associated with incident HIV. The lack of an association with symptomatic herpes is not surprising, as there were only eight occasions on which previously HIV-uninfected women were observed with a confirmed herpetic lesion. As three of seven HIV seroconversions in HIV/HSV-2-uninfected participants occurred simultaneously with HSV-2 seroconversion, it is possible that the two viruses were acquired concurrently.
In summary, a simple intervention consisting of regular 3-monthly STI/RTI screening and syndromic management in combination with HIV voluntary testing and counselling and HIV/STI information was well accepted by this cohort of barworkers and it is likely that it resulted in a decline in STI prevalence. The observed decline in HIV incidence may have been due partly to selective dropout of women with high-risk behaviours among the HIV-positive women.
1. Baeten JM, Richardson BA, Martin HL Jr, Nyang PM, Lavreys L, Ngugi EN, et al
. Trends in HIV-1 incidence in a cohort of prostitutes in Kenya: implications for HIV-1 vaccine efficacy trials. J Acquir Immune Defic Syndr 2000; 24:458–464.
2. Kaul R, Kimani J, Nagelkerke NJ, Fonck K, Keli F, MacDonald KS, et al
. Reduced HIV risk-taking and low HIV incidence after enrollment and risk-reduction counseling in a sexually transmitted disease prevention trial in Nairobi, Kenya. J Acquir Immune Defic Syndr 2002; 30:69–72.
3. Kaul R, Kimani J, Nagelkerke NJ, Fonck K, Ngugi EN, Keli F, et al
. Monthly antibiotic chemoprophylaxis and incidence of sexually transmitted infections and HIV-1 infection in Kenyan sex workers: a randomized controlled trial. JAMA 2004; 291:2555–2562.
4. Laga M, Alary M, Nzila N, Manoka AT, Tuliza M, Behets F, et al
. Condom promotion, sexually transmitted diseases treatment, and declining incidence of HIV-1 infection in female Zairian sex workers. Lancet 1994; 344:246–248.
5. Alary M, Mukenge-Tshibaka L, Bernier F, Geraldo N, Lowndes CM, Meda H, et al
. Decline in the prevalence of HIV and sexually transmitted diseases among female sex workers in Cotonou, Benin, 1993–1999. AIDS 2002; 16:463–470.
6. Ghys PD, Diallo MO, Ettiegne-Traore V, Satten GA, Anoma CK, Maurice C, et al
. Effect of interventions to control sexually transmitted disease on the incidence of HIV infection in female sex workers. AIDS 2001; 15:1421–1431.
7. Steen R, Vuylsteke B, DeCoito T, Ralepeli S, Fehler G, Conley J, et al
. Evidence of declining STD prevalence in a South African mining community following a core-group intervention. Sex Transm Dis 2000; 27:1–8.
8. Ghys PD, Diallo MO, Ettiegne-Traore V, Kale K, Tawil O, Carael M, et al
. Increase in condom use and decline in HIV and sexually transmitted diseases among female sex workers in Abidjan, Cote d'Ivoire, 1991–1998. AIDS 2002; 16:251–258.
9. Kapiga SH, Sam NE, Shao JF, Renjifo B, masenga EJ, Kiwelu IE, et al
. HIV-1 epidemic among female bar and hotel workers in northern Tanzania: risk factors and opportunities for prevention. J Acquir Immune Defic Syndr 2002; 29:409–417.
10. Nkya WM, Gillespie SH, Howlett W, et al
. Sexually transmitted diseases in prostitutes in Moshi and Arusha, Northern Tanzania. Int J STD AIDS 1991; 2:432–435.
11. Riedner G, Rusizoka M, Hoffmann O, Nichombe F, Lyamuya E, Mmbando D, et al
. Baseline survey of sexually transmitted infections in a cohort of female bar workers in Mbeya Region, Tanzania. Sex Transm Infect 2003; 79:382–387.
12. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol 1991; 29:297–301.
13. The United Republic of Tanzania MoH. National AIDS Control Programme HIV/AIDS/STI Surveillance Report. January – December 2000
, Report No 15. United Republic of Tanzania MoH 2000.
14. Lowndes CM, Alary M, Meda H, Gnintoungbe CA, Mukange-Tshibaka L, Adjovi C, et al
. Role of core and bridging groups in the transmission dynamics of HIV and STIs in Cotonou, Benin, West Africa. Sex Transm Infect 2002; 78 Suppl 1:i69–i77.
15. Morris M, Podhisita C, Wawer MJ, Handcock MS. Bridge populations in the spread of HIV/AIDS in Thailand. AIDS 1996; 10:1265–1271.
16. Meda N, Ndoye I, M'Boup S, Wade A, Ndiaye S, Niang C, et al
. Low and stable HIV infection rates in Senegal: natural course of the epidemic or evidence for success of prevention? AIDS 1999; 13:1397–1405.
17. Scheer S, Douglas JM Jr, Vittinghoff E, et al
. Feasibility and suitability of targeting young gay men for HIV vaccine efficacy trials. J Acquir Immune Defic Syndr Hum Retrovirol 1999; 20:172–178.
18. Grosskurth H, Mosha F, Todd J, Mwijarubi E, Klokke A, Senkoro K, et al
. Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: randomised controlled trial. Lancet 1995; 346:530–536.
19. Fleming DT, Wasserheit JN. From epidemiological synergy to public health policy and practice: the contribution of other sexually transmitted diseases to sexual transmission of HIV infection. Sex Transm Infect 1999; 75:3–17.
20. Rugpao S, Nagachinta T, Wanapirak C, Srisomboon J, Suriyanon V, Sirirojn B, et al
. Gynaecological conditions associated with HIV infection in women who are partners of HIV-positive Thai blood donors. Int J STD AIDS 1998; 9:677–682.
21. del Mar Pujades Rodriguez M, Obasi A, Mosha F, Todd J, Brown D. Herpes simplex virus type 2 infection increases HIV incidence: a prospective study in rural Tanzania. AIDS 2002; 16:451–462.