Article Text

Download PDFPDF

HIV and syphilis prevalence and associated risk factors among fishing communities of Lake Victoria, Uganda
  1. Gershim Asiki1,
  2. Juliet Mpendo2,
  3. Andrew Abaasa1,
  4. Collins Agaba1,
  5. Annet Nanvubya2,
  6. Leslie Nielsen2,
  7. Janet Seeley1,3,
  8. Pontiano Kaleebu1,2,
  9. Heiner Grosskurth1,4,
  10. Anatoli Kamali1
  1. 1Medical Research Council (MRC)/Uganda Virus Research Institute (UVRI)—Uganda Research Unit on AIDS, Entebbe, Uganda
  2. 2UVRI/International Aids Vaccine Initiative (IAVI)—HIV Vaccine Program, Entebbe, Uganda
  3. 3School of International Development, University of East Anglia, Norwich, UK
  4. 4London School of Hygiene and Tropical Medicine, London, UK
  1. Correspondence to Gershim Asiki, Medical Research Council Research Unit on AIDS, Uganda Virus Research Institute, PO Box 49 Entebbe, Uganda; gershim.asiki{at}


Objectives Recent publications suggest that fishing populations may be highly affected by the HIV epidemic. However, accurate data are scarce. The authors determined HIV and syphilis prevalence and associated risk factors in a fishing population of Lake Victoria in Uganda.

Methods 10 188 volunteers aged ≥13 years from a census carried out in five fishing communities between February and August 2009 were invited to attend central study clinics established in each community. After informed consent, 2005 randomly selected volunteers responded to socio-demographic and risk assessment questions, provided blood for HIV testing and 1618 volunteers were also tested for syphilis. Risk factors were analysed using logistic regression.

Results HIV and active syphilis (rapid plasma reagin titre ≥1:8) prevalences were 28.8% (95% CI 26.8 to 30.8) and 4.3% (95% CI 3.3 to 5.4), respectively, and high risk sexual behaviour was frequently reported. HIV prevalence was independently associated with female sex, increasing age, occupation (highest in fishermen), relationship to household head, self-reported genital sores and knowledge of an HIV infected partner. Alcohol consumption, syphilis and sexually transmitted infections (STIs) reported by health workers were associated with HIV in women, and genital discharge and inconsistent condom use in men. Syphilis prevalence was independently associated with age and alcohol consumption in women, and recent genital sores and sex under the influence of drugs in men.

Conclusion This fishing population characterised by a very high HIV prevalence, high syphilis prevalence and frequently reported sexual risk behaviours, urgently needs improved STI services and targeted behavioural interventions.

  • HIV
  • syphilis
  • risk factors
  • fishing communities
  • Uganda
  • Africa
  • prevalence

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Fisher-folk have been described as vulnerable to HIV infection and other sexually transmitted infections (STIs).1–4 In Uganda, as in other countries in sub-Saharan Africa, HIV/AIDS prevention and treatment services have not yet effectively reached fishing communities. Risk behaviour and infection rates are believed to be high; however, there is a paucity of robust data.5 6 We addressed this information gap by examining HIV and syphilis prevalence and associated risk factors in a fishing population from Uganda.


In order to identify participants for a cohort study to determine HIV incidence and to assess their suitability for future intervention studies including HIV vaccine research, a cross-sectional study was conducted among a representative sample of adults from five intentionally selected fishing communities situated on the shores of Lake Victoria in southwest Uganda. This provided an opportunity to investigate the prevalence of HIV infection and sero-syphilis, and the risk factors associated with these infections. Detailed study objectives of this larger study and the criteria used for the selection of communities and participants are provided in the online supplementary material. Between February and August 2009, a census was conducted in these communities and a random sample of individuals aged ≥13 years (see online supplementary material), stratified by sex and age groups to reflect census distribution, were invited to attend central study clinics, one of which was established in each community. Consenting participants were interviewed on demographic factors, sexual behaviour and recently experienced STIs. Blood samples were collected to test for HIV and serological syphilis. A rapid HIV test was performed (Determine; Abbott Laboratories, Diagnostic division, Illinous, Chicago USA) and all positive specimens were confirmed by two independent ELISA tests (Vironostika HIV Uni-Form II plus 0 microelisa system, Biome′rieux, Boxtel, The Netherlands) and Murex HIV-1.2.0 Murex, Biotech Limited, Dartford, UK) and by western blotting if indeterminate. The rapid plasma reagin test (RPR; Biotec Laboratories Limited, Ipswich, Suffolk, UK) was used to test for syphilis and if positive was confirmed by the Treponema pallidum haemagglutination assay (TPHA; Biotec Laboratories Limited Ipswich Suffolk, UK). Specimens with positive RPR and TPHA test results were defined as indicative of active syphilis if the RPR titre was ≥1:8.

Univariable analyses were carried out separately for men and women. A likelihood ratio test (p<0.05) was used to select variables for multivariate analysis. Residence in a particular fishing community was included as a fixed effect in all models. For a sensitivity analysis, generalised estimating equation models were fitted with residence in a particular fishing community as a clustering variable. Residence was not fitted as a random effect due to concerns about the stability of the numerical approximation when using a small number of large clusters. Stepwise multiple logistic regression models were fitted to all variables identified as significant in the univariate analyses (see online supplementary material).

Ethical approval was obtained from the Science and Ethics Committee of UVRI and from the Uganda National Council of Science and Technology.


Demographic characteristics

The five communities had a total population of 15 415 people, 10 188 (66%) of whom were ≥13 years old. Of these, 5186 were men (51%), 30% of whom worked as fishermen. Only 56% of the adults were married, 7% had no formal education and 2% were regular frequent visitors. Most men (69%) and women (61%) were involved in small-scale businesses and 28% of the women were housewives with no own source of income. During the last month, more men than women (47% vs 25%) had spent two nights or more away from home. Of the 2074 (20%) individuals selected for the cross-sectional study, 2005 (97%) accepted and were enrolled. Of these, 1618 (81%) also agreed to be tested for serological syphilis, while the remainder opted out because results could not be made available the same day. The study population reflected the demographic characteristics found during the census, except for the 13–17 years age group that was under-represented in one community and fishermen and regular visitors who were somewhat over-represented in two others (see details in online supplementary table 1). The mean age of enrolled participants was 29 years (SD 8 years).

Behavioural characteristics

For the past 3 months, 87% of the participants reported to have been sexually active, with up to 12 different partners. Two or more partners were reported by 46% of male participants, ranging from 0.85% in the age group 13–17 years to 48% among those aged 25–34 years. Corresponding data from women were 15%, 4% and 42%, respectively. A high number of these partners (37%) were described as ‘new partners’, for whom only 9% of the respondents reported consistent condom use. Of the 4% of the participants (45 men and 43 women) who knew that they had an HIV-positive partner, only 9% (men 11%, women 7%) reported that they had used condoms consistently with them.

Use of alcohol, marijuana and khat was reported by 51%, 4% and 3%, respectively, and 35% admitted that they had sex under the influence of these substances. Condom use was lower among daily and weekly alcohol consumers compared with non-consumers (25% vs 50%). Many participants reported that they had either received or given gifts, for example, fish or clothing, in exchange for sex (31% and 37%, respectively). This applied to 41% and 21% of women and 21% and 49% of men, respectively.

HIV and associated factors

Overall, HIV prevalence was 28.8% (95% CI 26.8 to 30.8) and higher in women than men (33.9% vs 23.9%, p<0.001). On multivariate analysis, HIV prevalence was independently and significantly associated with age, occupation, relationship to head of household, knowledge of partner's HIV status and a recent history of STIs (table 1). It increased with age, was higher among fishermen and housewives than other groups, among female spouses of head of households, among those reporting a known HIV-positive partner and among those who had a genital sore during the last 3 months. In addition, among women alcohol consumption, reporting an STI diagnosed by a health worker during the past 3 months and having active syphilis were significantly associated with HIV infection, while among men an association was observed between HIV infection and genital discharge reported for the last 3 months. Inconsistent condom use among men also remained associated with HIV, although this reached only borderline significance (table 1).

Table 1

HIV prevalence and associated risk factors

Marital status, spending more than two nights/month away from home, reporting multiple or new sex partners and reporting gift exchange for sex were not significantly associated with HIV infection (data not shown).

Syphilis and risk factors

The prevalence of active syphilis was 4.3% (95% CI 3.3 to 5.4) with no significant difference between men and women (4.4% vs 4.1%). In the multivariate analysis, syphilis was independently and significantly associated with younger age in both sexes, reaching peak levels of 9.6% and 7.1% in men and women, respectively. Reporting genital sores and sex under the influence of drugs were significantly associated with syphilis in men. In women, alcohol consumption showed a borderline significant association with syphilis (see online supplementary table 2). The prevalence of serological syphilis with any RPR-/TPHA-positive result regardless of RPR titre was 5.9% (95% CI 4.8 to 7.2) with no difference between men and women (6.0% vs 5.9%).


The prevalence of HIV infection in these fishing communities was high (29%), exceeding the national average level among adults (6.4%) by far.7 It was higher in women than in men, reaching peak levels of around 40% in some communities. Comparatively high levels were seen already during early adulthood, with 12% and 26% in men and women aged 18–24 years, respectively. A high proportion of the study population reported risky sexual behaviour (multiple sex partners, low condom use even with partners known to be HIV infected, transactional sex and sex under the influence of alcohol or drugs). HIV infection was independently and significantly associated with recently observed symptoms or signs of a genital infection in both men and women, and with alcohol use among women.

The strengths of this study include the representativeness of the respondents as the stratified random sampling was guided by census data, the high participation rate and the combination of behavioural information with biological data. Information on all individuals rather than on selected occupational groups such as fishermen provided valuable insights into the socio-demographic fabric of this society, which will also assist the design of interventions. Robust epidemiological and behavioural data on fisher-folk in Africa are scarce and our study helps to close this gap. A limitation of the study was that nearly 20% of the respondents did not agree to be tested for syphilis because we were not able to provide serology results on the same day for logistical reasons, although this was possible for HIV tests. Furthermore, we were not able to perform laboratory tests for other STIs, due to logistical and financial reasons.

For Uganda, we know of only one other rather outdated study from fishing communities that reported an HIV prevalence of 24% across four villages situated at Lake Albert, slightly lower than that in our study.8 A more recent study from Kenya showed a similar prevalence (26%) among fishermen from Lake Victoria.9 The reported risk behaviour in that study was marked by high mobility, a high rate of partner change, a low level of condom use and frequently reported transactional sex, a pattern quite similar to that seen in our data set. The obvious scarcity of epidemiological data on HIV infection from fishing communities is surprising, given that HIV infection was first described in East Africa in a fishing population3 and that the importance of fishing populations for the HIV epidemic and its control have been emphasised by various authors over time.4–6 10 11

The prevalence of active (high titre) serological syphilis in our study was far less than that of HIV infection, but at 4.3% it was still high compared with the prevalence of active syphilis of 1% documented among the general rural population of the same area.12 Interestingly, the prevalence for TPHA-/RPR-positive sero-syphilis of any RPR titre, which is likely to include long-standing low-level active as well successfully treated cases with low RPR titres (the so-called sero-scars), was much lower in the fishing communities of our study than in the general population, that is, 6% versus 11%.12 The majority of the syphilis infections in our study were high titre cases. These observations suggest that syphilis has been introduced comparatively recently in fishing communities, and that it may still be on the rise. It would therefore be important to monitor this epidemic closely and counter it by establishing effective diagnostic and treatment services as soon as possible, to the benefit of both the fishing population itself and the wider population with which it is connected through trade and other forms of mobility.

As expected, a recent history of genital sores was associated with syphilis in men. The lack of such an association in women may be due to the female anatomy and the fact that syphilitic ulcers are usually not painful, and therefore may remain unnoticed. The pattern of behavioural risk factors was less obvious for syphilis than for HIV infection, but sex under the influence of drugs or alcohol played a role here as well. A history of other STI symptoms over the last 3 months was also frequent, and although we could not present laboratory-based data, it is likely that much of this may have been caused by bacterial or protozoal infections that would also call for the provision of effective STI services within fishing communities.

Our data show that it is not just fishermen who are vulnerable to HIV and STIs. Fishermen are part of a wider sexual network. They interlink with fish traders, food vendors, alcohol brewers and other service providers, many of whom are women. At least in part, the vulnerability of these subgroups stems from the complex nature of the fishing occupation and fishing lifestyle.13

The societal context is characterised by an unreliable income for both men and women, frequent absence from home for extended periods of time again among men and women, and in the case of the fishermen a frequent exposure to physical danger. From our experience, there is anecdotal evidence that fishermen discount the threat of HIV as a possibly lethal infection because they face the more immediate daily threat of drowning. Among women, the pressure to secure sufficient cash to complement the family income combines with their often weak status with regard to decision making and with a fairly open attitude to sexuality. All this facilitates a culture of exchanging favours including ‘sex for fish’.10 11

Alcohol consumption plays an important role in this society in general and in the acquisition of HIV among women in particular, and any intervention must take this issue into account. Alcohol use lowers risk inhibition, and reduces the ability to negotiate and make use of safe sex methods, and both are likely to affect women disproportionately.10 11 14

There are a number of similarities between fishing communities and other mobile occupational groups, in particular with mining populations, for example, in South Africa, which for some time have enjoyed much more intensive attention from health and welfare services.15 Indeed, many fishing villages in Uganda (including the communities in our study) and elsewhere in Africa have not yet been reached by public health services in general or focused HIV intervention efforts in particular.

It is high time to address this problem. However, our data suggest that a holistic approach is required, which combines good health services with well designed locally appropriate educational and community development efforts tailored to the entire fishing community including all its subgroups.13

Key messages

  • Fishing communities are at a significant risk of HIV but have received little attention from the existing HIV intervention programmes and research.

  • We determined prevalence and risk factors for HIV and syphilis in select fishing communities in Lake Victoria, Uganda, to inform health decisions on interventions.

  • Women and men aged 13 years and above had a HIV prevalence of 29% and active syphilis prevalence of 4%, and high risk sexual behaviour was frequently reported.

  • Improved STI services and targeted behavioural and biomedical interventions are urgently needed for this population.


We acknowledge the support given by the European and Developing Countries Clinical Trials Partnership (EDCTP), which funded this study, and the UK Medical Research Council (MRC), which provided institutional and salary support. We thank the study participants, field study teams, and the laboratory, data management and administrative staff for their contribution to the work.



  • Funding European and Developing Countries Clinical Trials Partnership(EDCTP), The Hague, The Netherlands.

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the Uganda Virus Research Institute.

  • Provenance and peer review Not commissioned; externally peer reviewed.