Skip to main content
  • Loading metrics

The epidemiology of HIV infection among female sex workers in Nairobi, Kenya: A structural determinants and life-course perspective

  • Tara S. Beattie ,

    Contributed equally to this work with: Tara S. Beattie, Wendy Adhiambo

    Roles Conceptualization, Formal analysis, Funding acquisition, Investigation, Writing – original draft

    Affiliation Department of Global Health and Development, London School of Hygiene and Tropical Medicine, London, United Kingdom

  • Wendy Adhiambo ,

    Contributed equally to this work with: Tara S. Beattie, Wendy Adhiambo

    Roles Investigation, Writing – original draft

    Affiliation Partners for Health and Development in Africa, Nairobi, Kenya

  • Rhoda Kabuti,

    Roles Investigation, Project administration, Validation, Writing – review & editing

    Affiliation Partners for Health and Development in Africa, Nairobi, Kenya

  • Alicja Beksinska,

    Roles Methodology, Writing – review & editing

    Affiliation Department of Global Health and Development, London School of Hygiene and Tropical Medicine, London, United Kingdom

  • Pauline Ngurukiri,

    Roles Data curation, Validation

    Affiliation Partners for Health and Development in Africa, Nairobi, Kenya

  • Hellen Babu,

    Roles Investigation, Project administration, Validation, Writing – review & editing

    Affiliation Partners for Health and Development in Africa, Nairobi, Kenya

  • Mary Kung’u,

    Roles Investigation, Validation, Writing – review & editing

    Affiliation Partners for Health and Development in Africa, Nairobi, Kenya

  • Chrispo Nyamweya,

    Roles Investigation, Validation, Writing – review & editing

    Affiliation Partners for Health and Development in Africa, Nairobi, Kenya

  • Anne Mahero,

    Roles Investigation, Validation, Writing – review & editing

    Affiliation Partners for Health and Development in Africa, Nairobi, Kenya

  • Maisha Fiti Study -Champions ,

    Membership of the Maisha Fiti Study Champions is provided in the Acknowledgments.

    Affiliation Partners for Health and Development in Africa, Nairobi, Kenya

  • Erastus Irungu,

    Roles Investigation, Resources

    Affiliation Partners for Health and Development in Africa, Nairobi, Kenya

  • Peter Muthoga,

    Roles Investigation

    Affiliation Partners for Health and Development in Africa, Nairobi, Kenya

  • Janet Seeley,

    Roles Conceptualization, Funding acquisition, Project administration, Writing – review & editing

    Affiliations Department of Global Health and Development, London School of Hygiene and Tropical Medicine, London, United Kingdom, Africa Health Research Institute, KwaZulu-Natal, South Africa

  • Joshua Kimani,

    Roles Conceptualization, Project administration, Supervision, Writing – review & editing

    Affiliations Partners for Health and Development in Africa, Nairobi, Kenya, UNITID, University of Nairobi, Nairobi, Kenya

  • Helen A. Weiss ,

    Roles Conceptualization, Funding acquisition, Supervision, Writing – review & editing

    ‡ These authors also contributed equally to this work.

    Affiliation Department of Infectious Disease Epidemiology and International Health, MRC International Statistics and Epidemiology Group, London School of Hygiene and Tropical Medicine, London, United Kingdom

  •  [ ... ],
  • Rupert Kaul

    Roles Conceptualization, Funding acquisition, Supervision, Writing – review & editing

    ‡ These authors also contributed equally to this work.

    Affiliation Department of Immunology, University of Toronto, Ontario, Canada

  • [ view all ]
  • [ view less ]


High HIV prevalence among female sex workers (FSWs) is heavily influenced by structural determinants (e.g. criminalisation of sex work; violence) and significant life events (e.g. orphanhood, teenage pregnancy). This study aims to understand the epidemiology of HIV among FSWs in Nairobi, Kenya using a structural determinants and life-course perspective. Baseline cross-sectional survey data were collected June–December 2019 for the Maisha Fiti study with 1003 FSWs (aged 18–45 years). Odds ratios and 95% confidence intervals were estimated using multivariable logistic regression with a hierarchical modelling approach. HIV prevalence was 28.0%, and increased with age (<25 years 5.7%, 25–34 years 19.0%, ≥35 years 40.6%). In adjusted analyses, HIV seroprevalence was associated with childhood and adolescence including violence from militia or soldiers (AOR = 1.60; 95%CI:1.00–2.53), young age at sexual debut (≤15 years old vs. ≥18 years AOR = 0.57; 95%CI:0.39–0.84) and teenage pregnancy (AOR = 1.37; 95%CI:1.00–1.88). For adulthood the factors included lower SES score (lowest vs. highest tertile AOR = 0.63; 95%CI:0.40–0.98); reduced housing insecurity (AOR = 0.52; 95%CI:0.54–0.79); lower alcohol/drug use score (AOR = 0.44; 95%CI:0.31–0.61); and a longer duration of selling sex (0–5 years vs. ≥11 years AOR = 2.35; 95%CI:1.44–3.82). Among HIV-negative FSWs, prevalence of HIV risk factors was high (recent hunger 32.3%; internalised 67.7% and experienced 66.0% sex work stigma; recent police arrest 30.1%; recent physical or sexual violence 65.6%, condomless last sex intimate partner 71.1%; harmful alcohol or substance use 49.1%). Only 24.6% of HIV-negative FSWs reported taking PrEP. Taken together, adverse events in childhood and adolescence were associated with increased odds of living with HIV, and were more strongly associated with HIV serostatus than adulthood structural or behavioural risk factors. HIV-negative FSWs remain at high risk of HIV acquisition. This study highlights the importance of addressing adverse events throughout the life course to reduce HIV risk, and the need to continue multi-level HIV prevention and treatment efforts.


In 2021, it was estimated that 38.4 million adults and children were living with HIV globally, with 20.6 million of these living in East and Southern Africa [1,2]. In 2021, almost half (44.6%) of new HIV infections globally were in people living in those regions [2]. The majority of countries in sub-Saharan Africa (SSA) have seen a decline in both HIV prevalence and incidence among the general population during the last two decades, although this progress may be slowing [1,3]. Female sex workers (FSWs) are among the most at-risk populations for HIV globally, with an estimated prevalence of 10.4% (95% CI 9·5–11·5) globally, and 33.3% (95% CI 29·2–37·6) in East and Southern Africa [4]. In 2019, FSWs had 30 times the risk of acquiring HIV compared with the general female population [5]. In Kenya, from where the data for this paper are drawn, HIV prevalence in 2021 was estimated as 29.3% among FSWs compared with 4.3% in the general population [6]. Unlike the general population, HIV prevalence among FSWs over the past decade has remained stable across most settings globally [4,7,8], with FSWs not equally benefitting from efforts to increase HIV service coverage [9]. FSWs who have sought HIV prevention and treatment services have reported violence, police harassment and discrimination from healthcare providers, all of which can deter them from accessing the care they require [10].

The high prevalence of HIV among FSWs in SSA is heavily influenced by structural determinants [1013], defined as factors that are external to the individual and operate outside the locus of their control [14]. For HIV, known structural determinants among FSWs have been conceptualised by Shannon et al. [14] as operating at the (i) macrostructural (e.g. social, economic and health policies, and laws governing sex work, migration, stigma, cultural norms on gender and sexuality), (ii) community organisation of sex work (e.g. sex work collectivisation and empowerment) and (iii) work environment levels (e.g. venue based characteristics, managerial practices, local policing, coverage and access of condoms, HIV/STI testing and anti-retroviral therapy (ART)). These structural determinants can dynamically interact and influence inter-personal (e.g. partner-level/dyad-level risks and protections, such as condom negotiation, sexual networks and patterning) and individual factors of sex workers, clients and their non-paying intimate partners including (i) behavioural (e.g. drug use, duration in sex work), (ii) biological (e.g. age, sex, race, HIV characteristics, STI co-infection) and (iii) host genotypic factors (e.g. host immunity). Together these work to shape HIV acquisition and transmission dynamics and epidemic trajectories at the individual and population levels [14]. In particular, the criminalisation of sex work in 116 countries globally not only undermines HIV prevention, treatment, care and support efforts [12,13,1518] but also means that FSWs have little recourse to report violence or abuse, including police violence and extortion [7,19,20]. However few studies among FSWs in SSA have included a structural determinants framework when examining associations with HIV [21]. In addition, life course theory—which takes into account significant events over someone’s entire life, including during childhood and adolescence, and which explicitly states the temporal ordering of exposure variables and their inter-relationships with the outcome measure–has rarely been applied to HIV studies among FSWs in SSA [22]. Instead, studies have tended to focus on individual and inter-personal behavioural and biological risks such as host immunity, condom use, hormonal contraceptive use and client volume. This paper helps address this evidence-gap by examining HIV risk through a structural determinants and a life-course lens.

In 2010, the Kenya National AIDS and STI Control Programme (NASCOP) developed a set of National Guidelines in response to the Kenya National HIV Strategic Plan (KNASP III), which identified FSWs as a key at-risk group for HIV [8]. Kenya’s current National HIV Programme for FSWs is a combination prevention approach that includes not only bio-medical and behavioural, but also structural interventions, and follows global guidance for programming for key populations [23]. This includes the prioritization of FSWs for PrEP and ART programming [3], although retention on daily oral PrEP has been challenging. Encouragingly, estimates suggest that HIV incidence has been decreasing among some groups of FSWs in Kenya [6,2426]. In Nairobi county, HIV prevalence in new FSW clinic enrolees peaked at 81% in 1986, but has been consistently below 50% since 1997 [24]. Approximately 39,600 women are estimated to sell sex, and HIV prevalence is estimated as 29.5% [27,28]. A 2015 HIV incidence study by McKinnon and colleagues [29] found that HIV acquisition was associated with shorter baseline duration of sex work, minimum charge/sex act, Neisseria gonorrhoea infection, sex with casual clients during menses, DepoProvera use, and unprotected regular partner contacts. As this was an incidence study using clinic-level data, structural determinants and childhood and adolescent experiences were not examined.

Maisha Fiti is a longitudinal, mixed-methods study which aims to examine the impact of violence experiences and harmful alcohol use on systemic and genital inflammation. In 2019, we conducted a baseline behavioural-biological survey among a representative sample of 1003 FSWs in Nairobi county. We collected structural, inter-personal and individual-level data including on childhood, adolescent, and adult events. The aim of the current study is to use the baseline cross-sectional survey data to understand the epidemiology of HIV among FSWs in Nairobi, Kenya, using a structural determinants and life-course perspective. Specifically, we examine (i) structural, inter-personal (sex work) and individual risk factors across the life-course associated with HIV serostatus, and (ii) the prevalence of HIV risk behaviours among HIV-negative participants.

Materials and methods

Study design

The Maisha Fiti study was designed in consultation with the FSW community in Nairobi, and peer educators and staff working at seven Sex Work Outreach Programme (SWOP) clinics. The data analysed in this paper are from the baseline behavioural-biological surveys, conducted June to December 2019.

Subjects and recruitment

SWOP clinics are health clinics specifically for FSWs across Nairobi county. Seven clinics serve approximately 29,000 (73%) of an estimated 39,000 sex workers, providing an accessible and free comprehensive package of HIV prevention and treatment services. Additional programmes provide services to other FSWs. Eligible women were aged 18–45 years, had attended one of the SWOP clinics in Nairobi in the past 12 months, were not pregnant or breast-feeding and did not have an underlying chronic illness (other than HIV) which could affect their immunology (diabetes, rheumatoid arthritis, asthma, TB infection, recent chirotherapy (to treat cervical cell abnormalities)). Women aged <25 were over-sampled to enable stratified analyses by age-group.

Sample size calculations have been described previously [30]. Eligibility of the 29,000 FSWs enrolled at a SWOP clinic was initially assessed using anonymized SWOP patient data. Of 10,409 eligible women, 1200 were randomly selected with probability weighted to clinic client volume, so that the sample was self-weighting. Selected women were contacted by telephone, told about the study, and those interested in participating were scheduled for a screening appointment at the dedicated study clinic in downtown Nairobi. At the screening visit, eligible women received detailed information about the study both verbally and through a written participant information leaflet. Consenting women undertook a pregnancy test, and those who were not pregnant or breast-feeding were enrolled in the study and completed a behavioural-biological survey.

Behavioural-biological survey

All interviews were conducted in Swahili or English by a trained research assistant or clinician. The behavioural questionnaire took approximately 1–1.5 hours to complete and contained sections on: childhood experiences, socio-demographic factors, sexual and reproductive health factors, violence, stigma, sex work and behavioural characteristics. Biological samples comprised: a urine to test for pregnancy (exclusion criteria) and for Neisseria gonorrhoeae (NG) and Chlamydia trachomatis (CT) using GeneXpert Assay. HIV status was screened by rapid HIV tests, with positive tests confirmed using HIV DNA Genexpert. Blood samples were collected to confirm new HIV diagnoses and to assess Treponema pallidum (syphilis) infection using the plasma reagin assay. Genital swabs were used to test for Trichomonas vaginalis (TV; OSOM Trichomonas Rapid Test; SEKISUI Diagnostics, LLC) and Bacterial vaginosis (BV; Gram’s stain and Nugent scoring). Given that the questionnaire contained sensitive questions, the study team underwent an intensive three-week training prior to the study start, which included pilot testing and refining the questionnaire and developing clear referral pathways for women who required additional support. Participants who reported recent violence, mental health problems or suicidal behaviours were referred to a trained counsellor employed as part of the study team. Women reporting ongoing violence experiences were also linked to the violence response team at the SWOP clinics. All women who tested positive for HIV were counselled and referred for HIV care at their chosen SWOP clinic. Women who tested positive for STIs were offered treatment free of charge. Participants were reimbursed for their time. Questionnaires were completed on paper, and double-entered using CS-PRO.


The main outcome variable of interest was HIV serostatus (defined using the laboratory tests described above).

To examine risk factors across the life-course associated with HIV status (objective 1), we developed a conceptual framework, drawing on current theories about risk factors for HIV among FSWs, using a structural determinants and life-course perspective (Fig 1). Thus we diagrammatically ordered classes of variables across the life course, prior to statistical analysis. We conceptualised risk factors as operating across the life course (childhood, adolescence and adulthood) and at a variety of hierarchical levels (structural, inter-personal (sex work) and individual) [14,3133]. Variables could be both ‘childhood’ and ‘structural’ e.g. education level; forced sexual debut, as well as ‘adulthood’ and ‘structural’ e.g. sex work stigma, police arrest. Exposure variables are shown in Fig 1, with further details on variable measurement provided in Table 1.

Fig 1. Conceptual hierarchical framework of risk factors across the life course for current HIV status among FSWs in Nairobi, Kenya.

Table 1. Definition of exposure variables used in the Maisha Fiti study with FSWs in Nairobi, Kenya.

To examine the prevalence of HIV risk behaviours among HIV-negative participants (objective 2), we developed a conceptual framework based on the literature, and drawing on life-course and structural determinants theory, and grouped exposure variables into three domains (Fig 2). Current risk factors for HIV infection were diagrammatically ordered as operating at the structural, sex work and individual (transmission co-factors) levels. Five intravaginal washing practices were ascertained (yes/no) using standardised questions [39].

Fig 2. Hierarchical conceptual framework to examine the potential effect and relationships between age and HIV infection among HIV negative FSWs in Nairobi, Kenya.

Statistical analysis

Data were double entered and statistical analyses were conducted in STATA version 16.1. (Stata Inc., College Station, TX, USA). Data were weighted to allow for over-sampling of participants aged <25 years old. To adjust for clustering by clinic, SWOP clinic was included as a fixed effect in all models. Descriptive statistics and the Wald test were used to describe the study cohort and HIV prevalence overall. Associations with HIV serostatus (objective 1) were estimated using odds ratios (OR) and 95%CI, with p-values obtained using a joint hypothesis test via the adjusted Wald test (to allow for sampling weights). Informed by the structural determinants and life-course theories, we categorised variables into three hierarchical levels (i) childhood factors (ii) adolescent sexual and reproductive and health factors and (iii) adulthood risk factors. We examined associations of these with HIV status using logistic regression. Age was included in all models as an a-priori confounder of other exposures of interest. We built models starting with variables earliest in the life-course and worked forward time-wise, so that we first estimated the direct effects for childhood and adolescent variables on HIV serostatus, independent of any mediating effect through more recent adulthood variables [40]. We then examined the direct effects of more recent adulthood variables on the outcome, after adjusting for upstream confounders. Thus, following the conceptual framework (Fig 1), we built multivariable models for the outcome (HIV serostatus) as follows: In model 1, the overall associations of HIV with level 1 variables (childhood factors) were examined and a core group adjusted for each other were retained if associated with HIV status (p-value <0.1). In model 2, level 2 variables (adolescent sexual and reproductive health factors) were examined adjusted for the core level 1 variables, age and clinic, and retained if independently associated with HIV status (p-value <0.1). Similarly, in models 3a-3c, level 3a-3c variables (adulthood individual, structural, and sex work (inter-personal) variables, respectively) were adjusted for the core level 1 and 2 variables and retained if independently associated with HIV status (p-value <0.1). Missing data were reported if >5% of observations were missing.

To examine the prevalence of risk behaviours among HIV-negative participants (objective 2), we conducted weighted descriptive analyses informed by our conceptual framework (Fig 2), overall and stratified by age. We stratified by age as we were a-priori interested in understanding if risk behaviours differed by age. P-values were obtained using a joint hypothesis test via the adjusted Wald test.

Ethics statement

All participants provided written formal consent. The Maisha Fiti study was approved by the Kenyatta National Hospital Ethics and Research Committee (KNH ERC P778/11/2018), the London School of Hygiene and Tropical Medicine (LSHTM) Ethics Committee (Approval number: 16229) and the University of Toronto ethics committee (Approval number: 37046).


Participant characteristics and HIV and STI status, overall and by age

Of 1200 women randomly sampled from 7 SWOP clinic sites in Nairobi, 1039 (86.6%) were eligible for the study, of whom 1003 (96.5%) were enrolled. Participant characteristics are detailed in Table 2. Almost all participants were born in Kenya (98.7%) and most were Protestant (54.4%) or Catholic (36.8%). Overall, 28.4% had an ACE score of 0–4 (low), 54.3% had an ACE score of 5–8 (moderate) and 17.3% had an ACE score of 9–12 (high), with little difference by age group (p-value = 0.767). Around two-thirds of participants (69.2%) had primary education or less, with increasing education levels among younger age groups (p-value <0.001) (Table 2). Most participants (81.3%) had previously been married, and lived alone or with children (83.0%); 59.8% said that they currently had an intimate partner.

HIV prevalence was 28.0%, and increased with age from 5.7% among those aged <25 years to 40.6% among those aged ≥35 years (p-value <0.001) (Table 2). Prevalence of bacterial STIs was relatively low: syphilis (2.1%), gonorrhoea (2.6%), chlamydia (7.6%), and was higher in younger compared with older women (p-value = 0.011). BV prevalence was 20.0%, and trichomonas prevalence was 3.0%, with no evidence of a difference by age group (p-value = 0.928; p-value = 0.448, respectively).

Associations with HIV status

We first examined risk factors associated with HIV serostatus, using a structural determinants and life-course perspective (objective 1) (Fig 1, Table 3). In adjusted analyses there was strong evidence that HIV serostatus was independently associated with increasing age (25–34 years vs. <25 years; AOR = 3.51; 95%CI:1.86–6.63; ≥35 years vs. <25 years: AOR = 9.77; 95%CI:5.23–18.27), and experiencing violence from soldiers in childhood (AOR = 1.60; 95%CI:1.00–2.53) (Model 1). There was weak evidence that low education level (completed secondary school or higher vs. did not complete primary school (AOR = 0.64; 95%CI:0.40–1.02)), was also associated with increased odds of HIV infection (Model 1). In model 2, HIV serostatus was independently associated with the following adolescent sexual and reproductive health risk factors: early age of sexual debut (age of sexual debut ≥18 years vs. ≤15 years AOR = 0.57; 95%CI:0.39–0.84) and early age at first pregnancy (<18 years of age vs ≥18: AOR = 1.37; 95%CI:1.00–1.88). When we examined associations between HIV serostatus and more recent adulthood risk factors, after adjusting for factors in models 1 and 2, in Model 3a, HIV serostatus was associated with being in the lowest compared with the highest wealth tertile (AOR = 0.63; 95%CI:0.40–0.98; test for trend; p = 0.041) and having low compared with medium/high levels of alcohol or drug use (AOR = 0.44: 95%CI:0.31–0.61). In model 3b, having fewer compared with ≥2 house moves in the past year (AOR = 0.52; 95%CI: 0.34–0.79) was strongly associated with HIV serostatus. In model 3c, we found evidence that increasing duration in sex work was associated with HIV serostatus (test for trend: p = 0.001) with HIV serostatus strongly associated with ≥11 years of sex work compared with selling sex for 0–5 years (AOR = 2.12; 95%CI: 1.31–3.44).

Table 3. Multivariable logistic regression–associations with HIV status.

HIV risk factors among HIV negative women

We next described the prevalence of current HIV risk behaviours among HIV-negative participants, using a structural determinants and life-course approach (objective 2) (Fig 2, Table 4). When we considered adulthood structural factors, one third (32.3%) of HIV negative participants reported food insecurity in the past week and one fifth (21.9%) reported housing insecurity in the past year. One third had more than 4 people dependent on their income, and just over half (55.5%) had no additional income to sex work. Both the number of dependents and recent food insecurity increased with increasing age (p-value = 0.001; p-value = 0.027, respectively). Two thirds of women reported internalised (67.7%) and experienced (66.0%) sex work stigma in the past 12 months, 30.0% reported recent police arrest and 6.9% reported having had sex with the police to avoid arrest in the past 6 months. Nearly two-thirds of HIV-negative participants (65.6%) reported physical or sexual violence in the past 6 months, with physical or sexual violence most common among women aged 25–34 years (69.9%) (p-value = 0.067). A substantial minority (6.3%) reported being raped in the past 7 days.

Table 4. HIV risk factors among HIV negative female sex workers in Nairobi, Kenya.

In terms of adulthood sex work factors, 41.6% reported having had 5 or more clients in the previous week and 29.3% reported soliciting clients from the streets, with street-based sex work most likely among women aged 25–34 years old (p-value = 0.088). Reported condom use at last sex with a client was 83.2%, but condom use with intimate partners was lower (28.9%). Anal sex with clients was reported by a small minority (4.1%). When we examined HIV transmission co-factors, 13.5% of HIV negative women had at least one STI and 17.1% had BV. The prevalence of STIs was highest among the <25 year old participants (p-value = 0.005). One fifth (20.0%) of participants reported current depo-provera use, with the highest prevalence among <25 year olds (27.5%) (p-value = 0.009). Using substances to clean or dry the vagina was common, with 37.9% of women reporting having used soap, antiseptic, lemon, cloth to clean or inserted drying substance into vagina in the past 30 days. Younger women were more likely to report intravaginal washing than older women (48.0% among <25 year olds vs 36.2% among ≥35 years olds) (p-value = 0.003). One-quarter (24.6%) reported currently taking PrEP, with little difference by age group (p-value = 0.378). Almost half of participants (49.1%) had an alcohol or substance use problem, and prevalence was highest among those aged <25 years (p-value = 0.022).


In this study among Kenyan FSWs, we found that key events in childhood and adolescence, such as violence from soldiers, early age at first sex and early age at first pregnancy, were associated with HIV serostatus, with weak evidence that not completing secondary school may also be a risk factor. These events were more closely related to HIV risk than more recent structural risk factors such as food insecurity, sex work stigma, police arrest, violence and gang-rape. The inverse association with alcohol and drug use problems has previously been reported by this team and is likely due to the intensive counselling that FSWs living with HIV receive by the SWOP clinic teams, due to the negative impacts of alcohol and drug use on anti-retroviral compliance and efficacy [41].

A review by Shannon et al. of 87 FSW studies which explicitly examined structural determinants of HIV noted that few such studies have been published in SSA [21]; our study adds to this literature. Education is one structural factor included in several SSA FSW studies; similar to here, studies with FSWs in Zimbabwe, Cameroon, South Africa, Nigeria, Mombasa, and Uganda, found that secondary school education or higher may be protective, although no associations between education level and HIV were found among FSWs in Lesotho, The Gambia or Rwanda [4250]. Historical gender based violence (physical or sexual) was examined in our study, as well as with FSWs in Zimbabwe and Rwanda, with positive associations found between forced sex and HIV in Rwanda only [43,48]. Sex-work related stigma was examined in our study, and in Cameroon and Lesotho, with positive associations with enacted sex work stigma seen in Lesotho only [42,44]. At the work environment level, physical or sexual violence by clients, police, managers or pimps was examined in our study, as well as in Cameroon, The Gambia and Zimbabwe, with no associations with HIV seen in any of these settings [42,43,45]. However, police arrest or incarceration (ever) was positively associated with HIV in Cameroon and Rwanda (but not in our study) [42,51]. In addition, to our knowledge, ours is the first study from SSA to examine HIV risk among FSWs from a life-course perspective. Previous studies from South Africa and Malawi among young adults from the general population have reported associations between a higher ACE or childhood trauma score and HIV risk behaviours [52,53]. We did not find associations between HIV and overall ACE score [54], although there was evidence that individual items (violence from soldiers or militia during childhood) were associated with HIV serostatus. Taken together, these findings highlight the importance of supporting adolescents and young women who experience key life events such as war, young age at sexual debut and teenage pregnancy to help reduce their risk of future HIV infection.

We also found that HIV negative sex workers reported high levels of exposure to recent structural HIV risk factors (food and housing insecurity, having two or more other people dependent on their income, internalised and experienced sex work stigma, recent police arrest, recent physical or sexual violence experience). We also found high levels of exposure to sex work risk factors (condom use particularly with regular partners) and HIV transmission co-factors (testing positive for an STI and/or bacterial vaginosis; PrEP use; harmful alcohol or drug use). Taken together, these findings suggest that many HIV negative FSWs in this setting remain at risk of HIV acquisition. The study with young FSWs in Zimbabwe reported similar concerns [43].

Since 2003, the Kenyan government, supported by overseas donors have implemented targeted HIV prevention interventions both for the general population, and for key populations including FSWs [17]. From 2008, these have included the widespread implementation of HIV testing and free ART services, the provision of FSW-specific clinics and outreach, and some structural interventions, including FSW-led community mobilisation and violence response interventions [26]. Similar to other settings, the HIV prevalence among the general population has fallen dramatically in Kenya over the past 25 years from 10.5% in 1995–1996 to 4.0% in 2021, with a 53% reduction in new HIV infections between 2010 and 2021 [55,56]. In addition, the prevalence among clients of FSWs is also estimated to have fallen across SSA from 10% prior to 2001 to 3% from 2011 onwards [57], although data is limited for Kenya. Contrary to many other FSW settings globally where HIV prevalence has remained stable [4,7], the HIV prevalence among FSWs in Kenya and in Nairobi has declined over the past two decades [2426,58]. This is likely due to sustained prevention efforts by the Government of Kenya, international donors such as the US President’s Emergency fund for AIDS Relief (PEPFAR) and the Global Fund to Fight AIDS, Tuberculosis and Malaria (GFATM). The population-level scale-up and roll-out of ART for people living with HIV, and the work of specialised sex work services and programmes implemented by more than 100 partners across Kenya will also have been important [3,8,17]. In addition, the heterogeneity in host HIV susceptibility may have contributed to this decline [59]. However, our findings, together with the literature suggest that we cannot become complacent; in order to prevent steep increases in new HIV infections with increasing age, multi-level FSW HIV intervention efforts need to continue. In addition, our findings suggest that interventions which address harmful alcohol and drug use as part of HIV prevention planning, as well as the decriminalisation of sex work, are also needed [12,17].

A key limitation of our study was the cross-sectional study design, which limits our ability to ascertain the direction of causality. This is particularly important when examining recent determinants of HIV risk, where HIV infection is likely to have preceded factors such as recent housing or food insecurity and current STI prevalence. Cohort studies are needed to determine causality but these can be challenging to conduct with hidden and hard-to-reach populations such as FSWs. In addition, large sample sizes and/or a long follow-up will be needed in most settings; while the Maisha Fiti study is a longitudinal study, the sample size is too small to be able to conduct HIV incidence analyses. Although our sample was random, it was drawn from FSWs registered at SWOP clinics across Nairobi county; we may have missed the most vulnerable FSWs who are not registered at a SWOP clinic, such as young or new FSWs. This may have led to an under-estimate of key HIV risk factors such as STI prevalence, violence and harmful alcohol/drug use. Many of our questions were sensitive, including those on ACEs, violence and mental health and this could have led to an underestimate of key exposure variables. The study team underwent three weeks of intensive training and were supported by weekly de-brief meetings throughout data collection. In addition, the study team included 10 FSWs employed as outreach workers at the SWOP clinics and as beauticians and receptionists at the study clinic. The SWOP clinics were established in 2010 and have built trust with the FSW community, especially vis a vis confidentiality. Together these should have helped reduce interviewer and reporting bias. Key strengths of this study was the large sample size and the random selection of FSWs from across Nairobi county. The comprehensive behavioural-biological survey data meant we were able to examine several variables across the life-course in our analyses.


This study adds to newly emerging literature on the importance of childhood and adolescent experiences as well as structural factors on shaping the HIV risk environment for high risk groups such as FSWs. We also demonstrate that despite national and county-level targeted HIV intervention programming for key populations and considerable declines in HIV prevalence over time, many HIV negative FSWs in this setting remain at high risk of HIV infection. Together this calls for the continuation of targeted multi-level FSW programming, which addresses structural as well as behavioural and biological determinants of risk. ‘Upstream’ HIV prevention programmes which can identify and specifically work with adolescents and young adults who have experienced adverse life events such as violence from soldiers or militia, early sexual debut and/or teen pregnancy are also needed.


The authors thank the women in Nairobi who participated in this study. The Maisha Fiti Study Champions were: Demtilla Gwala, Daisy Oside, Ruth Kamene, Agnes Watata, Agnes Atieno, Faith Njau, Elizabeth Njeri, Evelyn Orobi and Ibrahim Lwingi.


  1. 1. UNAIDS. IN DANGER: UNAIDS Global AIDS Update 2022. 2022. Geneva: Joint United Nations Programme on HIV/AIDS. Accessed 7 October 2022.
  2. 2. UNAIDS. Fact Sheet 2022. 2022. Geneva: UNAIDS, Accessed 17 October 2022.
  3. 3. LeVasseur MT, Goldstein ND, Welles SL. A public health perspective on HIV/AIDS in Africa: Victories and unmet challenges. Pathophysiology. 2014;21(3):237–56. Epub 2014/08/07. pmid:25096828.
  4. 4. Shannon K, Crago AL, Baral SD, Bekker LG, Kerrigan D, Decker MR, et al. The global response and unmet actions for HIV and sex workers. Lancet. 2018;392(10148):698–710. Epub 2018/07/25. pmid:30037733; PubMed Central PMCID: PMC6384122.
  5. 5. UNAIDS. HIV and sex work. 2021. Geneva: UNAIDS. Accessed 6th October 2022.
  6. 6. Kenya Ministry of Health. Kenya World AIDS Day Progress Report 2013–2021 Nairobi, Kenya 2021 [ Accessed 02 Oct 2023].
  7. 7. Viswasam N, Rivera J, Comins C, Rao A, Lyons CE, Baral S. The Epidemiology of HIV Among Sex Workers Around the World: Implications for Research, Programmes, and Policy. In: Goldenberg SM, Morgan Thomas R, Forbes A, Baral S, editors. Sex Work, Health, and Human Rights: Global Inequities, Challenges, and Opportunities for Action. Cham: Springer International Publishing; 2021. p. 15–39.
  8. 8. Baral S, Beyrer C, Muessig K, Poteat T, Wirtz AL, Decker MR, et al. Burden of HIV among female sex workers in low-income and middle-income countries: a systematic review and meta-analysis. Lancet Infect Dis. 2012. Epub 2012/03/20. pmid:22424777.
  9. 9. Lancaster KE, Cernigliaro D, Zulliger R, Fleming PF. HIV care and treatment experiences among female sex workers living with HIV in sub-Saharan Africa: A systematic review. Afr J AIDS Res. 2016;15(4):377–86. Epub 2016/12/16. pmid:27974017; PubMed Central PMCID: PMC5541376.
  10. 10. Jin H, Restar A, Beyrer C. Overview of the epidemiological conditions of HIV among key populations in Africa. Journal of the International AIDS Society. 2021;24 Suppl 3(Suppl 3):e25716. Epub 2021/07/01. pmid:34190412; PubMed Central PMCID: PMC8242974.
  11. 11. Stangl AL, Pliakas T, Izazola-Licea JA, Ayala G, Beattie TS, Ferguson L, et al. Removing the societal and legal impediments to the HIV response: An evidence-based framework for 2025 and beyond. PloS one. 2022;17(2):e0264249. Epub 2022/02/23. pmid:35192663; PubMed Central PMCID: PMC8863250.
  12. 12. Lyons CE, Schwartz SR, Murray SM, Shannon K, Diouf D, Mothopeng T, et al. The role of sex work laws and stigmas in increasing HIV risks among sex workers. Nat Commun. 2020;11(1):773. Epub 2020/02/20. pmid:32071298; PubMed Central PMCID: PMC7028952.
  13. 13. Ngugi EN, Roth E, Mastin T, Nderitu MG, Yasmin S. Female sex workers in Africa: epidemiology overview, data gaps, ways forward. SAHARA J: journal of Social Aspects of HIV/AIDS Research Alliance / SAHARA, Human Sciences Research Council. 2012;9(3):148–53. Epub 2012/12/15. pmid:23237069; PubMed Central PMCID: PMC4560463.
  14. 14. Shannon K, Goldenberg SM, Deering KN, Strathdee SA. HIV infection among female sex workers in concentrated and high prevalence epidemics: why a structural determinants framework is needed. Current opinion in HIV and AIDS. 2014;9(2):174–82. Epub 2014/01/28. pmid:24464089; PubMed Central PMCID: PMC4286412.
  15. 15. UNAIDS. HIV Criminalization. 2021. Geneva: UNAIDS. Accessed 6th October 2022.
  16. 16. Glick JL, Russo RG, Huang AK, Jivapong B, Ramasamy V, Rosman LM, et al. ART uptake and adherence among female sex workers (FSW) globally: A scoping review. Glob Public Health. 2022;17(2):254–84. Epub 2020/12/11. pmid:33301704; PubMed Central PMCID: PMC8190161.
  17. 17. Musyoki H, Bhattacharjee P, Sabin K, Ngoksin E, Wheeler T, Dallabetta G. A decade and beyond: learnings from HIV programming with underserved and marginalized key populations in Kenya. Journal of the International AIDS Society. 2021;24 Suppl 3(Suppl 3):e25729. Epub 2021/07/01. pmid:34189847; PubMed Central PMCID: PMC8242977.
  18. 18. UNAIDS. The GAP Report. Geneva, Switzerland: Accessed 19 May 2023, 2014.
  19. 19. Deering KN, Amin A, Shoveller J, Nesbitt A, Garcia-Moreno C, Duff P, et al. A systematic review of the correlates of violence against sex workers. Am J Public Health. 2014;104(5):e42–54. Epub 2014/03/15. pmid:24625169; PubMed Central PMCID: PMC3987574.
  20. 20. Decker MR, Wirtz AL, Pretorius C, Sherman SG, Sweat MD, Baral SD, et al. Estimating the impact of reducing violence against female sex workers on HIV epidemics in Kenya and Ukraine: a policy modeling exercise. Am J Reprod Immunol. 2013;69 Suppl 1:122–32. Epub 2013/02/15. pmid:23387931.
  21. 21. Shannon K, Strathdee SA, Goldenberg SM, Duff P, Mwangi P, Rusakova M, et al. Global epidemiology of HIV among female sex workers: influence of structural determinants. Lancet. 2015;385(9962):55–71. Epub 2014/07/26. pmid:25059947; PubMed Central PMCID: PMC4297548.
  22. 22. Ben-Shlomo Y, Kuh D. A life course approach to chronic disease epidemiology: conceptual models, empirical challenges and interdisciplinary perspectives. International Journal of Epidemiology. 2002;31(2):285–93. pmid:11980781
  23. 23. Musyoki H, Bhattacharjee P, Blanchard AK, Kioko J, Kaosa S, Anthony J, et al. Changes in HIV prevention programme outcomes among key populations in Kenya: Data from periodic surveys. PloS one. 2018;13(9):e0203784. Epub 2018/09/20. pmid:30231072; PubMed Central PMCID: PMC6145580.
  24. 24. Kimani J, Kaul R, Nagelkerke NJ, Luo M, MacDonald KS, Ngugi E, et al. Reduced rates of HIV acquisition during unprotected sex by Kenyan female sex workers predating population declines in HIV prevalence. Aids. 2008;22(1):131–7. Epub 2007/12/20. pmid:18090401.
  25. 25. Manguro GO, Gichuki C, Ampt FH, Agius PA, Lim MS, Jaoko WG, et al. HIV infections among female sex workers in Mombasa, Kenya: current prevalence and trends over 25 years. Int J STD AIDS. 2020;31(14):1389–97. Epub 2020/10/27. pmid:33103582.
  26. 26. Tago A, McKinnon LR, Wanjiru T, Muriuki F, Munyao J, Gakii G, et al. Declines in HIV prevalence in female sex workers accessing an HIV treatment and prevention programme in Nairobi, Kenya over a 10-year period. Aids. 2021;35(2):317–24. Epub 2020/11/10. pmid:33165030.
  27. 27. Musyoki H, Kellogg TA, Geibel S, Muraguri N, Okal J, Tun W, et al. Prevalence of HIV, sexually transmitted infections, and risk behaviours among female sex workers in Nairobi, Kenya: results of a respondent driven sampling study. AIDS Behav. 2015;19 Suppl 1(Suppl 1):S46–58. Epub 2014/11/28. pmid:25428282; PubMed Central PMCID: PMC4786175.
  28. 28. National AIDS and STI Control Programme. Key Population Mapping and Size Estimation in Selected Counties in Kenya: Phase 1. 2019. Nairobi, Kenya: Accessed 21 November 2022.
  29. 29. McKinnon LR, Izulla P, Nagelkerke N, Munyao J, Wanjiru T, Shaw SY, et al. Risk Factors for HIV Acquisition in a Prospective Nairobi-Based Female Sex Worker Cohort. AIDS Behav. 2015;19(12):2204–13. Epub 2015/06/21. pmid:26091706.
  30. 30. Beksinska A, Jama Z, Kabuti R, Kungu M, Babu H, Nyariki E, et al. Prevalence and correlates of common mental health problems and recent suicidal thoughts and behaviours among female sex workers in Nairobi, Kenya. BMC PSYCHIATRY. 2021;21(1). WOS:000707340000001. pmid:34649544
  31. 31. Ben-Shlomo Y, Kuh D. A life course approach to chronic disease epidemiology: conceptual models, empirical challenges and interdisciplinary perspectives. Int J Epidemiol. 2002;31(2):285–93. Epub 2002/05/01. pmid:11980781.
  32. 32. El-Bassel N, Mukherjee TI, Stoicescu C, Starbird LE, Stockman JK, Frye V, et al. Intertwined epidemics: progress, gaps, and opportunities to address intimate partner violence and HIV among key populations of women. Lancet HIV. 2022;9(3):e202–e13. Epub 2022/02/14. pmid:35151376.
  33. 33. Heise LL. Violence against women: an integrated, ecological framework. Violence Against Women. 1998;4(3):262–90. Epub 2002/09/26. pmid:12296014.
  34. 34. World Health Organisation. Adverse Childhood Experiences International Questionnaire (ACE-IQ). 2020. Geneva: Accessed 31st October 2022.
  35. 35. Beksinska A, Jama Z, Kabuti R, Kungu M, Babu H, Nyariki E, et al. Prevalence and correlates of common mental health problems and recent suicidal thoughts and behaviours among female sex workers in Nairobi, Kenya. BMC Psychiatry. 2021;21(1):503. Epub 2021/10/16. pmid:34649544; PubMed Central PMCID: PMC8518166.
  36. 36. Kenya National Bureau of Statistics. The 2014 Kenya Demographic and Health Survey. 2015. Kenya: Accessed 30 June 2022.
  37. 37. Garcia-Moreno C, Jansen HA, Ellsberg M, Heise L, Watts CH. Prevalence of intimate partner violence: findings from the WHO multi-country study on women’s health and domestic violence. Lancet. 2006;368(9543):1260–9. pmid:17027732.
  38. 38. World Health Organisation. The Alcohol, Smoking and Substance Involvement Screening Test (ASSIST): manual for use in primary care. 2010. France: Accessed 12 July 2022.
  39. 39. Low N, Chersich MF, Schmidlin K, Egger M, Francis SC, van de Wijgert JH, et al. Intravaginal practices, bacterial vaginosis, and HIV infection in women: individual participant data meta-analysis. PLoS Med. 2011;8(2):e1000416. Epub 2011/03/02. pmid:21358808; PubMed Central PMCID: PMC3039685.
  40. 40. Victora CG, Huttly SR, Fuchs SC, Olinto MT. The role of conceptual frameworks in epidemiological analysis: a hierarchical approach. Int J Epidemiol. 1997;26(1):224–7. Epub 1997/02/01. pmid:9126524.
  41. 41. Beksinska A, Nyariki E, Kabuti R, Kungu M, Babu H, Shah P, et al. Harmful Alcohol and Drug Use Is Associated with Syndemic Risk Factors among Female Sex Workers in Nairobi, Kenya. International journal of environmental research and public health. 2022;19(12). Epub 2022/06/25. pmid:35742558; PubMed Central PMCID: PMC9223659.
  42. 42. Bowring AL, Ketende S, Billong SC, Mfochive Njindam I, Rao A, Decker MR, et al. Characterizing Sociostructural Associations With New HIV Diagnoses Among Female Sex Workers in Cameroon. J Acquir Immune Defic Syndr. 2019;80(3):e64–e73. Epub 2019/02/15. pmid:30762674.
  43. 43. Chabata ST, Hensen B, Chiyaka T, Mushati P, Mtetwa S, Hanisch D, et al. Changes Over Time in HIV Prevalence and Sexual Behaviour Among Young Female Sex-Workers in 14 Sites in Zimbabwe, 2013–2016. AIDS Behav. 2019;23(6):1494–507. Epub 2019/02/23. pmid:30788641; PubMed Central PMCID: PMC6535417.
  44. 44. Moazzami M, Ketende S, Lyons C, Rao A, Taruberekera N, Nkonyana J, et al. Characterizing Multi-level Determinants of HIV Prevalence Among Female Sex Workers in Maseru and Maputsoe, Lesotho. AIDS Behav. 2020;24(3):714–23. Epub 2019/05/02. pmid:31041624; PubMed Central PMCID: PMC6821589.
  45. 45. Peitzmeier S, Mason K, Ceesay N, Diouf D, Drame F, Loum J, et al. A cross-sectional evaluation of the prevalence and associations of HIV among female sex workers in the Gambia. Int J STD AIDS. 2014;25(4):244–52. Epub 2013/08/24. pmid:23970652.
  46. 46. Vandepitte J, Bukenya J, Weiss HA, Nakubulwa S, Francis SC, Hughes P, et al. HIV and other sexually transmitted infections in a cohort of women involved in high-risk sexual behavior in Kampala, Uganda. Sex Transm Dis. 2011;38(4):316–23. Epub 2011/04/01. pmid:23330152; PubMed Central PMCID: PMC3920055.
  47. 47. Auvert B, Marais D, Lissouba P, Zarca K, Ramjee G, Williamson AL. High-risk human papillomavirus is associated with HIV acquisition among South African female sex workers. Infect Dis Obstet Gynecol. 2011;2011:692012. Epub 2011/08/02. pmid:21804752; PubMed Central PMCID: PMC3143430.
  48. 48. Braunstein SL, van de Wijgert JH, Vyankandondera J, Kestelyn E, Ntirushwa J, Nash D. Risk Factor Detection as a Metric of STARHS Performance for HIV Incidence Surveillance Among Female Sex Workers in Kigali, Rwanda. Open AIDS J. 2012;6:112–21. Epub 2012/10/12. pmid:23056162; PubMed Central PMCID: PMC3465816.
  49. 49. Eluwa GI, Strathdee SA, Adebajo SB, Ahonsi B, Azeez A, Anyanti J. Sexual Risk Behaviors and HIV Among Female Sex Workers in Nigeria. JAIDS Journal of Acquired Immune Deficiency Syndromes. 2012;61(4):507–14. pmid:22918155-201212010-00015.
  50. 50. Graham SM, Raboud J, McClelland RS, Jaoko W, Ndinya-Achola J, Mandaliya K, et al. Loss to follow-up as a competing risk in an observational study of HIV-1 incidence. PloS one. 2013;8(3):e59480. Epub 2013/04/05. pmid:23555041.
  51. 51. Braunstein SL, Ingabire CM, Geubbels E, Vyankandondera J, Umulisa MM, Gahiro E, et al. High burden of prevalent and recently acquired HIV among female sex workers and female HIV voluntary testing center clients in Kigali, Rwanda. PloS one. 2011;6(9):e24321. Epub 2011/09/29. pmid:21949704.
  52. 52. Gibbs A, Dunkle K, Washington L, Willan S, Shai N, Jewkes R. Childhood traumas as a risk factor for HIV-risk behaviours amongst young women and men living in urban informal settlements in South Africa: A cross-sectional study. PloS one. 2018;13(4):e0195369. Epub 2018/04/07. pmid:29624612; PubMed Central PMCID: PMC5889178.
  53. 53. VanderEnde K, Chiang L, Mercy J, Shawa M, Hamela J, Maksud N, et al. Adverse Childhood Experiences and HIV Sexual Risk-Taking Behaviors Among Young Adults in Malawi. Journal of interpersonal violence. 2018;33(11):1710–30. Epub 2018/05/10. pmid:29739289; PubMed Central PMCID: PMC6158791.
  54. 54. Hughes K, Bellis MA, Hardcastle KA, Sethi D, Butchart A, Mikton C, et al. The effect of multiple adverse childhood experiences on health: a systematic review and meta-analysis. Lancet Public Health. 2017;2(8):e356–e66. Epub 2017/12/19. pmid:29253477.
  55. 55. UNAIDS. Country factsheets. Kenya. 2021. Geneva: UNAIDS. Accessed 7 Oct 2022.
  56. 56. Kenya National AIDS Control Council NAaSCP. The Kenya AIDS epidemic. Update 2011. 2012. Nairobi, Kenya: NACC and NASCOP. Accessed 7 October 2022.
  57. 57. Wulandari LPL, Guy R, Kaldor J. The burden of HIV infection among men who purchase sex in low- and middle-income countries—a systematic review and meta-analysis. PloS one. 2020;15(9):e0238639. Epub 2020/09/05. pmid:32886695; PubMed Central PMCID: PMC7473528.
  58. 58. Vandenhoudt HM, Langat L, Menten J, Odongo F, Oswago S, Luttah G, et al. Prevalence of HIV and other sexually transmitted infections among female sex workers in Kisumu, Western Kenya, 1997 and 2008. PloS one. 2013;8(1):e54953. Epub 2013/02/02. pmid:23372801; PubMed Central PMCID: PMC3553007.
  59. 59. Nagelkerke N, de Vlas SJ, Jha P, Luo M, Plummer FA, Kaul R. Heterogeneity in host HIV susceptibility as a potential contributor to recent HIV prevalence declines in Africa. Aids. 2009;23(1):125–30. Epub 2008/12/04. pmid:19050394; PubMed Central PMCID: PMC2764323.