Sage Journals: Discover world-class research

Abstract

Background

Mobility challenges HIV prevention efforts through associated risk behaviours and sexually transmitted infections (STIs). We characterized relationships between mobility and sexual risks on STI prevalence over time in East Africa.

Methods

Geospatial mobility and sexual risk behaviours were collected in 12 communities using a sex- and HIV-stratified random sub-sample of 2750 adults from a longitudinal cohort (2015–2019) of a HIV test-and-treat trial in Kenya and Uganda. Annual Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) testing was performed and relationships of prevalent STIs with mobility, sexual concurrency, and higher HIV-risk sexual partners (defined as one night stand, stranger, commercial sex worker/client, casual partner, or inherited partner/inheritor) were examined.

Results

The annual prevalence of CT or NG among 2665 participants tested was 3.1% (95% CI: 2.5–3.9) at baseline, 3.3% (95% CI: 2.6–4.0) at year 1, 4.4% (95% CI: 3.0–5.2) at year 2, and 4.8% (95% CI: 4.0–5.7) at year 3. STI (CT, NG) prevalence was associated with migration in the past year, sexual partnership concurrency, being single, higher HIV-risk partners, age >25, low household wealth, and the relationship between gender and work-related travel in past 6 months. The association between select STI prevalence and past 6-months travel was mediated by higher HIV-risk sexual partners, partnership concurrency, out-of-town partner, and higher HIV-risk transactional sex partners.

Conclusions

Geospatial mobility, sexual concurrency, and higher HIV-risk partnerships significantly influence longitudinal CT and/or NG prevalence in East Africa.

Keywords

STIs mobility sexual risks sexual concurrency gender mediation

Introduction

Human population mobility, inclusive of both changes of residence across geopolitical boundaries (whether temporary or permanent), and complex forms of localized mobility (travel requiring overnight stays away from a primary residence), is common across sub-Saharan Africa (SSA).^1,2,3 Men predominantly travel for work-related reasons,⁴ while women often travel for non-work reasons, such as to care for family members^3,5,6; however, women increasingly participate in labour-related migration.⁵ The HIV epidemic in SSA has been closely linked to population mobility.^1,7,8,9,10 Regions with high levels of mobility tend to be associated with HIV transmission hot-spots.^11,12,13 Mobility links geographically separate epidemics and creates dynamics where some locations import and export HIV risk.^12,14,15 Mobility therefore challenges efforts to prevent and contain HIV transmission.^1,16

Mobility has also been linked with behaviours that increase the risk of HIV acquisition. Mobile persons are more likely to engage in more concurrent sexual relationships,^17,18,19 sex with higher HIV-risk sexual partners (such as commercial sex workers, widowed persons, casual partners),^19,20 transactional sex,^18,21 and age-disparate relationships,²² all of which increases the risk of HIV acquisition.

The primary livelihood in Lake Victoria shoreline communities in Western Kenya is fishing, which requires mobility both to obtain catch and engage in processing, transport and selling; mobility of both men and women is highly prevalent in the area.^{1,17,19,20,21} A local transactional sex economy, in which female fish traders exchange sex with fishermen in order to receive preferential access to fish catch,^6,21,23 has contributed to a high HIV burden as well: HIV prevalence in this population (9.5–19.1%) is about two to four times the national average in Kenya (4.5%)²⁴ and HIV incidence in this area continues to be high (>2.61 per 1000).²⁵

Sub-Saharan Africa has the highest incidence of trichomoniasis (Trichomonas vaginalis), chlamydia (Chlamydia trachomatis, CT), syphilis (Treponema pallidum), and gonorrhoeae (Neisseria gonorrhoea, NG) with 96 million cases reported in 2020.²⁶ Reproductive-age women have high rates of STIs.²⁶ Recent meta-analyses document pooled prevalence rates for STIs among reproductive-aged women in sub-Saharan Africa, including CT (7.8% (95% CI: 5.6–10.6),²⁷ NG (3.28% (95% CI: 2.61–3.94),²⁸ and syphilis among pregnant women in East Africa of 3.2% (95% CI: 2.3–4.2)²⁹ with some syphilis estimates exceeding 7.5% (95% CI: 5.5%, 10.1%).³⁰ Having an STI can increase HIV risk and vice versa³¹; 10% of people living with HIV are co-infected with another STI,^32,33 and the presence of an STI is a good indicator of sexual risk behaviours.

The relationship between population mobility and STI prevalence has been well-characterized in cross-sectional studies,³⁴ in particular among truckers,^35,36 commercial sex workers,^35,37 women who engage in transactional sex,^38,39 and cross-border traders.⁴⁰ These relationships, however, have not been characterized longitudinally.^41,42 In this study we investigated patterns of select STI (CT, NG) acquisition over time and among men and women living in rural communities in Kenya and Uganda, and assessed the relationships between geospatial mobility, and sexual risk behaviour, on select STI prevalence over 3 years. In the context of high mobility, we characterized relationships between forms of mobility, sexual risks, and gender with objective longitudinal STI (CT, NG) point prevalence as the outcome.

Methods

Study design, participants, and setting

Data are from a longitudinal cohort study of mobility (R01MH104132) embedded within a population-based HIV universal testing and treatment (“test and treat”) trial in Kenya and Uganda (the Sustainable East Africa Research in Community Health [SEARCH] study, NCT# 01864603). A multi-level stratified random sampling design (based on region, sex, and baseline mobility status) was used to select the analytic sample of 2750 adults aged 16 years or older from the larger census-enumerated adult population of 12 SEARCH communities. The selection strata based on couple geospatial mobility status were: (1) both couple members are mobile, (2) both not mobile, (3) female mobile and male not mobile, and (4) male mobile and female not mobile. These communities were selected purposively to sample underlying heterogeneity in forms of mobility. The analysis prioritized investigation of forms of of mobility and select STI prevalence over an analysis estimating precise population prevalences. Baseline mobility was defined as being away from the household for 6 months or more in the past year and/or fewer than half of nights spent in the household in the past 4 months.

Data collection

From 2015 to 2019, survey data on sexual risk behaviours and mobility were collected every 6 months via interviewer-administered surveys, with annual CT and NG urine testing using Cepheid GeneXpert (Xpert® CT/NG). Data were collected using detailed classifications and quantifications of geospatial mobility to understand the impact of mobility on sexual HIV risk behaviours, HIV incidence and prevalence, HIV care engagement, and components of the test and treat intervention strategy. Survey data were collected with programmed tablets and took approximately 90 minutes to complete. Topics covered included demographics, sexual behaviours, mobility in the past 6 months, and migration histories.

Measures

Demographics

SEARCH trial socio-demographic data were used (including age, marital status, household wealth), with some data reconfirmed during data collection. The household wealth variable was divided into the lowest quintile compared to all other wealth quintiles.

Geospatial mobility

The baseline survey obtained participants’ migration histories over their lifetime by asking their birthplace and the names of places they had lived (county/district/nation recorded) including their age when they changed residences from childhood to the present. Migration was defined as a change of primary residence across pre-specified geopolitical boundaries (in Kenya, the sub-county, and in Uganda, district; these most closely matched one another in size and population density). Migration within countries was classified as internal migration while migration between countries was classified as international migration.

Participants were also surveyed about geospatial mobility in the 6 months prior to the study visit, including labour-related and non-labour-related mobility that required nights sleeping away from the main residence.¹ Mobility was defined as travel that included time spent away from the primary places of residence, without the intention to change main residence (locations and movements between multiple homes considered main residences were also recorded). This excluded commuting for work or other reasons, as mobility is recorded only if the travel involved sleeping away from the primary residence(s) for one or more nights. Labour-related mobility was defined as travel “for business or to earn money”, including travel to seek employment and for food production/farming. Mobility that was non-labour-related was defined as travel for all other reasons. As described elsewhere by our group,¹ interviewers collected information on where participants travelled, the number of trips taken and the number of nights per trip.

Sexual behaviour

We adapted a calendar-based data collection tool that has previously been used in the region^19,20 to gather relationship and behavioural histories for sexual relationships since January 2011. This tool was based on by a relationship history calendar developed by Luke and colleagues⁴³ that has been shown to reduce social desirability bias and to improve the reporting of sexual behaviour and relationships.^19,20,43 The calendar tool was used to collect data on partnerships, including relationship type as defined in the relationship history calendar (e.g. casual, one night stand, boyfriend/girlfriend, fiancé, spouse), partnership concurrency (i.e. overlapping partnerships within 1-month periods), and mobility of partners in monthly increments. Data is recorded in the survey in monthly intervals because often relationships last for less than a full year. We measured changes in behaviours and relationship type over the course of each sexual relationship in the 5 years before the baseline survey.

Data analysis

Repeated measures mixed-effects logistic models adjusting for clustering by individuals were used to examine bivariate and multivariate relationships of prevalent STIs (CT, NG, or any CT or NG) with recent migration (change of residence over geopolitical boundaries) and local mobility (overnight travel away from home), sexual concurrency (overlapping partnerships), sexual partner residence (local town vs non-local), higher-HIV risk sexual partners (one night stand, stranger, commercial sex worker/client, casual partner, or inherited partner/inheritor), and demographics. To examine potential mediation of the relationship between STI prevalence and measures of mobility by sexual behaviour, we performed a causal mediation analysis using VanderWeele’s approach.⁴⁴ Causal mediation analysis was conducted using the PARAMED command in STATA.⁴⁵ All analyses were conducted with STATA version 17 (College Station, TX).

Ethical approvals

Ethical approvals were received from the Ethical Review Committee of the Kenya Medical Research Institute (KEMRI/SERU/CMR/3052), the Ugandan National Council for Science and Technology (HS 1834), the Makerere University School of Medicine Research and Ethics Committee (2015-040), and the University of California San Francisco Committee on Human Research (14-15058). All study participants provided written informed consent.

Results

Sample demographics

The analytic sample consisted of 2728 persons (99.8%) out of 2750 study participants who had an STI test in any of the four study years, of which 1302 were men and 1426 were women. Only 22 (0.8%) had no testing. Annual STI testing in all 4 years was available for 2344 individuals (85%), with 95% of participants having three or more years of annual STI testing. Overall, the median age was 38 (IQR: 29–50) and 75.3% were married (Table 1). Most participants had some primary or secondary level education (81.3%). About 5% of both men and women were engaged in informal sector, high HIV-risk occupations.

Table 1.

Participant characteristics.

Characteristics at study entry	Overall	Men	Women	p
N (number of participants)	2728	1302	1426	na
Age 10-years age band, n (%)
16–24	402 (14.7%)	170 (13.1%)	232 (16.3%)	0.001
25–34	720 (26.4%)	311 (23.9%)	409 (28.7%)
35–44	690 (25.3%)	350 (26.9%)	340 (23.8%)
45–54	451 (16.55)	234 (18.0%)	217 (15.2%)
55–64	250 (9.2%)	119 (9.1%)	131 (9.2%)
65 and older	215 (7.9%)	118 (9.1%)	97 (6.8%)
Marital status
Divorced, separated, widowed, missing	410 (15.0%)	65 (5.0%)	345 (24.2%)	<0.001
Currently married	2053 (75.3%)	1094 (84.0%)	959 (67.3%)
Currently single	265 (9.7%)	143 (11.0%)	122 (8.5%)
Education level
No schooling	341 (12.5%)	100 (7.7%)	241 (16.9%)	<0.001
Primary/Secondary	2219 (81.3%)	1105 (84.8%)	1114 (78.1%)
Post-secondary	106 (3.9%)	66 (5.1%)	40 (2.8%)
Household wealth: Poorest quantile	63 (2.3%)	32 (2.5%)	31 (2.2%)
Occupation
Informal sector (low HIV-risk)	2117 (77.6%)	944 (72.5%)	1173 (82.3%)	<0.001
Informal sector (high HIV-risk)	144 (5.3%)	74 (5.7%)	70 (4.9%)
Formal sector	467 (17.1%)	284 (21.8%)	183 (12.8%)
Baseline geospatial mobility
Migration last year	201 (7.4%)	116 (8.9%)	85 (6.0%)	0.003
Any travel last 6 months	1385 (50.8%)	618 (47.5%)	767 (53.8%)	0.001
Any work travel last 6 months	323 (11.8%)	283 (21.7%)	40 (2.8%)	<0.001
Any non-work travel last 6 months	564 (20.7%)	203 (15.6%)	361 (25.3%)	<0.001
Sexual behaviour
Number of partners last 5 years
0–1	2042 (74.8%)	815 (62.6%)	1227 (86.0%)	<0.001
2	420 (15.4%)	265 (20.4%)	155 (10.9%)
3 or more	266 (9.8%)	222 (17.0%)	44 (3.1%)
Higher HIV-risk partner^a	248 (9.1%)	144 (11.1%)	104 (7.3%)	0.001
Out of town partner	578 (21.2%)	309 (23.7%)	269 (18.9%)	0.002
Higher risk transactional sexual partner^b	1509 (55.3%)	838 (64.4%)	671 (47.1%)	<0.001
Any concurrent partner	348 (12.8%)	283 (21.7%)	65 (4.65%)	<0.001
HIV prevalence (%)	1110 (40.7%)	520 (39.9%)	590 (41.4%)	0.446

Sexual partnership concurrency is defined as having two or more sexual partnerships within any 1 month over a given time period.

Informal sector (low risk): construction/artisanal labour, student, shopkeeper/market vendor, farming/livestock, household worker/housewife.

Informal sector (high risk): transport driver/tourism, hotel/restaurant/bar worker, fishing/fish trade.

Formal sector: factory worker/mining, government/teacher/military/healthcare.

^a“Higher HIV-risk sexual partner” = one night stand, commercial sex worker or client, casual partner, or inherited partner/inheritor.

^b“Transactional sex partner” = any partner with whom money or goods was paid was exchanged within a sexual relationship.

Men and women exhibited differences in baseline mobility and sexual behaviours (Table 1). While 7.4% of participants reported migration within the past year, more men than women had migrated (8.9% vs 6.0%, respectively, p = 0.003). Over the past 6 months, women reported more travel for any purpose (53.8%) than men (47.5%) (p = 0.001). While more men than women reported work-related travel (21.7% vs 2.8%, p < 0.001), more women than men reported non-work travel (25.3% vs 15.6%, p < 0.001). In addition, 37.4% of men reported two or more sexual partners in the past 5 years (compared to women, 14.0%, p < 0.001), and more men than women reported higher HIV-risk partnerships (11.1% men vs 7.3% women, p = 0.001), and concurrent partnerships (21.7% vs 4.7%, p < 0.001).

Mobility and migration

Men and women exhibited significant differences in migration and mobility across all study follow-up rounds. At all follow-up rounds, women reported more travel within the past 6 months (Table 2). In addition, women reported more non-work travel, while men reported more work-related travel, at all-time points.

Table 2.

Migration and mobility measures by round.

Characteristics at study entry	Baseline	Year-1	Year-2	Year-3
N (number of participants with STI evaluation)	2616	2600	2567	2539
Total (%)
Any migration last 6 months	86 (3.3%)	88 (3.4%)	89 (3.5%)	85 (3.4%)
Any intra-district/Migration last year/6 months	60 (2.3%)	17 (0.7%)	9 (0.4%)	3 (0.1%)
Any inter-district/Migration last year/6 months	57 (2.2%)^a	70 (2.7%)^a	90 (3.5%)	56 (2.2%)^a
Any travel last 6 months	1338 (51.2%)^a	1332 (51.2%)^a	1501 (58.5%)^a	1409 (55.5%)^a
Any work travel last 6 months	310 (11.9%)^a	276 (10.6%)^a	306 (11.9%)^a	305 (12.0%)^a
Any non-work travel last 6 months	1126 (43.1%)^a	1158 (44.5%)^a	1317 (51.3%)^a	1199 (47.2%)^a
Men (%)
N (number of participants)
Any migration last 6 months	47 (3.8%)	46 (3.7%)	47 (3.9%)	44 (3.7%)
Any intra-district/Migration last year/6 months	30 (2.4%)	4 (0.3%)	2 (0.2%)	1 (0.1%)
Any inter-district/Migration last year/6 months	35 (2.9%)	17 (1.4%)	34 (2.8%)	16 (1.3%)
Any travel last 6 months	589 (47.3%)	569 (46.0%)	636 (52.2%)	569 (47.4%)
Any work travel last 6 months	271 (21.8%)	224 (19.1%)	253 (20.8%)	248 (20.7%)
Any non-work travel last 6 months	395 (31.7%)	416 (33.6%)	468 (38.4%)	388 (37.3%)
Women (%)
N (number of participants)
Any migration last 6 months	39 (2.9%)	42 (3.1%)	42 (3.1%)	41 (3.1%)
Any intra-district/Migration last year/6 months	30 (2.2%)	13 (1.0%)	7 (0.5%)	2 (0.2%)
Any inter-district/Migration last year/6 months	21 (1.5%)	53 (3.9%)	56 (1.2%)	40 (3.0%)
Any travel last 6 months	749 (54.7%)	763 (56.0%)	865 (64.2%)	840 (62.8%)
Any work travel last 6 months	39 (2.9%)	52 (3.8%)	53 (3.9%)	57 (4.3%)
Any non-work travel last 6 months	731 (53.4%)	742 (54.4%)	849 (63.0%)	811 (60.7%)

^a(significantly different between men & women p < 0.05).

Longitudinal and multi-year patterns of STI (CT/NG) prevalence

The prevalence of STIs across all years was just under 2% (CT: 1.98%, NG: 1.79%, CT + NG: 0.15%). The STI prevalence across all years did not significantly differ by sex, HIV status, or within years. In Supplemental Figure 1, we present the baseline STI prevalence by age and sex. In general, younger age groups (15–29 years) had the highest prevalence, particularly with an early spike among young females aged 15–19 years. Looking at STI prevalence by year and sex (Supplemental Figure 2), women had higher STI prevalence than men over time and there was an increasing trend in STI prevalence (overall Cochran–Armitage test for trend p = 0.0002, also significant with sex strata 0.0086 for males and 0.0095 for females). In Supplemental Figure 3 displaying multi-year patterns of STI prevalence, the majority (86%) of those with a STI had only one annual survey test positive for an STI over all years.

Bivariate analysis

We identified significant bivariate associations between STIs and mobility and sexual behaviour measures (Figure 1). Bivariate associations with STIs were observed for baseline migration within the past one, two, and 5 years as well as the number of migrations within the past one, two, and 5 years. Significant associations were also observed for work-related travel within the past one and 6 months, for non-work travel in the past 6 months, and for any travel within the past 6 months. Various demographic and behavioural measures also had significant bivariate associations with an increased STI risk, including age less than 25, reporting a higher HIV-risk sexual partner, being single, reporting an out-of-town or any concurrent partner, and having a higher number of partners. Demographic and behavioural measures also had significant bivariate associations with decreased risk of STIs, including reporting condom use at last sex, a sexual partner age difference of greater than 5 years, and being married.

Figure 1.

Associations of mobility and sexual behaviour measures with STI (CT/NG) prevalence.

Multivariate analysis

Significant independent longitudinal associations with any STI were: any migrations in past year, sexual relationship concurrency, being single, higher HIV-risk sexual partners, age less than 25 years, low household wealth, and the relationship between gender and any travel in the past 6 months for work (Table 3). Being older than 25 years was protective from STIs (aOR = 0.51, 95% CI: 0.36–0.71, p < 0.001) while those who were single (aOR = 1.47, 95% CI: 1.03–2.11, p = 0.034) and those reporting low household wealth (aOR = 1.41, 95% CI: 1.06–1.88, p = 0.017) had increased odds of an STI. In addition, those with higher HIV-risk sexual partnerships (aOR = 1.52, 95% CI: 1.10–2.11, p = 0.012) and those reporting concurrent relationships (aOR = 1.75, 95% CI: 1.30–2.36, p < 0.001) had increased odds of an STI. There were also significant interactions between gender and any travel in the past 6 months. Compared to men who did not travel, women who did not travel (aOR = 1.65, 95% CI: 1.26–2.17, p < 0.001) and men who travelled (aOR = 1.76, 95% CI: 1.25–2.50, p = 0.001) had higher odds of an STI. In particular, women who travelled in the past 6 months had more than three times the odds of an STI versus men who did not travel (aOR = 3.31, 95% CI: 1.90–5.76, p < 0.001).

Table 3.

Multivariate analysis results.

Characteristic	aOR (95% CI)	p
(n = 10,319 observations)^a
Age ≥25 years	0.51 (0.36-0.71)	<0.001
Single	1.47 (1.03-2.11)	0.034
Low household wealth index	1.41 (1.06-1.88)	0.017
Concurrent relationship	1.75 (1.30-2.36)	<0.001
Higher HIV-risk partnership	1.52 (1.10-2.11)	0.012
Migration in past year	1.53 (1.08-2.18)	0.016
Past 6-months travel for work by sex
No past 6-months work travel & male	Reference group
No past 6-months work travel & female	1.65 (1.26-2.17)	<0.001
Past 6-months work travel & male	1.76 (1.25-2.50)	0.001
Past 6-months work travel & female	3.31 (1.90-5.76)	<0.001

^aAdjusted for clustering by individual subject id.

Mediation analysis

The causal mediation analysis examined mediation of the relationship between STI prevalence and measures of mobility by key sexual behaviours (higher HIV-risk sexual partner, out-of-town partner, any concurrent partners) (Table 4). For clarity, we present mediation of the STI prevalence at baseline and at last measured STI follow-up association. Overall, we saw evidence of mediation of the relationship between STI prevalence and measures of past 6-months travel by all four sexual behaviour measures examined as indicated by significant natural indirect effect odds ratios at baseline and again at last STI follow-up measurement. The percent mediated by the different sexual behaviours ranged from a low of 10.8% to a high of 22.1%. When we separately examined work related travel and non-work-related travel, we saw strong evidence of mediation by sexual behaviour measures at baseline for non-work-related travel that was less apparent at time of last measured STI follow-up. The percent mediated at baseline ranged from 21.9% to 39.6%. We did not observe significant mediation effects for work related travel at baseline of time of last STI follow-up.

Table 4.

Mediation analysis of STI (CT/NG) prevalence and mobility by sexual behaviours.

Direct effect variable	Mediating variable	STI baseline				STI last follow-up
Direct effect variable	Mediating variable	Total effect OR	Natural direct effect OR	Natural indirect effect OR	% mediated	Total effect OR	Natural direct effect OR	Natural indirect effect OR	% mediated
Any travel	Higher HIV-risk partner	1.58	1.5	1.05	13.7	1.71	1.63	1.05	11.9
	Out of town partner	1.59	1.47	1.08	20.24	1.68	1.61	1.05	10.8
	Higher-risk transactional sex partner	1.59	1.44	1.1	24	1.74	1.62	1.07	16
	Any concurrent partner	1.55	1.43	1.08	22.1	1.65	1.47	1.06	13.5
Any work travel	Higher HIV-risk partner	2.18	2.14	1.02	4.32	1.78	1.77	1	0.84
	Out of town partner	2.15	2.07	1.04	7.19	1.75	1.77	0.99	* - med
	Higher-risk transactional sex partner	2.23	2.26	0.99	*-Med	1.83	1.8	1.02	3.9
	Any concurrent partner	2.06	1.88	1.09	16.7	1.75	1.69	1.04	8.3
Any non-work travel	Higher HIV-risk partner	1.25	1.2	1.05	21.9	1.45	1.4	1.04	11.8
	Out of town partner	1.26	1.18	1.07	32.3	1.4	1.37	1.04	12.1
	Higher-risk transactional sex partner	1.26	1.15	1.09	39.6	1.46	1.36	1.07	21.6
	Any concurrent partner	1.25	1.17	1.06	29.4	1.42	1.36	1.05	14.8

Bold = p < 0.05 significance.

ALL Models include Age<25 and Sex.

^* = negative mediation observed.

Discussion

Gendered patterns of geospatial mobility, non-local partnerships, sexual concurrency, higher HIV-risk sexual partnerships, and age significantly influenced longitudinal STI (CT, NG) prevalence in rural East African communities, with evidence of mediation of the relationship between STI prevalence and mobility by sexual behaviour risks among men and women. These findings are corroborated by other cross-sectional STI studies in the region,^22,35,41 although ours is one of the few longitudinal studies of STI risk in SSA, allowing us to examine patterns of STI acquisition over time. A large majority of participants who had an STI (86%) experienced just one STI positive test per annual examination. Repeat STIs during the 4-year follow-up were uncommon. However, the observed significant and increasing STI prevalence for males and females reflects global STI trends and associated increases in population mobility and multi-drug-resistant gonorrhoea.^26,46

We also found differences in STI prevalence among men and women. Young women experienced an early spike in STIs, which is consistent with data in the literature for STIs and HIV age prevalence patterns.^30,32 In addition, at all follow-up time points, women had higher STI prevalence than men. This is similar to findings from a cross-sectional study of STI prevalence among men and women in South Africa.⁴⁷

In multivariate models, any migrations in the past year, those with higher HIV-risk sexual partnerships, and those reporting concurrent relationships had increased odds of an STI. Similarly, prior cross-sectional studies have reported associations between STIs/HIV and mobility,^1,3,15,17 higher HIV-risk sexual partnerships,³⁵ and concurrent relationships.^22,36 Being older than 25 years was protective from STIs, while being single and reporting low household wealth both increased STI odds. The finding that being younger than 25 was protective for STIs is in contrast to other published literature that has reported a spike in STIs/HIV among younger age groups,^26,48 and young women in this setting often engage in age-disparate relationships, which may put them at heightened risk for STIs.^49,50

In addition, there were significant interactions between gender and any travel in the preceding 6 months. Compared to men who did not travel, women who did not travel and men who travelled had higher odds of an STI. This may reflect men’s sexual behaviours while travelling^3,51 and women’s underlying risk for STIs at home. Women who travelled in the past 6 months had three times the odds of an STI compared to men who did not travel. This may reflect greater sexual risks engaged in by mobile women. The literature has many examples of how mobility can increase HIV and STI risks for female market traders, whose mobility affords opportunities for transactional sex,⁵² and female fish traders who exchange sex for access to fish.^5,21

In causal mediation analysis we found evidence of mediation of the relationship between STI (CT, NG) prevalence and past 6-months travel by multiple sexual risk behaviours (higher HIV-risk sexual partner, out-of-town partner, any concurrent partners). This finding supports the hypothesis that the influence of mobility on higher STI prevalence reflects differences in sexual behaviours among mobile persons. Further investigation is needed to determine if these differences are general for mobile persons or occur primarily when traveling away from the primary residence. Separating work from non-work travel reduced evidence of mediation; this may be due to the gendered nature of STI risks and the contexts of mobility for men and women.⁶

This study has some limitations. While the study spanned 4 years of longitudinal data collection, the findings reported here represent annual STI point prevalence. In addition, the study results are reflective of western Kenya and may not be generalizable to other geographic locations in Africa. However, the prevalence of geographic mobility is reflective of many of the regions and countries in the area. In the analysis, we included known confounders for CT and NG acquisition risks, but it is possible there were other unmeasured confounders that could have improved the regression analysis. The potential bias associated with self-report of mobility was not thought to be a major issue as mobility is not a stigmatized behaviour and is common in the region. At the start of the study, all participants with positive STI test results were referred to and received STI treatment. Because annual point-prevalence was measured, intervening STI acquisition and self-clearance in the absence of treatment would have been missed and this could lead to an underestimate of true STI incidence and prevalence.^53,54

The results of this study highlight the globally observed problem of increasing STI rates and the influence of mobility and female gender on STI acquisition. Offering a heightened package of sexual and reproductive health services for mobile women and men that includes STI screening and treatment is warranted. Potential approaches to addressing the increased risk for mobile persons, particularly mobile women, could include targeted provision of STI services at common transit hubs or on market days (when and where mobile women congregate) to offer needed services such as rapid/mobile STI testing⁵⁵ and treatment to reduce onwards transmission of STIs and HIV.

Supplemental material

Supplemental material - Geospatial mobility, non-local partners, concurrent sexual partnerships, and gender influence longitudinal STI prevalence in rural eastern Africa

Supplemental material for Geospatial mobility, non-local partners, concurrent sexual partnerships, and gender influence longitudinal STI prevalence in rural eastern Africa by Edwin D. Charlebois, Sarah A. Gutin, Torsten B. Neilands, Monica Getahun, Adam Akullian, Irene Maeri, Patrick Eyul, Daniel Omoding, Jaffer Okiring, Craig R. Cohen, Monica Gandhi, Elizabeth A. Bukusi, Moses R. Kamya and Carol S. Camlin in International Journal of STD & AIDS

Footnotes

Acknowledgements

We gratefully acknowledge the communities and participants who made this study possible, the Ministries of Health of Uganda and Kenya, and the Mobility in SEARCH and SEARCH research teams, advisory boards and collaborators. We also thank S. Ssali and W. Owino for their contributions to this research.

ORCID iDs

Edwin D. Charlebois

Sarah A. Gutin

Ethical considerations

Ethical approvals were received from the University of California San Francisco Committee on Human Research (14-15058), Ethical Review Committee of the Kenya Medical Research Institute (KEMRI/SERU/CMR/3052), Makerere University School of Medicine Research and Ethics Committee (2015-040), and the Uganda National Council for Science and Technology (HS 1834). The procedures used in this study adhere to the tenets of the Declaration of Helsinki.

Consent to participate

All study participants provided written informed consent.

Author contributions

CSC, EDC, and TBN conceptualized and designed the study with helpful contributions from EAB, CRC, MRK, and AA. PE, IM, DO, JO and MG oversaw study implementation, including supervision of field activities and quality assurance. CSC, SAG, and MG verified the data. TBN and EDC designed the analytic data strategy and conducted the analyses. EDC and SAG led the writing of the manuscript. All authors edited, helped with revisions, and approved the final manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a grant from the National Institutes of Health, NIMH under award number R01MH104132 (Camlin, Mobility in SEARCH) and K24MH126808 (Camlin). SAG was supported by the National Institutes of Mental Health under grant K01MH132435. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data and statistical code underlying this analysis are available in the Dryad Digital Repository at https://doi.org/10.5061/dryad.dv41ns26f.*

Supplemental material

Supplemental material for this article is available online.

References

Camlin

Akullian

Neilands

, et al. Gendered dimensions of population mobility associated with HIV across three epidemics in rural eastern Africa. Health Place 2019; 57: 339–351. https://doi.org/10.1016/j.healthplace.2019.05.002

Camlin

Snow

Hosegood

. Gendered patterns of migration in rural South Africa. Popul Space Place 2014; 20: 528–551. https://doi.org/10.1002/psp.1794

Dzomba

Tomita

Govender

, et al. Effects of migration on risky sexual behavior and HIV acquisition in South Africa: a systematic review and meta-analysis, 2000–2017. AIDS Behav 2019; 23: 1396–1430. https://doi.org/10.1007/s10461-018-2367-z

Camlin

Charlebois

. Mobility and its effects on HIV acquisition and treatment engagement: recent theoretical and empirical advances. Curr HIV AIDS Rep 2019; 16: 314–323. https://doi.org/10.1007/s11904-019-00457-2

Kwena

Nakamanya

Nanyonjo

, et al. Understanding mobility and sexual risk behaviour among women in fishing communities of Lake Victoria in East Africa: a qualitative study. BMC Public Health 2020; 20: 1–10. https://doi.org/10.1186/s12889-020-09085-7

Camlin

Kwena

Dworkin

, et al. “She mixes her business”: HIV transmission and acquisition risks among female migrants in western Kenya. Soc Sci Med 2014; 102: 146–156. https://doi.org/10.1016/j.socscimed.2013.11.004

Olawore

Tobian

AAR

Kagaayi

, et al. Migration and risk of HIV acquisition in Rakai, Uganda: a population-based cohort study. Lancet HIV 2018; 5: e181–e189. https://doi.org/10.1016/S2352-3018(18)30009-2

Glynn

Pönnighaus

Crampin

, et al. The development of the HIV epidemic in Karonga district, Malawi. AIDS 2001; 15: 2025–2029. https://doi.org/10.1097/00002030-200110190-00016

Jochelson

Mothibeli

Leger

. Human immunodeficiency virus and migrant labor in South Africa. Int J Health Serv 1991; 21: 157–173. https://doi.org/10.2190/11UE-L88J-46HN-HR0K

10.

Coffee

Garnett

Mlilo

, et al. Patterns of movement and risk of HIV infection in rural Zimbabwe. JID (J Infect Dis) 2005; 191(Suppl): S159–S167. https://doi.org/10.1086/425270

11.

Valdano

Okano

Colizza

, et al. Using mobile phone data to reveal risk flow networks underlying the HIV epidemic in Namibia. Nat Commun 2021; 12: 2837. https://doi.org/10.1038/s41467-021-23051-w

12.

Ratmann

Kagaayi

Hall

, et al. Quantifying HIV transmission flow between high-prevalence hotspots and surrounding communities: a population-based study in Rakai, Uganda. Lancet HIV 2020; 7: e173–e183. https://doi.org/10.1016/S2352-3018(19)30378-9

13.

Grabowski

Lessler

Bazaale

, et al. Migration, hotspots, and dispersal of HIV infection in Rakai, Uganda. Nat Commun 2020; 11: 1–12.

14.

Okano

Sharp

Valdano

, et al. HIV transmission and source-sink dynamics in Sub-Saharan Africa. Lancet HIV 2020; 7: e209–e214. https://doi.org/10.1016/S2352-3018(19)30407-2

15.

Cassels

Jenness

Biney

AAE

, et al. Geographic mobility and potential bridging for sexually transmitted infections in agbogbloshie, Ghana. Soc Sci Med 2017; 184: 27–39. https://doi.org/10.1016/j.socscimed.2017.05.003

16.

Sileo

Wanyenze

Kizito

, et al. Multi-level determinants of clinic attendance and antiretroviral treatment adherence among fishermen living with HIV/AIDS in communities on Lake Victoria, Uganda. AIDS Behav 2019; 23: 406–417. https://doi.org/10.1007/s10461-018-2207-1

17.

Kwena

Camlin

Shisanya

, et al. Short-term mobility and the risk of HIV infection among married couples in the fishing communities along Lake Victoria, Kenya. PLoS One 2013; 8: e54523. https://doi.org/10.1371/journal.pone.0054523

18.

Fiorella

Camlin

Salmen

, et al. Transactional fish-for-sex relationships amid declining fish access in Kenya. World Dev 2015; 74: 323–332. https://doi.org/10.1016/j.worlddev.2015.05.015

19.

Camlin

Akullian

Neilands

, et al. Population mobility associated with higher risk sexual behaviour in eastern African communities participating in a universal testing and treatment trial. JIAS 2018; 21: e25115. https://doi.org/10.1002/jia2.25115

20.

Gutin

Neilands

Charlebois

, et al. Mobility is associated with higher-risk sexual partnerships among both men and women in Co-resident couples in rural Kenya and Uganda: a longitudinal cohort study. AIDS Behav 2022; 27: 1418–1429, Published Online First. https://doi.org/10.1007/s10461-022-03878-0

21.

Camlin

Kwena

Dworkin

. Jaboya vs. jakambi: status, negotiation, and HIV risks among female migrants in the “sex for fish” economy in nyanza province, Kenya. AIDS Educ Prev 2013; 25: 216–231. https://doi.org/10.1521/aeap.2013.25.3.216

22.

Okiring

Getahun

Gutin

, et al. Sexual partnership concurrency and age disparities associated with sexually transmitted infection and risk behavior in rural communities in Kenya and Uganda. Int J Infect Dis 2022; 120: 158–167. https://doi.org/10.1016/j.ijid.2022.04.038

23.

Kwena

Shisanya

Bukusi

, et al. Jaboya (“sex for fish”): a qualitative analysis of contextual risk factors for extramarital partnerships in the fishing communities in Western Kenya. Arch Sex Behav 2017; 46: 1877–1890. https://doi.org/10.1007/s10508-016-0930-0

24.

National AIDS and STI Control Programme (NASCOP) . Kenya Population-based HIV impact assessment KENPHIA 2018, 2022.

25.

National AIDS Control Council (NACC) . Kenya aids response progress report 2018. 2018.

26.

World Health Organization . Global progress report on HIV, viral hepatitis and sexually transmitted infections, 2021. Accountability for the global health sector strategies 2016–2021: actions for impact. World Health Organization, 2021. https://www.who.int/publications/i/item/9789240027077

27.

Hussen

Wachamo

Yohannes

, et al. Prevalence of chlamydia trachomatis infection among reproductive age women in sub Saharan Africa: a systematic review and meta-analysis. BMC Infect Dis 2018; 18: 596. https://doi.org/10.1186/s12879-018-3477-y

28.

Kassa

Hussen

Hadra

, et al. Prevalence of Neisseria gonorrhoeae infection among women of reproductive age in Sub-Saharan Africa: a systematic review and meta-analysis. Eur J Contracept Reprod Health Care 2020; 25: 365–371. https://doi.org/10.1080/13625187.2020.1779688

29.

Hussen

Tadesse

. Prevalence of syphilis among pregnant women in Sub-Saharan Africa: a systematic review and meta-analysis. BioMed Res Int 2019; 2019: 4562385. https://doi.org/10.1155/2019/4562385

30.

Torrone

Morrison

Chen

P-L

, et al. Prevalence of sexually transmitted infections and bacterial vaginosis among women in sub-Saharan Africa: an individual participant data meta-analysis of 18 HIV prevention studies. PLoS Med 2018; 15: e1002511. https://doi.org/10.1371/journal.pmed.1002511

31.

Sexton

Garnett

Røttingen

J-A

. Metaanalysis and metaregression in interpreting study variability in the impact of sexually transmitted diseases on susceptibility to HIV infection. Sex Transm Dis 2005; 32: 351–357. https://doi.org/10.1097/01.olq.0000154504.54686.d1

32.

Culbreth

Swahn

Salazar

, et al. Risk factors associated with HIV, sexually transmitted infections (STI), and HIV/STI Co-infection among youth living in the slums of kampala, Uganda. AIDS Behav 2020; 24: 1023–1031. https://doi.org/10.1007/s10461-019-02444-5

33.

Kalichman

Pellowski

Turner

. Prevalence of sexually transmitted co-infections in people living with HIV/AIDS: systematic review with implications for using HIV treatments for prevention. Sex Transm Infect 2011; 87: 183–190. https://doi.org/10.1136/sti.2010.047514

34.

Camlin

Getahun

Alullian

, et al. Measures of recent migration predict STI (GC/CT) among adults in East African communities participating in the SEARCH study. IAS 2017- 9th IAS Conference on HIV Science, 23 - 26 July 2017. Abstract TUPEC0808 pg. 296. https://ias2017.org/Portals/1/Files/IAS2017_LO.compressed4c6a.pdf?fileticket=m3LSDs1z4QY%3d&tabid=577&portalid=1

35.

Matovu

JKB

Ssebadduka

. Sexual risk behaviours, condom use and sexually transmitted infection treatment-seeking behaviours among female sex workers and truck drivers in Uganda. Int J STD AIDS 2012; 23: 267–273. https://doi.org/10.1258/ijsa.2011.011313

36.

Morris

Ferguson

. Sexual and treatment-seeking behaviour for sexually transmitted infection in long-distance transport workers of East Africa. Sex Transm Infect 2007; 83: 242–245. https://doi.org/10.1136/sti.2006.024117

37.

Kiyingi

Witte

Nabunya

, et al. Predictors of mobility among women engaged in commercial sex work in Uganda using generalized estimating equations model. Int J STD AIDS 2023; 34: 633–640. https://doi.org/10.1177/09564624231167910

38.

Khalifa

Findley

Gummerson

, et al. Associations between mobility, food insecurity, and transactional sex among women in cohabitating partnerships: an analysis from 6 African countries 2016-2017. J Acquir Immune Defic Syndr 2022; 90: 388–398. https://doi.org/10.1097/QAI.0000000000002995

39.

Fiorella

Desai

Miller

, et al.

A review of transactional sex for natural resources: under-researched, overstated, or unique to fishing economies?

Glob Public Health 2019; 14: 1803–1814. https://doi.org/10.1080/17441692.2019.1625941

40.

Salia

Sidat

Dias

, et al. High mobility and STIs/HIV among women informal cross border traders in southern Mozambique: exploring knowledge, risk perception, and sexual behaviors. Int J Environ Res Publ Health 2020; 17: 17134724. https://doi.org/10.3390/ijerph17134724

41.

Diabaté

Chamberland

Geraldo

, et al. Gonorrhea, chlamydia and HIV incidence among female sex workers in cotonou, Benin: a longitudinal study. PLoS One 2018; 13: e0197251. https://doi.org/10.1371/journal.pone.0197251

42.

Wang

Tapia

Oluoch

, et al. Adolescent girls and young women in Kenya demonstrate rapid STI incidence following first sex: data from a longitudinal cohort. J Adolesc Health 2023; 72: 568–574. https://doi.org/10.1016/j.jadohealth.2022.10.026

43.

Luke

Clark

Zulu

. The relationship history calendar: improving the scope and quality of data on youth sexual behavior. Demography 2011; 48: 1151–1176. https://doi.org/10.1007/s13524-011-0051-2

44.

VanderWeele

. Explanation in causal inference: methods for mediation and interaction. Oxford University Press, 2015.

45.

Emsley

Liu

. PARAMED: stata module to perform causal mediation analysis using parametric regression models. Statistical Software Components, 2013.

46.

WHO. Multi-drug resistant gonorrhoea. 2025. https://www.who.int/news-room/fact-sheets/detail/multi-drug-resistant-gonorrhoea (Accessed 4 Feb2026).

47.

Francis

Mthiyane

Baisley

, et al. Prevalence of sexually transmitted infections among young people in South Africa: a nested survey in a health and demographic surveillance site. PLoS Med 2018; 15: e1002512. https://doi.org/10.1371/journal.pmed.1002512

48.

Grabowski

Mpagazi

Kiboneka

, et al. The HIV and sexually transmitted infection syndemic following mass scale-up of combination HIV interventions in two communities in southern Uganda: a population-based cross-sectional study. Lancet Global Health 2022; 10: e1825–e1834. https://doi.org/10.1016/S2214-109X(22)00424-7

49.

Stoner

MCD

Nguyen

Kilburn

, et al. Age-disparate partnerships and incident HIV infection in adolescent girls and young women in rural South Africa. AIDS 2019; 33: 83–91. https://doi.org/10.1097/QAD.0000000000002037

50.

Schaefer

Gregson

Eaton

, et al. Age-disparate relationships and HIV incidence in adolescent girls and young women: evidence from Zimbabwe. AIDS 2017; 31: 1461–1470. https://doi.org/10.1097/QAD.0000000000001506

51.

Deane

Samwell Ngalya

Boniface

, et al. Exploring the relationship between population mobility and HIV risk: evidence from Tanzania. Glob Public Health 2018; 13: 173–188. https://doi.org/10.1080/17441692.2016.1178318

52.

Camlin

El Ayadi

Kwena

, et al. High mobility and HIV prevalence among female market traders in East Africa in 2014. J Acquir Immune Defic Syndr 2017; 74: e121–e128. https://doi.org/10.1097/QAI.0000000000001252

53.

Jordan

Wilson

Ren

, et al. Natural clearance of Chlamydia trachomatis infection is associated with distinct differences in cervicovaginal metabolites. J Infect Dis 2023; 228: 1119–1126. https://doi.org/10.1093/infdis/jiad155

54.

Teker

de Vries

Heijman

, et al. Spontaneous clearance of asymptomatic anogenital and pharyngeal neisseria gonorrhoeae: a secondary analysis from the NABOGO trial. Sex Transm Infect 2023; 99: 219–225. https://doi.org/10.1136/sextrans-2022-055488

55.

Peters

RPH

Klausner

Mazzola

, et al. Novel lateral flow assay for point-of-care detection of Neisseria gonorrhoeae infection in syndromic management settings: a cross-sectional performance evaluation. Lancet 2024; 403: 657–664. https://doi.org/10.1016/S0140-6736(23)02240-7

Geospatial mobility,non-local partners,concurrent sexual partnerships,and gender influence longitudinal STI prevalence in rural eastern Africa

Abstract

Background

Methods

Results

Conclusions

Keywords

Introduction

Methods

Study design, participants, and setting

Data collection

Measures

Demographics

Geospatial mobility

Sexual behaviour

Data analysis

Ethical approvals

Results

Sample demographics

Mobility and migration

Longitudinal and multi-year patterns of STI (CT/NG) prevalence

Bivariate analysis

Multivariate analysis

Mediation analysis

Discussion

Supplemental material

Supplemental material - Geospatial mobility, non-local partners, concurrent sexual partnerships, and gender influence longitudinal STI prevalence in rural eastern Africa

Footnotes

Acknowledgements

ORCID iDs

Ethical considerations

Consent to participate

Author contributions

Funding

Declaration of conflicting interests

Data Availability Statement

Supplemental material

References