Combined impact of recent sexual activity and vaginal product use on genital immune activation in South African women: implications for HIV risk

Abstract

Adolescent girls and young women (AGYW) in South Africa experience a disproportionate burden of HIV infection. Genital inflammation has been implicated in increased HIV susceptibility through enhanced availability of target immune cells; however, the effects of recent sexual activity and vaginal-stimulating product (VSP) use on mucosal immune responses remain incompletely defined. This study characterised mucosal immune profiles following recent sexual exposure and VSP use in adolescent and adult women. HIV-negative sexually active adolescents (14–19 years) and adult women (25–35 years) were enrolled in a longitudinal cohort in KwaZulu-Natal, South Africa. Cervical cytobrush samples and cervicovaginal secretions were collected at baseline (≥ 2 weeks abstinence) and following reported sexual activity and/or VSP use. Cervical T-cell activation markers (CD38, HLA-DR, CCR5, α4β7) were quantified by flow cytometry, and 18 cytokines and chemokines were measured using a multiplex assay. Associations with time since last sexual intercourse were assessed using correlation analyses, and factors associated with genital inflammation were evaluated using weighted logistic regression. Recent sexual activity showed inverse associations with several pro-inflammatory cytokines, including MIP-1α, IL-1α, IL-1β, GM-CSF, and IL-17, consistent with a transient pattern of mucosal immune activation following exposure. In multivariable models, recent sexual activity (OR 0.58; p = 0.009) and bacterial vaginosis (OR 2.38; p = 0.017) were independently associated with genital inflammation, whereas VSP use was not statistically significant after adjustment (OR 1.63; p = 0.187). Although higher cytokine concentrations were observed among VSP users in unadjusted analyses, including IL-1α in adolescents (p = 0.001) and MIP-1α in adults (p = 0.017), these associations did not remain significant after correction for multiple testing. Overall, these findings suggest that recent sexual activity is associated with transient changes in mucosal immune mediators, while VSP use does not independently predict genital inflammation in this cohort after adjustment for confounding and multiple testing. These results highlight the importance of timing of exposure in interpreting mucosal immune measurements and support further longitudinal studies with precise sampling intervals to better define post-coital immune dynamics and the role of behavioural practices in shaping genital immunity.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1038/s41598-026-60851-w.

Introduction

South Africa remains the epicenter of the global HIV epidemic, with an estimated 7.7 million people living with HIV by 2023¹. Adolescent girls and young women (AGYW) are disproportionately affected, accounting for approximately 89,000 new infections that year alone¹. Despite advances in prevention and treatment, AGYW continues to acquire HIV at significantly higher rates than adult women and their male peers¹, reflecting the complex interplay of social, behavioral, and biological factors.

The adolescent female genital tract (FGT) is thought to be biologically more susceptible to HIV, potentially due to several factors, including endogenous and exogenous sex hormones, sexually transmitted infections (STIs), genital microbiome composition, epithelial barrier integrity, increased frequencies of HIV target cells, and genital inflammation^2–7. In addition, cervical immaturity, characterized by a fragile single-layered columnar epithelium, may increase the likelihood of microabrasions during intercourse, facilitating pathogen entry⁸. Recent sexual activity can acutely alter the immune environment in the female genital tract, as seminal fluid has been linked to elevated vaginal pH, reduced antimicrobial peptides such as defensins⁹, and increased pro-inflammatory cytokines and chemokines (IL-1α, IL-1β, IL-6, GM-CSF)^10,11, which may recruit and activate HIV target cells (including CD4 + T cells, macrophages, and dendritic cells expressing CCR5 and CXCR4), potentially increasing susceptibility to infection. Recent work in STI-free heterosexual couples further demonstrated that cervicovaginal cytokines, including IP-10 and soluble E-cadherin, increase rapidly following penile–vaginal sex, regardless of condom use, alongside increases in endocervical HIV target cells such as Th17 cells and activated dendritic cells¹². While some of these responses resolve within 72 h, elevated levels of activated CD4 + T cells may persist for longer, suggesting a transient but potentially critical window of increased HIV susceptibility.

Furthermore, in regions with high HIV prevalence, the use of vaginally inserted sexual-enhancing products (VSPs) is common¹³ and has been associated with mucosal inflammation and elevated HIV risk^14,15. VSPs, often used for perceived sexual or hygienic benefits, may further alter mucosal immunity in ways that are poorly understood and under characterized in younger populations, which has implications for HIV acquisition risk.

Taken together, these observations highlight a critical knowledge gap regarding how recent sexual activity and VSP use influence genital tract immune activation, particularly in adolescent girls. This study aimed to characterize genital tract immune activation in adolescent and adult women following recent sexual intercourse, with and without VSP use. Specifically, we measured genital tract cytokine and chemokine profiles along with cytobrush-derived T cell activation to identify biological markers of inflammation and HIV vulnerability. The study is unique in assessing young women both before and after sexual activity and in simultaneously comparing two age groups, allowing investigation of whether immune effects of sexual activity are amplified by VSP use and how responses differ across stages of reproductive and cervical maturation. By integrating behavioural exposures with detailed biological profiling, this work addresses how co-occurring factors may synergistically increase HIV susceptibility and informs prevention strategies that consider both age-specific mucosal biology and modifiable behavioural risks.

Methods

Study design, participants, and sample collection

This study is nested within the Mucosal Injury from Sexual Trauma/Contact (MIST/C) study, a longitudinal cohort conducted at two sites in South Africa: the Centre for the AIDS Programme of Research in South Africa (CAPRISA) Vulindlela Clinic Research Site (KwaZulu-Natal, KZN) and the Desmond Tutu HIV Centre (Cape Town, Western Cape). A detailed description of the parent cohort has been published previously¹⁶. For the present analysis, only participants enrolled at the CAPRISA Vulindlela site were included, as VSP use was not reported among Cape Town participants. Eligible participants were HIV-negative adolescent girls and young women (AGYW; 14–19 years) and adult women (25–35 years) who reported prior sexual activity. Additional inclusion criteria included willingness to provide written informed consent (or assent with parental consent where applicable), willingness to comply with study procedures, not being pregnant, no antibiotic use within the past month, and no history of cervical pathology. Participants requiring STI treatment were treated and re-screened 30 days after completion of antibiotic therapy.

All participants provided written informed consent. Ethical approval was obtained from the Biomedical Research Ethics Committee of the University of KwaZulu-Natal (BREC/00001105/2020), and all procedures were conducted in accordance with the Declaration of Helsinki.

A total of 251 HIV-negative participants (188 AGYW and 63 adult women) were included in this analysis. HIV status was confirmed at screening using rapid diagnostic tests (Uni-Gold™ Recombigen^® HIV-1/2 [Trinity Biotech, Ireland] and Determine™ HIV-1/2 [Abbott Diagnostics, USA]), with pre- and post-test counselling provided. Only HIV-negative participants were enrolled. Structured questionnaires were administered by trained clinical staff in private settings to collect demographic, socio-behavioural, and clinical data. The questionnaire was developed based on established HIV risk factors and refined through expert consultation, as previously described¹⁶. To improve recall accuracy, participants were provided with diaries to prospectively record sexual activity, coital frequency, and vaginal product use. Participants were eligible for baseline sampling after abstaining from sexual activity and VSP use for at least two weeks, as determined by self-report. Participants were instructed to return to the clinic within 72 h following sexual activity and/or VSP use; however, some returned up to six days post-exposure. Samples were collected at two time points: baseline (≥ 2 weeks abstinence) and post-exposure. Cervicovaginal secretions were collected using menstrual cups inserted by a trained clinician and left in place for approximately 30 min. Samples were transported on ice, processed within 6 h of collection, and stored at − 80 °C until analysis. Prior to cytokine analysis, samples were thawed and filtered using SPIN-X 0.2 μm cellulose acetate filters (Sigma-Aldrich, USA). Cervical mononuclear cells (CMCs) were collected using a Digene cervical cytobrush (Qiagen) and processed as previously described¹⁷.

Sexually transmitted infection testing and genital health assessments

At each visit, participants were screened for sexually transmitted infections (STIs). In KwaZulu-Natal, urine samples were tested for Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) using the Xpert^® CT/NG assay (Cepheid, USA), which has high sensitivity and specificity. Detection of Trichomonas vaginalis (TV) was performed using lateral vaginal wall swabs assessed by wet mount microscopy in combination with the OSOM^® Trichomonas Rapid Test (Sekisui Diagnostics, USA). Bacterial vaginosis (BV) was assessed using Nugent scoring of Gram-stained vaginal smears. Vaginal pH was measured by clinical staff using pH indicator strips (Multistix^®, Siemens), by applying vaginal swab samples to the strip and comparing the resulting color change to a standardized reference chart. The presence of fungal elements (yeast and hyphae) was assessed using potassium hydroxide (KOH) microscopy on lateral vaginal wall swabs.

Flow cytometry and immune cell phenotyping

Multiparametric flow cytometry was used to assess immune cell phenotypes in cervical cytobrush-derived cells (CMCs). The optimised antibody panel included markers for T cell lineage (CD3, CD4, CD8), activation (CD38, HLA-DR), HIV co-receptor expression (CCR5), and mucosal homing (α4β7), as well as CCR6 and CCR10 to define Th17-like cells (CCR6⁺CCR10⁻). Exclusion markers included CD14 and CD19, as well as a fixable viability dye to exclude non-viable cells All antibodies were obtained from R&D Systems (USA) (Supplementary Table 1). Fluorescence minus one (FMO) controls were used to define gating thresholds. Compensation was performed using single-stained controls, and instrument performance was verified using calibration beads. Samples were acquired on a BD LSRFortessa™ flow cytometer (Becton Dickinson, USA) using FACSDiva software and analyzed using FlowJo software (version 10.7.1). The gating strategy used to define CD4 + and Th17-like T cell subsets and activation profiles is shown in Supplementary Fig S1.

Cytokine and chemokine quantification

Cytokine and chemokine concentrations were measured in cervicovaginal secretions using Bio-Rad multiplex cytokine/chemokine assay kits (Bio-Rad, Hercules, CA, USA) on the Bio-Plex™ 200 platform. Samples were thawed, centrifuged, and filtered prior to analysis. Analytes included interleukin (IL)-1α, IL-1β, IL-6, IL-8, IL-10, IL-17, IL-1 receptor antagonist (IL-1rα), tumor necrosis factor (TNF)-α, TNF-related apoptosis-inducing ligand (TRAIL), interferon (IFN)-α, interferon gamma-induced protein 10 (IP-10), monocyte chemoattractant protein-1 (MCP-1), macrophage inflammatory protein (MIP)-1α, MIP-1β, MIP-3α, granulocyte colony-stimulating factor (GM-CSF), and vascular endothelial growth factor (VEGF). Concentrations were calculated using five-parameter logistic (5-PL) standard curves.

Samples were assayed across multiple plates. To control for inter- and intra-assay variability, five samples were included in duplicate on each plate and inter-plate variability was minimized through inclusion of shared control samples across all plates and stringent quality control criteria. Acceptable assay precision was defined as < 25% coefficient of variation (CV). Values below the lower limit of detection were assigned half the lower limit of quantification (LLOQ/2). Cytokine concentrations were expressed as pg/mL without normalization to total protein concentration, as all samples were processed under standardized conditions. This approach is commonly used in mucosal secretion studies, where variation in total protein may reflect differences in secretion composition rather than true inflammatory activity.

Statistical analysis

Cytokine concentrations and immune cell marker expression were summarized using medians and interquartile ranges (IQRs), while categorical variables were summarized as frequencies and percentages. Group comparisons were performed using the Wilcoxon rank-sum test for continuous variables and Pearson’s chi-squared test or Fisher’s exact test for categorical variables, as appropriate. To evaluate changes associated with sexual activity and VSP use, cytokine concentrations were compared between paired samples collected at baseline (following ≥ 2 weeks of abstinence from sexual activity and VSP use) and post-exposure. Post-exposure samples were collected within 72 h of sexual activity and/or VSP use; however, some samples were collected up to 144 h later. Spearman’s rank correlation coefficient was used to assess associations between time since last sexual activity and levels of inflammatory cytokines and activated T cell subsets. Correlation coefficients (r) and corresponding p-values are reported. Genital inflammation was defined as the presence of ≥ 5 out of 9 pro-inflammatory cytokines above the 75th percentile, based on previously published criteria¹⁸. This composite biomarker-derived definition was used to classify participants as having high versus low inflammation and was not based on clinical or imaging assessments. Weighted logistic regression models were used to identify factors associated with inflammation, adjusting for potential confounders including age, bacterial vaginosis (BV) status, and STI status. To explore patterns of cytokine expression in relation to behavioural and clinical variables, heatmaps were generated in R (version 4.4.0) using hierarchical clustering based on Euclidean distance. To account for multiple comparisons, p-values were adjusted using the Benjamini–Hochberg false discovery rate (FDR) method. All statistical tests were two-tailed, and statistical significance was defined as p ≤ 0.05. Analyses were performed using GraphPad Prism v8 (GraphPad Software), STATA v17 (StataCorp, College Station, TX, USA), and R (version 4.4.0).

Results

Demographic, social behavioural, and clinical characteristics

Participant characteristics stratified by VSP use within adolescents and adults are shown in Table 1. Among adolescents (n = 188), no statistically significant differences were observed between VSP users and non-users in age, age at sexual debut, vaginal pH, number of lifetime sexual partners, condom use, contraceptive use, transactional sex, or prevalence of BV and STIs, including Chlamydia trachomatis, Neisseria gonorrhoeae, and Trichomonas vaginalis (all p > 0.05). Among adult women (n = 63), most demographic, behavioural, and clinical characteristics were also comparable between VSP users and non-users, including age, age at sexual debut, vaginal pH, number of lifetime sexual partners, contraceptive use, and prevalence of BV and STIs (all p > 0.05). However, in adults, differences in behavioural variables were observed, with higher reported transactional sex and lower condom use among VSP users compared with non-users; these differences did not reach statistical significance.

Table 1.

Baseline characteristics of participants stratified by VSP use.

Characteristic	Adolescents (N = 188)			Adults (N = 63)
	Non-users (n = 110)	VSP users (n = 78)	p-value	Non-users (n = 47)	VSP users (n = 16)	p-value
Age (years)	18 (17–18)	17 (16–18)	0.060	28 (26–30)	29 (27–32)	0.393
Age at sexual debut (years)	16 (15–17)	16 (15–17)	0.133	17 (15–19)	17 (16–18)	0.719
Vaginal pH	4.7 (4.4–5.3)	4.7 (4.4–5.0)	0.157	4.7 (4.7–5.0)	5.0 (4.6–5.3)	0.467
Lifetime sexual partners	2.0 (1.0–2.0)	1.0 (1.0–2.0)	0.372	3.0 (2.0–4.0)	4.0 (2.8–6.0)	0.162
Condom use			0.807			0.153
No	92 (88%)	69 (90%)		33 (73%)	15 (94%)
Yes	12 (12%)	8 (10%)		12 (27%)	1 (6.3%)
Contraceptive use			0.367			1.000
No	51 (46%)	31 (40%)		13 (28%)	4 (25%)
Yes	59 (54%)	47 (60%)		34 (72%)	12 (75%)
Transactional sex			1.000			0.075
No	106 (96%)	75 (97%)		40 (85%)	10 (63%)
Yes	4 (3.6%)	2 (2.6%)		7 (15%)	6 (38%)
Bacterial vaginosis (Nugent score)			0.343			0.863
BV	41 (41%)	27 (40%)		18 (39%)	7 (44%)
Intermediate	23 (23%)	22 (32%)		9 (20%)	2 (13%)
No BV	36 (36%)	19 (28%)		19 (41%)	7 (44%)
Neisseria gonorrhoeae			0.694			1.000
No	96 (91%)	67 (93%)		43 (96%)	16 (100%)
Yes	9 (8.6%)	5 (6.9%)		2 (4.4%)	0 (0%)
Chlamydia trachomatis			0.607			0.325
No	71 (68%)	46 (64%)		40 (87%)	16 (100%)
Yes	34 (32%)	26 (36%)		6 (13%)	0 (0%)
Trichomonas vaginalis			0.753			1.000
No	88 (94%)	56 (92%)		43 (93%)	15 (94%)
Yes	6 (6.4%)	5 (8.2%)		3 (6.5%)	1 (6.3%)

^#p-values were calculated using the Wilcoxon rank sum test, Pearson’s chi-squared test, or Fisher’s exact test as appropriate values are shown as median (IQR) or n (%), as applicable.

Patterns of VSP use among adolescents and adult women

Distinct patterns of VSP use were observed between adolescent girls and adult women at the time of the exposure visit (Fig. 1). Overall, VSP use was more prevalent among adolescents, with 41% (78/188) reporting use compared to 25% (16/63) of adult women. Adolescents also reported using a broader range of products, with 26 distinct VSPs identified compared to 10 among adult women. Among VSP users, the most commonly reported products included alum and ibhodwe labafazi. Adolescents more frequently reported using alum, a compound consisting of aluminum sulfate and potassium sulfate, whereas adult women primarily reported using ibhodwe labafazi (translated from isiZulu as “ladies’ pot”), a pink scented petroleum jelly of unknown composition. Other commonly reported products among adolescents included snuff (smokeless tobacco made from dried tobacco leaves). Some participants reported using more than one product type. These products are typically informally manufactured and lack standardized composition, which may contribute to variability in their biological effects on the vaginal mucosa. A more detailed description of these products, including reported motivations for their use, has been published previously¹⁹.

Fig. 1.

Types and distribution of vaginally inserted product (VSP) use among women in rural KwaZulu-Natal. (A) Adolescent girls (n = 77); (B) adult women (n = 16). Doughnut charts illustrate the proportion of reported VSP types among users. Product categories are listed below each chart. Some participants reported using more than one product.

Comparison of cytokine levels and immune markers between baseline and exposure visits

We assessed whether mucosal immune parameters differed between baseline and post-exposure visits. No statistically significant differences in cytokine concentrations were observed between the two visits in either VSP users or non-users (Supplementary Fig S2–3). Similarly, immune cell activation marker expression did not differ between baseline and post-exposure visits (data not shown). Given the absence of detectable differences over time, analyses focused on the exposure visit, which was closest to the reported sexual exposure.

Cytokine concentrations in VSP users and non-users

At the exposure visit, cytokine concentrations were generally higher among VSP users compared with non-users across both adolescents and adults; however, most differences did not reach statistical significance (Fig. 2). Among adolescents, IL-1α concentrations were significantly higher in VSP users compared with non-users (p = 0.001; Fig. 2A). Among adults, MIP-1α concentrations were significantly higher in VSP users compared with non-users (p = 0.017; Fig. 2B). No statistically significant differences were observed when adolescents and adults were analysed together (Supplementary Fig S4). Following FDR correction for multiple comparisons (Benjamini-Hochberg method), none of the cytokine associations remained statistically significant.

Fig. 2.

Cytokine concentrations in VSP non-users and users at the exposure visit in adolescents and adults. (A) Adolescents: VSP non-users (n = 66) vs. users (n = 31). (B) Adults: VSP non-users (n = 55) vs. users (n = 18). Cytokine concentrations are presented on a logarithmic scale. Box-and-whisker plots represent the 5th–95th percentiles. Group comparisons were performed using the Mann–Whitney U test, with p < 0.05 considered statistically significant.

Cervical immune cell activation in VSP users versus non-users

Cytobrush samples were available for 251 participants, of which 66 (26%) met the predefined threshold of > 100 CD3 + events for flow cytometric analysis. At the exposure visit, no statistically significant differences were observed in the proportion of cervical immune cell activation markers between VSP users and non-users, except for a lower proportion of CD4 + T cells among VSP users (p = 0.040; Fig. 3A). No statistically significant differences were observed in HLA-DR/CD38 co-expression on CD4 + T cells or Th17-like cell populations. A non-significant trend toward higher HLA-DR expression was observed in VSP users (Fig. 3A–B). No statistically significant differences were observed when analyses were stratified by age group (data not shown).

Fig. 3.

Frequency of cellular marker expression in VSP users versus non-users at the exposure visit. Comparisons between VSP users and non-users (all age groups combined) were performed for: (A) CD4⁺ T cells and (B) Th17-like cells. The total sample included 48 non-users and 18 users. Box-and-whisker plots (showing minimum to maximum values) illustrate the distribution of activation marker and chemokine expression within each group. Statistical comparisons were performed using the Mann–Whitney U test, with p < 0.05 considered statistically significant.

Association between time since sexual intercourse and genital immune responses

We assessed associations between time since last sexual intercourse and mucosal immune markers in the combined cohort. Several cytokines showed weak inverse associations with time since intercourse, including MIP-1α (p = 0.029, r = − 0.2150), IL-1α (p = 0.047, r = − 0.1968), IL-1β (p = 0.024, r = − 0.2211), GM-CSF (p = 0.011, r = − 0.2476), and IL-17 (p = 0.008, r = − 0.2568). However, after FDR correction, these associations were no longer statistically significant (Table 2). No significant associations were observed for immune cell activation markers (Supplementary Table 2). When analyses were stratified by age group, no statistically significant associations were detected (data not shown).

Table 2.

Spearman correlation between inflammation and days since sexual intercourse.

Cytokine	Overall
	95% confidence interval		R	P	FDR corrected P value
	Lower	Upper
Log MCP-1	− 0.3174	0.0747	− 0.1263	0.203	0.261
Log IL-8	− 0.3432	0.0458	− 0.1547	0.118	0.261
Log IP10	− 0.3237	0.0676	− 0.1332	0.179	0.261
Log IFN-α	− 0.3072	0.0859	− 0.1152	0.246	0.295
Log MIP-1α	− 0.3971	− 0.0166	− 0.2150	0.029*	0.131
Log MIP-1β	− 0.3198	0.0720	− 0.1289	0.194	0.261
Log MIP-3α	− 0.1769	0.2210	0.0229	0.818	0.818
Log IL-1α	− 0.3819	0.0033	− 0.1968	0.047*	0.165
Log IL-6	− 0.3259	0.0652	− 0.1356	0.171	0.261
Log TRAIL	− 0.3352	0.0548	− 0.1458	0.141	0.261
Log IL-1β	− 0.4025	− 0.0230	− 0.2211	0.024*	0.131
Log TNF-α	− 0.3432	0.0458	− 0.1547	0.118	0.261
Log GCSF	− 0.3284	0.0624	− 0.1383	0.163	0.261
Log GMCSF	− 0.4258	− 0.0510	− 0.2476	0.011*	0.099
Log VEGF	− 0.2666	0.1297	− 0.0712	0.474	0.518
Log IL-17	− 0.4337	− 0.0607	− 0.2568	0.008*	0.099
Log IL-1rα	− 0.1341	0.2681	0.0698	0.489	0.518
Log IL-10	− 0.3746	0.0098	− 0.1896	0.055	0.165

Heatmap analysis of inflammatory cytokines in relation to HIV risk factors

A heatmap was used to visualise inflammatory profiles defined a priori as having ≥ 5 of 9 HIV-associated cytokines above the 75th percentile (Masson et al., 2015). No clear separation was observed between adolescents and adults based on inflammatory profiles. Variability in inflammatory status was observed across participants irrespective of age group. Visual patterns suggested higher inflammatory profiles among some VSP users and among participants sampled closer to sexual exposure (< 48 h); however, no formal clustering statistics were applied and these observations are descriptive (Fig. 4).

Fig. 4.

Heatmap of inflammatory cytokine concentrations in VSP non-users and users across all participants. This heatmap displays nine inflammatory cytokines (IL-1α, IL-1β, TNF-α, IL-6, IL-8, IP-10, MCP-1, MIP-1α, and MIP-1β) in VSP non-users and users, and by time since last sexual intercourse. Hierarchical clustering was performed using an unsupervised approach to organise participants and cytokines according to similarity in expression patterns. Expression values are presented as percentiles, with most values falling between the 50th and 75th percentiles. Data are visualised on a continuous scale to highlight relative differences between groups.

Predictors of inflammation status identified by weighted logistic regression

Weighted logistic regression was used to identify factors associated with inflammatory status, accounting for class imbalance in the outcome variable. The model included age group, bacterial vaginosis (BV), STI status, time since sexual intercourse, and VSP use. BV (OR 2.38; 95% CI: 1.18–4.95; p = 0.017) and recent sexual intercourse (OR 0.58; 95% CI: 0.38–0.86; p = 0.009) were statistically significant predictors of inflammatory status. VSP use was not significantly associated with inflammation status (OR 1.63; 95% CI: 0.79–3.38; p = 0.187), although the direction of the association suggested a potential trend (Table 3).

Table 3.

Factors associated with inflammation status using the weighted logistic regression.

Variable	Logistic (weighted)	p-value (weighted)
BV status (positive)	OR = 2.38 (1.18–4.95)	0.017*
Hours post-sex	OR = 0.58 (0.38–0.86)	0.009*
VSP use (yes)	OR = 1.63 (0.79–3.38)	0.187
Age (adult)	OR = 1.35 (0.66–2.79)	0.413
STI status (positive)	OR = 1.40 (0.55–3.59)	0.480

Discussion

Adolescent girls and young women (AGYW) in sub-Saharan Africa experience a disproportionate burden of HIV infection, driven by intersecting behavioural, biological, and social factors that may influence genital mucosal immunity. Despite this, post-coital immune responses remain insufficiently characterised, particularly in real-world contexts where sexual behaviour and culturally embedded practices such as VSPs may contribute to variation in mucosal immune profiles. In this study, we investigated mucosal immune responses following recent sexual intercourse and VSP exposure in adolescent and adult women in a high HIV prevalence setting. Distinct patterns of VSP use were observed between age groups. Adolescents reported a broader range of products, including alum and snuff, primarily used for perceived vaginal tightening and drying, whereas adult women more commonly reported ibhodwe labafazi for sexual enhancement. These findings are consistent with prior qualitative work describing culturally mediated motivations for VSP use in the region^16,19,20.

We first assessed whether mucosal immune parameters differed between baseline and post-exposure visits. At a group level, no significant differences in cytokine concentrations or immune cell activation markers were observed between visits in either VSP users or non-users. This suggests broadly stable mucosal immune profiles across sampling time points when exposure timing is not explicitly considered. However, variability in the timing of sampling relative to sexual exposure may have influenced these observations. Participants presenting shortly after intercourse may have exhibited transient inflammatory responses, whereas later sampling may have reflected partial resolution. Self-reported abstinence prior to baseline may also have introduced misclassification, potentially attenuating differences between time points. Together, these factors suggest that temporal heterogeneity may have reduced sensitivity to detect short-lived mucosal immune changes.

To further explore temporal dynamics, we examined immune markers in relation to time since last sexual intercourse. Recent unprotected sex was associated with elevated levels of pro-inflammatory cytokines, including MIP-1α, IL-1α, IL-1β, GM-CSF, and IL-17. These cytokines are involved in mucosal immune signalling and inflammatory responses²¹. The inverse associations between cytokine concentrations and time since intercourse are consistent with a transient post-coital inflammatory response, in line with studies describing early mucosal immune changes following coitus¹¹. These responses may reflect epithelial microtrauma and exposure to seminal components with immunostimulatory properties^22,23. Experimental studies suggest that seminal plasma can induce pro-inflammatory signalling in vaginal epithelial cells, contributing to cytokine production within hours of exposure^24,25. Collectively, these findings support a model in which sexual intercourse induces a short-lived mucosal inflammatory state that gradually resolves over time. Although cytokine changes were evident, immune cell activation markers showed limited association with time since intercourse. This may reflect the rapid resolution, redistribution, or compartmentalisation of cellular immune responses relative to soluble mediators. Overall, these findings suggest temporally distinct components of mucosal immune responses, with cytokines reflecting early inflammatory activity and cellular changes potentially reflecting later regulatory or repair-associated processes.

We next examined whether VSP use was associated with differences in mucosal immune responses following sexual exposure. In unadjusted analyses, VSP users exhibited higher cytokine levels than non-users, with IL-1α elevated in adolescents and MIP-1α elevated in adults. However, these associations did not remain statistically significant after correction for multiple testing and should therefore be interpreted cautiously. Although differences between specific product types (e.g., alum versus ibhodwe labafazi) may exist, sample sizes within subgroups were insufficient for stratified analyses, necessitating evaluation at an aggregated exposure level. To summarise overall inflammatory status, we used a composite score based on elevated pro-inflammatory cytokines as a pragmatic measure of mucosal immune activation relevant to HIV susceptibility. This approach does not capture more complex or bidirectional cytokine patterns, such as those observed in conditions such as BV. While multivariate immune signature modelling or bidirectional pattern analyses may better resolve this complexity, these approaches were beyond the scope of the present study and the primary focus remained on coordinated pro-inflammatory upregulation, recognising that immunoregulatory or suppressive responses may also be biologically relevant.

Using weighted logistic regression, recent sexual activity and bacterial vaginosis were identified as factors associated with mucosal inflammation. VSP use was not significantly associated with inflammation in the multivariable model, although a positive trend was observed. Overall, these findings suggest that recent sexual exposure and bacterial vaginosis are more consistently associated with mucosal immune activation in this cohort, while VSP use may contribute to inter-individual variability without independently driving inflammation after adjustment.

This study has several limitations. Variability in the timing of sampling following sexual intercourse likely introduced heterogeneity that reduced sensitivity to detect transient immune changes. In addition, reliance on self-reported abstinence prior to baseline may have introduced misclassification, potentially attenuating observed differences between time points. The relatively small sample size further limited statistical power, particularly in subgroup and product-specific analyses.

In conclusion, this study provides insight into mucosal immune dynamics following sexual intercourse and VSP exposure in adolescent and adult women in a high HIV prevalence setting. We demonstrate a transient increase in pro-inflammatory cytokines following recent unprotected sex, consistent with a short-lived period of mucosal immune activation. While VSP use was associated with higher cytokine levels in unadjusted analyses, these associations were not robust after correction for multiple testing. Overall, our findings highlight the importance of exposure timing in interpreting mucosal immune responses and suggest that recent sexual activity is a key determinant of transient genital inflammation. Further longitudinal studies with precise temporal sampling and larger, stratified cohorts are needed to better define the dynamics of post-coital immune responses and the potential contribution of culturally embedded practices such as VSP use.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

All authors contributed significantly to this study and met the criteria for authorship. LM, J-AP, HJ, Q-AK, SN, RB, and PM conceived the study. SR and PM determined the methodology used. SR, PR, ZG, RT, MM, NM, HH, DP, AM, GM, and NS collected data. Then, SR and PM analyzed the data; LL, ML, BS, LM, J-AP, HJ, RB, and SN interpreted the data. The original draft of the manuscript was written by SR, and all authors reviewed and approved the final manuscript.

Funding

This study was supported by funding from the National Institute of Health through HJ and J-AP (R01Al128792).

Data availability

Supplementary material is available for the original contributions of this study. Raw data supporting the conclusions of this study are available upon request from the corresponding author. The data were not publicly available because of privacy concerns.

Declarations

Competing interests

The authors declare no competing interests.