Cervical cancer is one of the most common female malignancies in the world, ranking at the top among all female malignancies (1). Its incidence and mortality are still at a high level, and especially in developing countries, it poses a prominent threat to women’s health (2, 3). Persistent high-risk human papillomavirus (HR-HPV) infection is a major cause of cervical cancer initiation and progression, detectable in most cases (4–6). However, HR-HPV is not the only driver for the initiation and development of cervical cancer (7–10). Bacterial sexually transmitted pathogens in the reproductive tract may serve as vital co-risk factors, further leading to the initiation and development of cervical intraepithelial neoplasia and even cervical cancer through co-infection and interaction with HR-HPV (7–10). Of note, our previous research on patients in the gynecological outpatient clinic found that Chlamydia trachomatis and Ureaplasma urealyticum were detected more frequently in HPV-positive cervical samples, suggesting a certain tendency for co-infection with HPV (11).

HR-HPV, C. trachomatis, U. urealyticum, and Neisseria gonorrhoeae are pathogens of sexually transmitted infections (12–15). One of the important characteristics of sexually transmitted infections is its significant concealment, which can colonize at a low level for a long time or persist in the host without typical symptoms, thereby causing delayed diagnosis and treatment as well as increased risk of sustained transmission and difficulty of clinical identification without being easily detected (8, 16–18). Occult infections not only have adverse effects on individual reproductive health but also exacerbate the disease burden and complexity of prevention and control at the population level, which emerges as an urgent issue of concern in the field of global public health (8, 15, 19, 20). Previous studies on STIs have mainly focused on the outpatient population of Grade-A Tertiary Hospitals in cities (15, 21, 22). Our preliminary survey on 2,389 symptomatic women in Haikou City showed that the age group of 15–25 years is at high risk of infection, and mixed infections are correlated with an increased risk of HPV (11). However, such data based on passive hospital visits inevitably suffers from selection bias, making it difficult to represent the asymptomatic population in rural areas who may carry pathogens for a long time, thus potentially underestimating the true infection burden and co-infection patterns of this “occult infection” population. Especially in developing countries and rural underserved communities, STIs are often not adequately diagnosed and effectively managed due to limited access to healthcare services, socio-economic disparities, and cultural barriers, with the disease burden mainly borne by women in these areas (23–26). This urban–rural gap may be further driven by disparities in healthcare infrastructure and accessibility, as well as differences in screening coverage and testing strategies implemented in public health centers and primary healthcare facilities (27, 28).

Existing evidence suggests a possible synergistic promoting effect of C. trachomatis and U. urealyticum on the progression of cervical lesions (8, 13, 29). The inflammatory microenvironment induced by these pathogens may weaken local immune surveillance, providing favorable conditions for sustained HR-HPV infection and the progression of related lesions (8, 30–32). However, the specific roles exerted by C. trachomatis and U. urealyticum during this procedure are still controversial, and there is no consensus on whether they have direct carcinogenic potential (32–34). In light of this, this study aims to provide more direct pathological evidence to elucidate the correlations of C. trachomatis and U. urealyticum infection with the severity of cervical intraepithelial neoplasia.

From a geographical perspective, Hainan Island is the only tropical island province in China, which is geographically relatively independent, with certain natural environment and socioecological differences and isolation effects as compared to inland areas, and its pathogen epidemic characteristics may have regional specificity distinct from inland areas (35–37). Hainan is also a highly connected hub for tourism and population mobility, with frequent inflow and outflow through ports and transport corridors that may facilitate pathogen introduction and transmission. Such sustained mixing driven by travel and commerce may further amplify cross regional introduction and onward transmission of sexually transmitted pathogens. Complementing this mobility-driven pressure, the unique tropical ecological environment and high biodiversity of Hainan Island may provide a more unique ecological background for the transmission, colonization, and sustained prevalence of multiple pathogens (38–40). Studies in this context can help supplement existing evidence from a regional and ecological perspective and provide a scientific basis for the development and implementation of more targeted cervical health management strategies that integrate multiple pathogen screening in this region.

This study focuses on the asymptomatic female screening population in rural areas of Hainan Island, aiming to answer two key questions: firstly, to systematically characterize the prevalence and co-infection patterns of major STIs, and assess the impact of sociodemographic factors, such as age and educational attainment, as well as geographic heterogeneity, on the distribution of infection; secondly, to evaluate the association between specific STIs co-infection and the risk of significant cytological abnormalities and cervical precancerous lesions. Altogether, this study fills the evidence gap at the urban–rural and regional levels and improves the epidemiological landscape of related reproductive tract infections.

The retrospective analysis of rural women in Hainan Island provided insights into the epidemiological landscape of HR-HPV and other common bacterial sexually transmitted pathogens of the reproductive tract in a tropical island region. Among the 1,342 female participants, HR-HPV-positive individuals were older and had lower levels of education, with significant differences in infection rates across different cities and counties. Among the bacterial sexually transmitted pathogens detected in the reproductive tract, U. urealyticum showed the highest detection rate, followed by C. trachomatis. Meanwhile, in viral sexually transmitted pathogens, the majority of HR-HPV-positive individuals were infected with a single genotype; in the co-infection patterns of sexually transmitted pathogens, the co-infection of U. urealyticum and HR-HPV52/58 was the most common. It should be noted that in this study, C. trachomatis was linked to ASC-US, while U. urealyticum infection correlated with an enhanced risk of CIN1. These results suggest that co-infection of HR-HPV with C. trachomatis and U. urealyticum may synergistically affect the local inflammatory response and immune microenvironment, thereby correlating with the risk of occurrence and progression of early cervical lesions.

This retrospective analysis revealed that lower levels of education and residence in rural areas were related to higher detection rates of co-infection with HR-HPV and C. trachomatis, U. urealyticum, and N. gonorrhoeae. This finding is consistent with previous research, which suggests that highly educated individuals generally have better compliance with cervical cancer screening and are more inclined to adopt protective sexual activities (48–50). The results of this research suggested that females with limited educational opportunities may have a longer time interval between infection and diagnosis, allowing gradual accumulative co-infections. Geographical clustering also affects the distribution of infections, and economic inequality and inadequate access to medical services may contribute to the sustained transmission of HR-HPV and C. trachomatis, U. urealyticum, and N. gonorrhoeae in resource-limited areas.

Additionally, we observed a significantly higher average age of HR-HPV-positive females as compared to HR-HPV-negative females (49.01 years vs. 46.45 years), indicating that persistent HR-HPV infection may have an age-dependent cumulative increase. This finding is consistent with the “bimodal” age distribution of HPV infection reported in previous population studies: HPV infection exhibits a high incidence in young females and then decreases with age, with a second peak around 50–60 years; this phenomenon may be correlated with factors such as birth cohort effect, hormonal fluctuations during perimenopause, immunosenescence, and reactivation of latent HPV (51–54). On the contrary, we observed that the infections of C. trachomatis, U. urealyticum, and N. gonorrhoeae generally decreased with age.

The phenomenon mentioned above may be relevant to the current urban sexual health education and HPV vaccination strategies, mainly covering adolescents and people of early reproductive age; in contrast, a considerable proportion of females in rural areas were born before the “vaccine era,” and they are currently in the high-risk age window for cervical cancer and its precancerous lesions. The middle-aged and older adult population had limited access to standardized prevention services in the past, coupled with the hidden nature of infections and insufficient medical accessibility, which may lead to a long-term underestimation of their reproductive tract infection burden and related risks, thus becoming a weak link and “blind spot” of risk in the grassroots public health service system (55–57). Therefore, it is necessary to further tilt public health policy formulation and resource allocation towards middle-aged and older females in rural areas with scarce resources, making them the key target audience of cervical health management; in the meantime, the screening strategy should gradually shift from “opportunistic screening” triggered by medical visits to “full-coverage screening” based on population management, and improve the accessibility and efficiency of early diagnosis and treatment of precancerous lesions through standardized follow-up and stratified intervention, thereby reducing the long-term burden of cervical diseases in rural areas.

In this study cohort, the total detection rates of C. trachomatis, U. urealyticum, and N. gonorrhoeae were 3.20, 45.75, and 0.15%, respectively. The infection rates of C. trachomatis and U. urealyticum were remarkably higher in HR-HPV-positive samples than in HR-HPV-negative samples (p < 0.01), which was consistent with our previous research findings on gynecological outpatient patients (11). However, it is noteworthy that this study focuses on the asymptomatic female population, and the incidence of U. urealyticum remains strikingly high in this population. This phenomenon reminds us widespread occult transmission of U. urealyticum among “healthy” populations, and these asymptomatic carriers may have long been overlooked as potential sources of infection. The traditional detection and therapeutic strategies are only for symptomatic populations and may have limitations. We need to re-examine the prevention and control strategies against reproductive tract pathogens, strengthen public health education, and correct the misconception that “no symptoms mean no infection”. In contrast, the infection rate of N. gonorrhoeae was extremely low, basically consistent with the overall level shown by domestic surveillance data, but lower than 1–2% reported in some studies on the urban population (58, 59). This may also be related to the inclusion of a healthy female population in this study. Generally speaking, N. gonorrhoeae infection is often accompanied by more obvious clinical symptoms, and such patients are more likely to seek medical attention in medical institutions rather than “health screening”. Therefore, the infection rate of N. gonorrhoeae was relatively low in the healthy female population surveyed in this study.

In HR-HPV-positive individuals, U. urealyticum exhibited more prominent co-infection clustering characteristics compared to C. trachomatis, indicating a closer co-occurrence relationship with HR-HPV. Therefore, we incorporated the UpSet graph and co-occurrence chord diagram for conjoint analysis to characterize the ecological niche and topological characteristics of U. urealyticum in the HR-HPV co-infection network. The results displayed that HR-HPV52 and HR-HPV58 were dominant genotypes in rural populations of Hainan Island, forming the main co-infection pattern together with U. urealyticum, suggesting that these two may have specific biological synergistic effect, that is, U. urealyticum infection may provide “boosting” conditions for the invasion, colonization, or persistence of specific HR-HPV subtypes by altering the reproductive tract microenvironment, thereby performing a potential synergistic promoting role in the viral persistence and its pathogenic process. Therefore, U. urealyticum, as a key hub in the HR-HPV infection network, has important clinical guidance significance. In clinical practice, the detection of U. urealyticum should be regarded as not only a single pathogen infection but also as a “sentinel signal,” indicating that the patient may be in a complicated multi-pathogen exposure environment or a state of microecologic dysbiosis, thus requiring heightened alertness of clinical doctors and comprehensive intervention.

In this regard, we probed into the clinical relationship between U. urealyticum and cervical epithelial lesions. It has been reported that co-infection with U. urealyticum considerably raises the incidence of high-grade squamous intraepithelial neoplasia and invasive cervical cancer (8). A cross-sectional study in Brazil demonstrated that Ureaplasma parvum-positive females were significantly linked to the incidence of CIN1, and its co-infection with HR-HPV further augments the risk of CIN1 (8, 14). This study unveiled that U. urealyticum infection was associated with an increased risk of CIN1 (aOR = 1.79). U. urealyticum infection is often manifested with occult and chronic colonization characteristics, which rarely cause typical acute symptoms, making it more likely to continuously shape the local cervical microenvironment over a long period. Its long-term existence may interfere with epithelial cell homeostasis and repair processes through low-grade but sustained inflammatory stimulation and immune microenvironment remodeling, which further affects the fine regulation of cell cycle-related pathways, thereby providing a “soil” for the occurrence and maintenance of mild epithelial abnormalities. Therefore, from a pathological perspective, U. urealyticum appears to be more closely associated with low-grade lesions, typified by CIN1. In HR-HPV-positive individuals, this coinfection may be linked to the maintenance of pathological states characteristic of early-stage lesions. Consequently, we speculate that U. urealyticum might play a synergistic role in the progression of HR-HPV-associated cervical disease. Additionally, we did not observe a significant association between U. urealyticum infection and the risk of high-grade lesions or invasive cervical cancer. This result may be limited by both the research population and sample size. Firstly, this cohort mainly covered healthy individuals, who have a lower baseline probability of developing high-grade lesions than patients with clinical symptoms. Secondly, the number of high-grade lesions and cancer cases in the cohort was relatively small, resulting in insufficient statistical power and making it difficult to detect possible moderate or weak correlations.

Of note, although the overall positive rate of C. trachomatis is low (3.20%), it is more strongly linked to cytological abnormalities. Multivariable analysis suggested a significant correlation between C. trachomatis infection and ASC-US (aOR = 2.82). C. trachomatis infection often triggers relatively acute inflammatory responses, which can cause transient reactive changes and increased morphological heterogeneity in cervical epithelial cells. This type of inflammation-associated “reactive atypia” is more easily categorized into ASC-US in TCT detection and evaluation from the perspective of cytology. Furthermore, C. trachomatis infection may indirectly promote the susceptibility and persistent infection of HPV through the pathological “inflammation-barrier breakdown-immune microenvironment remodeling” axis, leading to a higher incidence of mild abnormalities (ASC-US) in cytology (31, 32, 60). Specifically, C. trachomatis-induced cervical inflammation weakens the integrity of the mucosal epithelial barrier and alters the local innate and adaptive immune status, creating favorable conditions for the colonization, replication, and sustained existence of HR-HPV (9, 31, 32, 34). While the discriminatory performance of C. trachomatis or U. urealyticum infection for cervical lesions is limited in the overall HR-HPV-positive population, co-testing significantly enhances predictive efficacy within the specific subgroup of HR-HPV 52/58 positive individuals. However, that given the overall low-to-moderate AUC values, these markers are best utilized as adjunctive tools within existing triage strategies. They aim to refine risk stratification rather than serving as independent decision-making criteria. Crucially, from the perspective of large-scale population screening, even this moderate statistical improvement translates into significant public health benefits. By refining risk stratification, this co-testing approach enables more precise resource allocation in settings with limited resources. For instance, regarding high-risk subgroups such as those co-infected with U. urealyticum and HPV52/58, clinical management can be optimized by intensifying surveillance through increasing the frequency of follow-up visits or shortening the intervals for colposcopy referrals. This strategy thereby maximizes the detection of early-stage lesions while avoiding over-management of low-risk populations.

Therefore, we recommend incorporating combined testing for C. trachomatis and U. urealyticum into the risk management workflow for HR-HPV 52/58 positive individuals, particularly those presenting with cytological abnormalities, persistent infection, or other high-risk factors. For patients with identified co-infections, intensified surveillance (e.g., shortened follow-up intervals) should be implemented, with prioritized referral for colposcopy where resources permit. Conversely, for those without co-infection and at lower overall risk, a relatively conservative management strategy may be adopted alongside standard follow-up. This stratified management approach aids in avoiding unnecessary overtreatment and alleviating patient psychological burden, thereby advancing cervical cancer screening toward precision and personalization.

Despite its significant public health implications, this study has certain limitations. Regarding the construction of the control group, due to cost-effectiveness considerations associated with multi-pathogen molecular screening in a large-scale population, we did not perform universal testing on the entire HR-HPV-negative cohort. Instead, we employed a random sampling strategy with an approximate 1.8:1 case-to-control ratio. However, a priori sample size estimation based on observed effect sizes confirmed that the actual sample size (n = 1,342) far exceeded the theoretical minimum (n = 420) required to satisfy statistical testing requirements. Further post-hoc power analyses revealed that the statistical power for the primary pathogens, C. trachomatis and U. urealyticum, approached 1.0 (>0.99). This suggests a negligible probability of Type II errors (false negatives) even with the current sample size, ensuring the high robustness of our conclusions. Although statistical power for N. gonorrhoeae was limited due to its low prevalence, exact probability tests (Fisher’s exact test) were applied to ensure the accuracy of statistical inferences. The control group was an unbiased subset selected from the same source cohort based on strict random sampling principles. Given the relative stability of baseline infection rates for genital pathogens in the general population, the current control sample size was sufficient to establish a reliable baseline reference. Furthermore, we acknowledge the observed age disparity between the case and control groups (49.01 ± 8.68 vs. 46.45 ± 8.23 years, p < 0.001). While this variance was rigorously controlled for as a covariate to eliminate confounding effects across all key analyses, including inter-group comparative assessments, multivariate logistic regression, and ROC predictive modeling, future investigations would benefit from implementing stricter age-matching designs during the recruitment phase to further minimize potential residual confounding. Beyond these design considerations, the specific molecular mechanisms by which C. trachomatis or U. urealyticum, together with HR-HPV, synergistically expedite the occurrence and progression of cervical cancer have not been fully elucidated. Longitudinal cohort studies and in vitro/in vivo functional experiments are necessary for further verification of their causal relationship and interpretation of potential mechanistic pathways. In addition, the limited number of high-level lesions and cancer cases restricts the analysis of associations with more severe diseases.

Another important consideration is that several potentially relevant determinants of co-infection were not captured in the current dataset and may have contributed to the observed heterogeneity in co-infection rates. Information on sexual risk-related behaviors, including the number of sexual partners and condom use, was not routinely collected during cervical cancer screening registration and therefore could not be incorporated into our analyses. Because these factors may influence population exposure patterns and co-infection transmission dynamics, future screening registration and follow-up systems should incorporate standardized behavioral data collection to better inform targeted public health interventions.

Moreover, our virological assessment primarily focused on HR-HPV, precluding evaluation of the potential contribution of LR-HPV to infection dynamics and lesion trajectories. Prior studies have suggested that, compared with HR-HPV infection alone, women with concomitant LR-HPV detection may have a lower subsequent risk of invasive cervical cancer, raising the possibility that LR-HPV co-infection could partially delay or attenuate HR-HPV-driven carcinogenic progression (61). At the molecular level, E2 proteins from both HR-HPV and LR-HPV types have been reported to broadly suppress innate antiviral pathways, including RIG-I/MDA5-MAVS, TLR3-TRIF, cGAS-STING, and JAK–STAT signaling, thereby facilitating viral persistence (62). However, existing evidence largely derives from studies of individual HPV types, and direct data demonstrating clear synergistic or antagonistic effects of HR-HPV/LR-HPV co-infection on immune evasion remain limited.

Taken together, future work integrating LR-HPV genotyping with prospective follow-up and mechanistic profiling of molecular and immune features is warranted to clarify whether LR-HPV co-detection modifies HR-HPV persistence and its associations with cervical epithelial lesions. At the population level, such studies could also compare HPV co-infection patterns between clinic-based populations and the general population to better characterize differences associated with healthcare-seeking contexts. Concurrently, metagenomic sequencing of cervicovaginal samples may further elucidate microbial community structure and functional pathways, providing deeper insights into the underlying mechanisms.

Among rural female populations in Hainan Island, a tropical island region, STIs infection shows the characteristic of a public health issue intertwined with multiple factors, and its epidemic distribution is jointly shaped by demographic characteristics, behavioral factors, and regional differences. In this population, U. urealyticum is the main prevalent bacterial sexually transmitted pathogen and exhibits a directional co-infection pattern of U. urealyticum plus HR-HPV52/58. More importantly, C. trachomatis and U. urealyticum may participate in the occurrence and evolution of early cervical lesions through biological pathways, thus playing different roles in the natural history of the disease. Hence, in the future, the formulation of cervical disease prevention and control policies and resource allocation will further tilt towards healthy female populations and rural areas with relatively scarce resources, rendering them as key target audience of cervical health management. More importantly, it is necessary to go beyond the single perspective of “only focusing on HPV” and further integrate joint screening and stratified intervention for the key reproductive tract pathogens on the basis of optimizing HR-HPV screening strategies, so as to enhance the sensitivity of risk identification and improve early targeted management, achieving more efficient and sustainable cervical health protection.