Association between Trichomonas vaginalis infection and spontaneous preterm labour in Lagos, Nigeria: an analytical cross-sectional study

Abstract

Background

Preterm labour (PTL) leading to preterm birth is a grave complication of pregnancy that attracts long-term medical and social sequelae. However, there is still a paucity of data on the impact of Trichomonas vaginalis infection on PTL especially among women in sub-Saharan Africa.

Objective

This study determined the association between T. vaginalis infection and PTL among pregnant women in Lagos, Southwest, Nigeria using a molecular-based rapid antigen detection technique to diagnose T. vaginalis infection.

Methods

This was a multicenter analytical cross-sectional study carried out among parturients managed in the labour ward units of three hospitals in Lagos, Southwest Nigeria between April and December 2019. High vaginal swab samples were collected from n = 105 pregnant women with PTL and an equal number of women who had term labour between April and December 2019. T. vaginalis Antigen Rapid test was used for the detection of T. vaginalis in the vaginal samples. Multivariable binary logistic regression analyses were used to control for all potential confounders in the association between T. vaginalis infection and PTL. Statistical significance was reported at p < 0.05.

Results

The prevalence of T. vaginalis infection was significantly higher in women with PTL compared to those with term labour (12.4% vs. 2.9%, p = 0.009). On adjustments for age, parity, booking and educational status, abnormal vaginal discharge and occurrence of prelabour rupture of membranes, there was no association between T. vaginalis infection and PTL (adjusted Odds ratio = 0.94, 95% CI: 0.10–9.29).

Conclusions

The study showed no independent association between T. vaginalis and PTL. We recommend future adequately powered longitudinal studies with consideration for other potential confounding factors, such as sexual lifestyle and STIs, to further explore any potential relationship between T. vaginalis and PTL.

Introduction

Preterm labour (PTL) is the spontaneous onset of uterine contractions with progressive cervical changes and descent of the presenting parts after the threshold of fetal viability, generally considered to be 22–24 weeks of gestation, and before 37 completed weeks of gestation [1]. PTL leading to preterm birth is a grave complication of pregnancy that attracts long-term medical and social sequelae such as an increased risk of neurodevelopmental impairments, respiratory complications, cardiovascular diseases, and metabolic disorders [2,3,4]. Additionally, it places a significant burden on healthcare systems due to prolonged neonatal intensive care unit stays and long-term medical follow-up [3].

Preterm birth accounts for 40–60% of all perinatal deaths and 50% of long-term neurological impairment in under-five children [5, 6]. Of the more than 13 million babies born prematurely in 2020, nearly one million die due to complications related to preterm birth [3] with over 60% of these occurring in sub-Saharan Africa and South Asia [2, 7]. The burden of preterm birth remains high in many low- and middle-income countries, with recent estimates suggesting an increasing trend [3, 8]. In Nigeria, the rate of preterm birth is estimated at 12.2 per 100 live births [2], and data indicate that the incidence has been rising over the past three decades [9]. Several risk factors have been linked to spontaneous PTL and preterm births [7, 10] among which include genitourinary tract infections such as Bacterial vaginosis and Trichomoniasis [1, 2, 10].

Trichomoniasis, caused by Trichomonas vaginalis, is the most prevalent, non-viral sexually transmitted infection [11] affecting an estimated 143 million people worldwide [12]. In Nigeria, the prevalence ranges from 2.8% in Benin City [13], 15.4% in Lagos [14] to 11.0% in Maiduguri [15]. These variations in prevalence may be attributed to differences in study populations, diagnostic methods, sample sizes, and geographic or socio-demographic factors, including sexual behaviour, healthcare access, and screening practices. T. vaginalis is a unicellular, flagellated, microaerophilic, parasitic protozoa that primarily infects the squamous epithelium of the genital tract [16, 17]. Its microaerophilic nature allows it to thrive in the low-oxygen environment of the genital tract, facilitating its persistence and transmission [16]. It is transmitted among humans primarily by sexual intercourse [17] and has an incubation time that is generally between 4 and 28 days. However, the infection may persist for possibly months or years in some women [17]. T. vaginalis infection is asymptomatic in about 40–80% of women [18] and when symptomatic, it is associated with symptoms such as frothy, foul smelling yellowish/grey-green vaginal discharge, vulvovaginal irritation, dysuria and lower abdominal pain [11, 17, 18].

Based on findings from observational studies and meta-analyses, T. vaginalis infections in pregnancy have been associated with some adverse outcomes including prelabour rupture of membranes (PROM), preterm delivery, low birth weight, and perinatal morbidity [17,18,19]. A 2014 systematic review and meta-analysis estimated that women with T. vaginalis infections were 1.4 times more likely to have preterm birth compared to women without infection [19]. Perinatal transmission of T. vaginalis infection may result in vaginal and respiratory infections in neonates [17, 20] and may also increase the risk of HIV acquisition and transmission, especially in HIV-endemic regions [19, 21, 22]. This increased risk, the extent of which may vary by population and HIV prevalence, is thought to be mediated by inflammation and mucosal disruption, which facilitate viral entry [23, 24]. While some studies in other settings have suggested a possible impact of T. vaginalis infection on PTL [1, 11, 17, 18, 25, 26], limited research has specifically examined this relationship within sub-Saharan Africa.

This underscores the need for region-specific investigations to better understand the potential role of T. vaginalis in PTL using a molecular-based rapid antigen detection technique for T. vaginalis. This newer molecular-based diagnostic assay is faster, cheaper, and can eliminate the need for highly skilled microbiologists compared to using the centralized laboratory-based traditional wet mount microscopy and culture techniques [27]. They also demonstrate better diagnostic accuracy, with a sensitivity of 87–98% and a specificity of 99–100%, whereas wet mount microscopy has a significantly lower sensitivity of 50–70% and culture techniques range from 44 to 75% sensitivity [17, 28, 29]. This analytical cross-sectional study, therefore, tested the hypothesis that genital infection with T. vaginalis impacts the occurrence of PTL by determining the association between T. vaginalis infection and PTL among pregnant women in Lagos, Southwest, Nigeria.

Materials and methods

Study design and settings

This was a multicenter analytical cross-sectional study carried out among parturients managed in the labour ward unit of three hospitals in Lagos, Southwest Nigeria between April and December 2021. The hospitals were Lagos University Teaching Hospital (LUTH) Idi-Araba, Lagos Island Maternity Hospital (LIMH) Lagos Island, and Maternal and Child Centre (MCC) Surulere. These hospitals capture a diverse patient population across different levels of healthcare delivery and reflect the spectrum of maternal healthcare services in Lagos. LUTH is a tertiary referral center that represents high-risk obstetric cases and specialised maternal care. LIMH is a secondary-level facility that serves as a major maternity referral center within the state, while MCC Surulere is a comprehensive maternal and child health facility which provides obstetric care to the large urban population of Surulere and its adjoining Local Government Areas.

Study population and eligibility criteria

The study population comprised consecutively consenting pregnant women aged 15 to 49 years old with preterm labour (PTL) and their concurrently enrolled comparison counterparts with term labour at the study sites. Women with PTL were those who had spontaneous onset of uterine contractions with progressive cervical changes and descent of the presenting parts at or after 28 weeks (age of fetal viability) and before 37 completed weeks of gestation [1]. The threshold for fetal viability aligns with existing clinical definitions, as fetal viability is generally considered to begin around 28 weeks in many low-resource settings including Nigeria [30, 31]. Additionally, outcomes for neonates born before this gestational age are significantly poorer, with higher rates of morbidity and mortality [31]. Gestational age was determined using the last menstrual period and validated by ultrasonography before 20 weeks gestational age. Inclusion criteria were women with singleton gestation with sonographic confirmation of gestational age and cervical dilatation ≥ 4 cm. Excluded from the study were women with systemic diseases like hypertension, diabetes mellitus, cardiopulmonary disease, renal disease, sickle cell disease, those with multiple gestations and pre-existing reproductive tract abnormalities, and those who used antibiotics in the two weeks preceding enrolment.

Sampling and participants’ enrolment

The minimum sample size [32] for each group in the study (n = 99) was calculated using an odd ratio of 2.38 derived from the study by Hosny et al. [1] to achieve a power of 80% (0.842) at a type 1 error rate of 5% (1.96) and 95% confidence level. With further adjustment for a non-response or data-recording error rate of 5%, a total sample size of n = 105 will be enrolled into each of the preterm (PTL) and term labour groups. All the potentially eligible women who presented to the labour ward units of the participating hospitals during the study period were screened for eligibility by the study investigators and then enrolled by consecutive sampling technique until the sample size required for the two study groups was achieved. To minimize potential selection bias associated with consecutive sampling, recruitment was conducted continuously throughout the study period, spanning different days and times to include a diverse patient population. Additionally, participants were enrolled across the three study sites to enhance representativeness and reduce site-specific biases.

Participants’ data collection

Data was collected using a structured pretested questionnaire administered by the study investigators comprising one clinician (obstetrician) and a trained research assistant in each of the participating hospitals to collect sociodemographic information (age in years, parity, booking, occupation, marital, and educational status) and clinical information (presence of abnormal discharge and PROM). The questionnaire was pre-tested among a small sample of participants (n = 10) to ensure clarity and reliability before full implementation in the study. In the context of this study, “booking” refers to the initial antenatal care visit, during which a comprehensive assessment of the pregnant woman’s medical, obstetric, and social history is conducted. Following data collection, permission was obtained from each woman for a pelvic examination and a sterile bivalve speculum was introduced into the vagina to expose the cervix and a high vaginal swab was obtained from the posterior fornix of the vagina using sterile cotton swabs.

Laboratory analysis

T. vaginalis infection was diagnosed by detection of its antigen in a vaginal swab using the Strong Step^® Trichomonas vaginalis Antigen Rapid Test by Liming Bio-Products Co., Ltd Jiangsu China according to the kit’s instructions. The Trichomonas Rapid Test utilizes colour immunochromatographic, capillary flow, and “dipstick” technology. The test was performed at the bedside immediately after sample collection in the labour ward by an investigator or trained research assistant. Twenty drops (0.5 ml) of extraction buffer were added into a flexible extraction tube, the vaginal swab sample was dipped into the tube and mixed vigorously and then soaked for one minute in the buffer. The swab was removed, discarded appropriately, and the extraction tube closed. The test cassette was then removed from its sealed pouch and placed on a level surface. Three drops of the extracted sample in the extraction tube were dropped in the sample well on the test cassette and the result read after 15 min. A blue test line and a red control line indicated a positive Trichomonas antigen detection while a red control line without a blue test line was interpreted as a negative result. To ensure accuracy and reliability of results, all test kits were stored and used according to the manufacturer’s guidelines. Positive and negative control samples provided by the manufacturer were used periodically to confirm the validity of the test kits. Additionally, duplicate testing was performed on 10% of randomly selected samples as an internal quality control measure. Any discordant results were further reviewed by a senior Microbiologist to ensure consistency.

Statistical analysis

Data from the questionnaire were transferred into an Excel spreadsheet and then inputted into IBM SPSS version 29.0 software for Windows (Armonk, NY: IBM Corp) for analysis. The data were summarised, and categorical variables were presented as frequencies and percentages while continuous variables were presented as means (± standard deviation). Pearson’s chi-square test or Fisher’s exact test, where appropriate, was used to determine the association between categorical variables and PTL. We selected reference categories for the variables based on standard epidemiological practices and clinical relevance to ensure meaningful comparisons. The reference age category (< 35 years) was selected because younger maternal age is generally associated with lower risks of adverse pregnancy outcomes compared to advanced maternal age (≥ 35 years). Multiparity was chosen as the reference group for parity as nulliparity is typically associated with an increased risk of adverse pregnancy outcomes, including PTL. Booked participants were used as the reference antenatal booking group as they are expected to receive standard antenatal care, which reduces the likelihood of PTL, this allowing for a meaningful comparison with unbooked participants. “Less than tertiary education” was chosen as the reference category as a lower level of education is often linked to poorer health outcomes, providing a basis for assessing the impact of higher education on PTL. Skilled/professional workers were used as the reference occupation group since they are more likely to have better socioeconomic status and access to healthcare than unskilled workers. Women without vaginal discharge were selected as the reference group to evaluate whether the presence of vaginal discharge increases the risk of PTL. Women without PROM were used as the reference group to assess the impact of PROM on PTL, given its strong association with preterm birth. A multivariable logistic regression model with a backward conditional approach was developed to obtain the odds ratio (OR) at a 95% confidence interval (CI), with adjustments made for possible confounders. The backward conditional method allows for the initial inclusion of all potential confounders and the systematic removal of non-significant variables, ensuring a parsimonious model while retaining the key confounding variables in the final model. Confounders were selected based on a p-value < 0.10 in the preceding bivariable model, as well as their established relevance in prior literature [1, 17, 25], and clinical judgment. The variables included in the final multivariable model included participants’ parity, booking and educational status, presence of abnormal vaginal discharge, and PROM. We assessed the calibration or goodness-of-fit statistic of the final model using the Hosmer-Lemeshow test, while the model’s explanatory power was reported with Nagelkerke’s pseudo-R². A stratified/interaction analysis was conducted to reveal possible interaction effects of identified confounding factors (including unbooked antenatal status, presence of abnormal vaginal discharge, and PROM) on the association between T. vaginalis infection and PTL. Interaction terms were included in the regression models, and statistical significance was assessed using likelihood ratio tests. Statistical significance was set at p < 0.05.

Results

As shown in Table 1, the mean age of the participants with preterm labour was statistically different from that of their comparison group with term labour (27.6 versus 30.2 years). There were also statistically significant differences in parity, booking status, marital status, abnormal vaginal discharge and occurrence of prelabour rupture of membranes (PROM). There were, however, no differences in the educational and occupational status, religion and tribe of women in both the preterm and term labour groups.

Table 1 Sociodemographic and clinical characteristics of study participants

The prevalence of T. vaginalis infection was higher among women with preterm labour (n = 13) than in their counterparts with term labour (n = 3) (12.4% versus 2.9%; P = 0.009) as shown in Fig. 1.

Fig. 1

Clustered bar chart showing the proportion of women with T. vaginalis infection in the preterm and term participants group (12.4% versus 2.9%; P = 0.009)

As shown in Fig. 2, there was no significant difference in the prevalence of T. vaginalis infection in the different gestational age categories of women with spontaneous preterm labour (P = 0.117).

Fig. 2

Clustered bar chart showing the distribution of T. vaginalis infection in different gestational age categories among women with preterm labour (P = 0.117)

After adjusting for age, parity, booking and educational status, abnormal vaginal discharge and PROM in the multivariable model, there was no statistically significant association between T. vaginalis infection and PTL (adjusted Odds ratio = 0.94, 95% CI: 0.10–9.29). However, there were significant associations between being unbooked (adjusted Odds ratio = 60.81, 95% CI: 20.42–181.08), having abnormal vaginal discharge (adjusted Odds ratio = 4.55, 95% CI: 1.72–12.05), and PROM (adjusted Odds ratio = 60.81, 95% CI: 15.99–454.42) and experiencing spontaneous PTL [Table 2]. Stratified/interaction analyses showed no interaction effect of identified confounding factors (including unbooked antenatal status, presence of abnormal vaginal discharge, and PROM) on the association between T. vaginalis infection and PTL, indicating true independence of these factors. The final logistic regression model demonstrated a good overall fit to the data, as indicated by the Hosmer-Lemeshow test (χ² =162.70, p < 0.001) and explained 55.89% of the variance in preterm labour risk (Nagelkerke’s pseudo-R² = 0.5589), suggesting strong predictive power.

Table 2 Bivariable and multivariable analyses of the association between participants’ characteristics and preterm labour

Discussion

Our current study reported no statistically significant association between Trichomonas vaginalis (T. vaginalis) infection, a common sexually transmitted infection [33], and preterm labour (PTL) in Lagos, Southwest Nigeria. Also, our study did not identify a clear gestational age-related trend in the prevalence of T. vaginalis in cases of spontaneous PTL.

The findings of this study provide region-specific insights by demonstrating a significant association between T. vaginalis infection and PTL in our setting, thereby expanding the existing body of knowledge on the role of T. vaginalis in adverse pregnancy outcomes. The lack of statistically significant difference in the prevalence of T. vaginalis infection in women with and without spontaneous PTL and across the different gestational age categories suggests that the presence of T. vaginalis infection does not appear to correlate with specific gestational ages. Contrary to previous studies that have suggested a possible link between T. vaginalis infection and PTL [1, 11, 17, 18, 25], the findings of our study did not demonstrate a statistically significant association between T. vaginalis and the risk of PTL. This, however, corroborated the findings by Azargoon et al. in Semnan, Iran [26] which suggest that while T. vaginalis infection remains a concerning public health issue due to its association with other adverse reproductive outcomes, its direct impact on PTL in some populations may be limited. Several factors may account for the discrepancies observed between our study and prior research. One possible explanation is the difference in diagnostic techniques. Our study utilised a molecular-based detection method, which enhances accuracy by minimising false-negative results, compared to traditional microscopy or culture-based methods [27, 34, 35]. Additionally, variations in study populations, including demographic and geographic differences, differences in sample size across studies, disparities in co-infections, immune response, or access to antenatal care may influence the observed associations across the studies [1, 11, 17, 18, 25].

Women who were not booked for antenatal care have more than 60 folds risk of PTL in our study, similarly to the reported findings in previous studies that revealed that even when birth takes place in the hospital, lack of or low attendance to antenatal care carries a substantially elevated risk of severe adverse pregnancy outcomes [36, 37]. This may be because there was limited opportunity for prevention, identification, and prompt treatment of risk factors of PTL, such as abnormal vaginal discharge and PROM, in these women. Also, the more than four-fold increase in the risk of PTL among women having abnormal vaginal discharge as reported in our study is in similarity to the findings of several previous studies [1, 38,39,40]. This is thought to be linked to maternal innate immune inflammatory response following infection, such as T. vaginalis, as was observed in this study. Markers of neutrophil activation, including high levels of cervical interleukin-8 and vaginal defensins, have been demonstrated in pregnant women with asymptomatic T. vaginalis infection [19]. These markers have also been associated with amniotic fluid infection, PROM and preterm birth [19, 38]. Furthermore, the strong association observed between PROM and PTL (OR = 85) in our study suggests that PROM may serve as a mediating factor in the relationship between T. vaginalis infection and PTL. This aligns with existing evidence indicating that T. vaginalis infection is associated with increased membrane fragility and inflammatory responses [16], both of which predispose to PROM, thereby indirectly contributing to PTL [41]. Previous studies have also highlighted this pathway, emphasising that the risk of PTL in the presence of T. vaginalis infection may be significantly amplified with PROM [25, 40].

A key strength of this study is the use of a molecular-based diagnostic technique for T. vaginalis, the Strong Step^® Trichomonas vaginalis Antigen Rapid Test, which offers superior sensitivity and specificity compared to traditional methods such as wet mount microscopy and culture. This molecular technique detects low parasite loads, with reported sensitivities exceeding 90%, thereby reducing the risk of false negatives [34, 35]. This, in turn, minimises misclassification bias and ensures a more accurate assessment of the association between T. vaginalis infection and preterm labour. It is, however, important to interpret our findings in the context of the study limitations. Firstly, the cross-sectional nature of the study design could limit the establishment of causality between T. vaginalis infection and PTL as reported in the results. Secondly, as the pregnant women in the study were enrolled in hospitals within the Lagos metropolis, thus excluding a major proportion of women who mostly deliver at home in slums and suburban settings of Lagos, the findings are, therefore, generalisable only to these clinical settings. Thirdly, the use of consecutive sampling technique in the study may introduce selection bias, which could potentially limit the generalizability of our findings beyond the sampled population. Fourth, we did not screen the participants for other sexually transmitted infections other than T. vaginalis, thus making their systematic exclusion and control as confounding variables challenging. Finally, although a priori sample size calculation was performed to ensure adequate power for detecting association between T. vaginalis infection and PTL, the wide confidence interval for the adjusted odds ratio of T. vaginalis infection and PTL suggests low precision, likely due to the sample size and event frequency. Despite these, our study provides valuable insights into the association between T. vaginalis infection and spontaneous PTL in Lagos, Nigeria, highlighting the need for further research with more robust study designs.

Conclusions

Our study reported no association between T. vaginalis infection and PTL. Therefore, while T. vaginalis remains a prevalent STI, it appears to have a limited role in the occurrence of PTL. We suggest the need for future adequately powered longitudinal studies with consideration for other potential confounding factors, such as sexual lifestyle and coexisting STIs, to be conducted in more diverse populations within clinical and community settings in Nigeria to further explore the impact of coinfections or interactions between other local or systemic pathogens and T. vaginalis infection on the occurrence of PTL.