We found approximately 10% of women in Lower River Region of The Gambia to have at least one STI, with the highest prevalence being for TV whilst no TP was detected.
The proportion of women testing positive for an STI was comparable to findings from other STI prevalence studies from the region. A meta-analysis of data from several countries estimated the WHO Africa region (AFRO) prevalence of CT, NG, TV and TP to be 5.0 (3.8–6.6), 1.9 (1.3–2.7), 11.7 (8.6–15.6) and 1.6% (1.2–2.0%), respectively, in 2016.(1) Our findings are 2.1 (0.9 − 4.0), 0.2 (0.0 − 1.3), 7.6 (5.3 − 10.6) and 0% (0.0 − 0.9%), respectively, thus following a similar trend in relative prevalence. The WHO AFRO covers a large and diverse region, and the population of The Gambia only makes up a small proportion of the regional population, so comparison to these regional estimates should be considered accordingly.
Despite the variation in sampling and testing methods used in other studies, the similarity between these findings and previous studies was notable. At almost 8%, our data suggest TV is the most prevalent STI in this population. Other studies in West Africa estimate the prevalence of TV to range from 2 − 6%, although several used wet mount microscopy rather than molecular tests,(24–27) so the estimates are not directly comparable.(28) TV is often found to be highly prevalent, particularly in Africa.(1) At < 1%, the proportion of women with NG in this population from The Gambia was low, and lower than several similar studies from the region, where estimates ranged from 1 − 4%,(26, 27, 29, 30) including a recent study from an urban coastal region in The Gambia where the prevalence of NG was 2%.(13) We found an MG prevalence of 0.2%. This is lower than expected, as from the limited population-based studies reported elsewhere, prevalence is estimated to be 3.9% across low-income settings.(31) The proportion of CT-infected women in this study was 2.1% (0.9 − 4.0%). Although other estimates of CT prevalence from the region show considerable heterogeneity, this is generally in keeping with prevalence estimates from neighbouring countries.(26, 27, 29, 30) To find no women with active TP infection was surprising given other estimates of TP prevalence from elsewhere in The Gambia. It is possible that the three positive non-treponemal test results were active cases of TP. However, because of the well-described risk of false positive non-treponemal test results from other non-treponemal infections, such as Malaria,(32) in the absence of a positive treponemal test result we cannot say definitively that these were caused by TP. This diagnostic algorithm for TP is in keeping with recent guidelines from other organizations.(33) The four positive treponemal test results in the absence of positive non-treponemal test results suggests some historic exposure to TP but not active infection. Together, these data suggest some ongoing TP transmission in the population, albeit likely at a low level given the general low prevalence relative to estimates found elsewhere.
WHO identified strengthening STI surveillance and improving knowledge of prevalence (alongside symptom aetiology and antimicrobial resistance) as priorities in their STI strategy for 2016–2021.(34) Sexual and reproductive health services in The Gambia should be strengthened to manage this burden of STIs. Controlling the prevalence of TV should be a particular focus. Having TV may increase risk of HIV acquisition (35) and may have adverse impact on pregnancy outcomes,(36) although the evidence is inconsistent.(37) Only 5% of TV infections currently exhibit some resistance to first-line therapy,(38, 39) but there remain concerns of antimicrobial resistance (AMR) emerging.(40) CT is a pathogen which is generally thought to be increasing in some countries,(3, 4) so the low prevalence found here is notable. We found serovars D and G in this sample set, which were the two most common CT serovars in a 2012 study from nearby Guinea Bissau.(41) The low NG prevalence is reassuring, given NG’s propensity to develop extensive drug resistance(42, 43) resulting in it being on WHO’s “high priority” list for new antibiotics.(44) MG is another pathogen where AMR is of increasing concern.(43, 44) However, in a systematic review of mutations associated with macrolide and fluoroquinolone resistance in MG, AMR data for the African region were only available from Kenya and South Africa.(45) Integrating AMR surveillance with prevalence and incidence monitoring would be valuable in this setting, particularly considering the amount of community-wide exposure to antimicrobials imparted by neglected disease mass drug administration campaigns in The Gambia and its neighbouring countries.(46) As part of this work, we attempted to amplify AMR-associated single nucleotide polymorphism-containing regions from the residual ex-diagnostic DNA eluate of positive samples. We found a product to amplify in < 50% samples and generated several low-quality sequences among the gene segments which did amplify. Therefore, more reproducible methods should be employed for future AMR surveillance.
The study has two key strengths. First, data in this study were collected from a sample of the general population in sub-Saharan Africa. General population samples are more generalizable to the wider population than some other common STI survey sampling methods, such as sampling of clinic attendees or high-risk groups. No published evidence of STI prevalence in this population were available at the time of the study, and very few data on STIs were available from The Gambia as a whole. Also, LMICs in sub-Saharan Africa are under-represented in global estimates of STI burden. Second, recommended diagnostic approaches were used to detect STIs in this study: nucleic acid amplification tests for diagnosis of CT, NG, MG and TV and a specific serological diagnostic algorithm for detection of TP. These methods are essential for detecting asymptomatic carriage of STIs and offer improved sensitivity over culture or antigen detection methods.
There were also limitations to this study. First, for detection of CT, NG, TV and MG, we tested FVU. Despite its convenience as a sampling method, FVU is recognised as a difficult sample type to amplify DNA from and a less sensitive specimen type for detection of STIs than vaginal swab samples.(47, 48) This may be related to presence of inhibitors, or because of small proportion of the sample tested (in this case, 100 µL from several millilitres of FVU was tested). Second, the participant sampling technique was not random, so we cannot assume the data are representative of the local or national Gambian populations. Although the samples were taken before any parent trial intervention, the trial was examining an HPV vaccine and recruitment relied on voluntary self-presentation to the study, both of which may have led to recruitment bias. Also, pregnant women and women who were known to be HIV positive were excluded. The influence of these factors should be considered when interpreting the results. Finally, as a result of the nature of the parent trial from which these specimens were taken, men were not included in our sampling frame.
Several unanswered questions remain. First, larger, randomly sampled surveys and longitudinal prevalence monitoring are needed to expand the map of STI prevalence in The Gambia. Compared to the general population, high-risk populations such as female sex workers often bear a disproportionately high proportion of STIs. Inclusion of these groups in future surveillance efforts is important. Men should also be included. Second, several key infections of the reproductive tract (for example, Human Immunodeficiency Virus [HIV], Herpes Simplex Virus [HSV], bacterial vaginosis [BV], Candida albicans) were not tested for in this study. More data will be needed on their prevalence to ensure appropriate service provision. Third, prospectively designed STI surveys have increased scope to examine symptoms and behaviours related to STIs. Learning more about the epidemiology of these infections will be important in planning effective public health interventions. Finally, STIs, including TV which was the most common in this study, are often asymptomatic,(33, 49) so proactive screening of asymptomatic men and women should be core to monitoring strategies. Given the documented limitations of urine testing, supplementing urine testing with swab sample testing in subsequent studies would be valuable. Several studies have demonstrated that self-collected vaginal swabs are acceptable and feasible.(50, 51)
Our data, combined with those from elsewhere, will contribute to baseline STI prevalence estimates in The Gambia, laying a foundation for more systematic monitoring of the key challenges facing STI prevention, such as characterisation of antimicrobial susceptibility profiles among pathogen isolates and determining whether prevalence is changing over time. There is growing concern about STIs at the global level; the increasing evidence base in The Gambia should provide leverage for policy makers to ensure STIs are appropriately prioritised.