Genital chlamydia trachomatis (CT) infection is the most common, curable sexually transmitted disease (STD) worldwide (Woodhall et al., 2018). As up to 80% of cases are asymptomatic, CT infection often remains undetected or not diagnosed, which in turn results in the wide spread and delayed treatment of CT infection (Marcone et al., 2012). Furthermore, untreated CT infection can cause scarring of the upper reproductive tract in women and lead to serious complications, such as pelvic inflammatory disease, ectopic pregnancy, and tubal factor infertility (Haggerty et al., 2010). In men, CT infection is the most common cause of non-gonococcal urethritis and accessory gland infection (Wagenlehner et al., 2006). CT infection is also reported to be associated with increased risk of cervical cancer and acquisition of human immunodeficiency virus (HIV) infection (Koskela et al., 2000; Anttila et al., 2001; Petersen et al., 2020).

Chlamydia trachomatis infection is increasing worldwide from about 89 million new cases in 1995 to 127.2 million in 2016 (Newman et al., 2015; Rowley et al., 2019). In 2016, based on nucleic acid testing data, the global prevalence of C. trachomatis was estimated to be 3.8% in 15–49-year-old women and 2.7% in men whereas the global incidence rate was 34 cases per 1,000 person-years in women and 33 per 1,000 person-years in men, respectively (Rowley et al., 2019). In 2018, the European Centre for Disease Prevention and Control (ECDC) reported a crude C. trachomatis prevalence of 0.15% in 26 EU/EEA countries, and 5.3% increase of the prevalence from 2009 to 2018.¹ In the United States, there were more than 1.7 million C. trachomatis infections in 2018 (Torrone et al., 2014).

Given the significant burden of C. trachomatis infection and the serious sequelae, national C. trachomatis screening program has been implemented mainly targeting young women under 25 years old in several countries including England, Netherlands, and United States (Marcone et al., 2012; Workowski and Bolan, 2015) because they usually compose the majority of the reported C. trachomatis new infections (Fine et al., 2012). In these countries, a substantial decrease of CT infection-related complications has been observed at the end of the 1980s and early 1990s and a further decrease of pelvic inflammatory disease and ectopic pregnancy has been documented since 2000 (Chandra et al., 2005; Sutton et al., 2005; Bender et al., 2011; French et al., 2011; Marcone et al., 2012).

Unfortunately, there is no nationwide screening program of CT infection in China where bacterial sexually transmitted infections (STIs) are highly prevalent and have rapidly spread since early 1980s (Chen et al., 2000, 2011). Chlamydia is not a notifiable infectious disease in China although it has been included as a reportable STI in the national STI surveillance program since 2008 (Yue et al., 2020). Available data on the prevalence of CT infection are limited and a few prevalence surveys specifically targeted ad hoc populations, such as female sex workers (FSWs), men who have sex with men (MSM), immigrants, clinical patients, or pregnant women (Chen et al., 2011). A population-based study conducted in 1999 to 2000 that enrolled 3,426 Chinese residents using nucleic acid test showed a prevalence of 2.1% in men and 2.6% in women (Parish et al., 2003). In addition, several studies indicated that the CT prevalence ranged from 1.5 to 5.4% in the general population of women (Xia et al., 2015; Zhang et al., 2017; Rowley et al., 2019). So far, there is no longitudinal study about C. trachomatis infection in China.

Previous studies usually adapted nucleic acid amplification test (NAAT) to detect current C. trachomatis infection, rather than past infection. It is resource intensive and impossible to conduct population-based screening of C. trachomatis infection by using NAAT (Fine et al., 2012). In contrast, serological assays can determine the prevalence of both current and past C. trachomatis infection (Horner, 2017) and have been successfully used in several countries to assess C. trachomatis epidemiology (Lyytikäinen et al., 2008a; Horner et al., 2013, 2016; van Aar et al., 2014; Righarts et al., 2017; Woodhall et al., 2017). The performance of serological testing for C. trachomatis has been dramatically improved due to the identification of plasmid gene product 3 (Pgp3) antigen to avoid cross-reactivity with other Chlamydia spp. (Wills et al., 2009) and to improve suboptimal sensitivity (Woodhall et al., 2018). Immunoglobulin G (IgG) antibody against Pgp3 has been found to be the most reliable marker of C. trachomatis infection because Pgp3 is highly conserved across isolates and rarely found in C. pneumoniae (Gwyn et al., 2017; Kaur et al., 2018). Serological assays based on C. trachomatis Pgp3 antigen also outperformed other C. trachomatis antigens (Horner et al., 2016).

Chlamydia trachomatis serological assay provides a tool to investigate the prevalence and incidence of C. trachomatis infection. A thorough investigation of C. trachomatis infection including prevalence and incidence in general population is crucial to identify unmet needs for precise prevention and treatment services (Woodhall et al., 2018), especially in China where there is huge population size but no population-based survey of C. trachomatis infection. In this study, we explored the seropositive frequency and incidence of CT infection in the general population aged 18–65 years in a longitudinal community cohort from 2014 to 2018 in Northern China to achieve deep insight into the CT epidemic and anti-Pgp3 antibody dynamics in China.

To the best of our knowledge, this is the first longitudinal study about the seroepidemiology of C. trachomatis infection in the general population of China. We found that at least one quarter of the people aged 18–65 years have been infected with C. trachomatis over their lifetime based on the frequency of anti-Pgp3 antibody, while all age groups are susceptible to C. trachomatis infection, although the incidence and reinfection rate are relatively low (about 10 per 1,000 person-year) in the community of Northern China.

The most significant advantage of seroepidemiologic analysis is its capability to reflect both current and past C. trachomatis infection. Although there are no serologic assays approved by the US Food and Drug Administration (FDA), the preliminary evaluation of our in-house LISA by using sera from C. trachomatis NAAT-positive CT-infected women and low-risk children showed high sensitivity and specificity, which are comparable with the commercial Mikrogen ELISA. Although our “in-house” LISA used one Pgp3 protein from serovar E of C. trachomatis, it does show comparable performance with the commercial Mikrogen ELISA, which uses multiple C. trachomatis antigens (MOMP, TARP, and CPAF). Of note, Jones et al. (2020) recently reported the variation of Pgp3 protein with a genetic distance of 0.013 at the amino acid level among different C. trachomatis serovars. However, the small genetic distance of Pgp3 gene and the corresponding 98.7% sequence identity maintains a 94% probability of cross-reactivity within C. trachomatis species according to the formula

reported by Rahman et al. (2015, 2018b). In our study, we also validate the cross-reactivity of Pgp3 protein among different serovars analyzed. Eight NAAT-positive women (6.4%), including one infected with serovar E and seven with unknown serovars, were negative for both Pgp3-based LISA and Mikrogen ELISA. The lack of CT-specific antibody response has been previously reported in 5.5–10.2% of CT-NAAT positive subjects (Wang and Grayston, 1974; Komoda, 2007; Geisler et al., 2013; Bakshi et al., 2017) due to early CT infection without antibody response, reduced antigenic burden resulting from early therapy, inadequate dose of C. trachomatis to cause productive infection, or immunosuppressive diseases to prevent the production of CT specific antibodies (Geisler et al., 2012). However, in our study, the CT-NAAT-positive but antibody-negative patient was most likely in the early stage of CT infection since we found that the patient was negative for IgG antibody against Pgp3, but positive for IgM antibody against Pgp3. Our preliminary data supported the “in-house” Pgp3-based LISA as a suitable assay for the detection of anti-C. trachomatis antibody.

In our study, the overall seropositive frequency of C. trachomatis infection was 28.1% for the people aged 18–65 years old at baseline. To compare our results with previous studies, we adjusted the C. trachomatis seropositive frequency, incidence, and reinfection rate according to the assay performance, age groups, and gender within the study populations (Supplementary Table 5). The adjusted CT seropositive frequency in our study was similar to that in non-STD clinic outpatients from Tianjin, China (17.6% vs. 22.9%, p = 0.077) (Zhang et al., 2016), but significantly lower than that in the British general population (17.9% vs. 30.4% in men, p = 0.001 and 22.5% vs. 34.8% in women, p = 0.008) (Woodhall et al., 2017) and in the females from the general population of the United States (19.5% vs. 34.4%, p = 0.004) (Petersen et al., 2020). However, the adjusted frequency in our study was still relatively higher than that in the general population of Netherlands, especially more significant in men (16.6% vs. 8.1%, p = 0.004) (van Aar et al., 2014). Our results indicated that as the most common STI, exposure to C. trachomatis at some point over the lifetime is quite common in Northern China and could be as high as 50% in the sub-population aged 60–65 years.

Furthermore, our study uncovered the C. trachomatis incidence of 11 per 1,000 person-years in China based on the detection of anti-Pgp3 antibody seroconversion and the C. trachomatis reinfection rate of 14 per 1,000 person-years based on the fourfold increase of anti-Pgp3 antibody titers. The adjusted incidence was similar to that observed in the general population of New Zealand (14 per 1,000 vs. 19 per 1,000 person-year, p = 0.326) based on anti-Pgp3 seroconversion (Righarts et al., 2017) but significantly lower than that obtained in the community population of Australia (55 per 1,000 person-year, p < 0.001) based on the NAAT (Silver et al., 2015). The adjusted reinfection rate (18 per 1,000 person-years) in our study was similar to that in the American soldiers aged 16–51 years (21 per 1,000 person-year, p = 0.655) based on NAAT (Barnett and Brundage, 2001). Of note, the CT incidence identified in our study was 20–30-fold higher than that (0.37–0.55 per 1,000 person-year) reported from the national STI surveillance in China from 2015 to 2019 (Yue et al., 2020), suggesting a serious underestimation of C. trachomatis new infection rate in China partly due to 80% asymptomatic CT infection and the large number of undiagnosed CT infection as well as low reporting rate of the current surveillance system.

Like other STIs, there is a trend of age-related increasing seropositive frequency of C. trachomatis infection in the general population (van Aar et al., 2014; Woodhall et al., 2017). Previous studies found that CT seropositivity usually reached the peak at the age of 30–35 years followed by a slight decrease in older age groups (Woodhall et al., 2017). In our study, the CT seropositive frequency increased over ages, suggesting a broad susceptibility of all age groups to C. trachomatis infection in China. The susceptibility of all age groups to C. trachomatis infection was also supported by the similar CT incidence among different age groups in our study. In general, C. trachomatis incident cases (Scott Lamontagne et al., 2007; Walker et al., 2012; Silver et al., 2015) and NAAT-positive detections (Marcone et al., 2012; Sonnenberg et al., 2013) are more often found in teenagers (15–19 years) and declined with age thereafter. Two studies based on seroconversions also reported that the highest incidence occurred in those under 26 years in New Zealand (Righarts et al., 2017) and pregnant women under 23 years in Finland (Lyytikäinen et al., 2008b). These results are in line with the recommendation of the annual STI screening for sexually active women under the age of 25 in the United States [Centers for Disease Control and Prevention (CDC), 2019] and European countries [European Centre for Disease Prevention and Control (ECDC), 2008]. However, in our study, we failed to see the significantly higher incidence of CT infection around the age of 25 years, although STI surveillance data showed a 25-year-old population as a high risk of CT infection in China (Yue et al., 2020).

The results about the association of CT infection and genders are inconsistent. Several studies found lower CT seropositive frequency in men than in women and no decrease of C. trachomatis frequency over ages in men (van Aar et al., 2014; Horner et al., 2016; Woodhall et al., 2017). The lower CT seropositive frequency in men may be due to lower antibody-inducing infection rate (Närvänen et al., 1997), weaker antibody response post-infection (Bessho and Matsumoto, 1990), and the sub-performance of serological assays in men (Wills et al., 2009). However, in our study, slightly lower seropositive frequency was observed in men than in women (25.4% vs. 31.3%), but the difference was not statistically significant (p = 0.078). In addition, the C. trachomatis incidence was also similar in men and women (11 per 1,000 vs. 10 per 1,000 person-year, p = 0.833) in our study. Our results were in line with other studies based on anti-Pgp3 seroconversions or NAAT testing (Kim et al., 2011; de Walque et al., 2012; Sonnenberg et al., 2013; Rutherford et al., 2014; Corsenac et al., 2015; Silver et al., 2015; Ikonomidis et al., 2016; Matteelli et al., 2016; Babinská et al., 2017; Righarts et al., 2017; Huneeus et al., 2018), indicating no gender-related significant difference of CT infection.

Chlamydia trachomatis incidence and reinfection are important for CT prevention. Two studies in England and Australia reported a remarkably higher reinfection rate than incidence (22.3–29.9 per 100 person-year vs. 4.9–6.4 per 100 person-year in England; 22.3 per 100 person-year vs. 4.4 per 100 person-year in Australia.) in women aged 16–25 years (Scott Lamontagne et al., 2007; Walker et al., 2012). These results indicated the susceptibility of at least part (20–30%) of CT-infected subjects to the reinfection of C. trachomatis or the insufficient protection of CT-related immunity to C. trachomatis infection. A possible explanation is the “immunity arrest” effect of the timely antibiotic treatment (Vickers and Osgood, 2014) since nearly 20% untreated C. trachomatis infections can resolve spontaneously and the protective immunity in “self-cure” individuals has been previously reported (Geisler et al., 2013). However, quite similar C. trachomatis incidence (11 per 1,000 person-year) and reinfection rate (14 per 1,000 person-year) were found in the community population of northern China in our study. Previous studies found that for C. trachomatis new infection, the subjects who have been infected with CT experienced a nearly fivefold high risk than those who have never been infected (Scott Lamontagne et al., 2007; Walker et al., 2012); however, no difference of C. trachomatis incidence and reinfection rate were observed in our study, which may indicate the protective immunity after C. trachomatis infection or the lack of treatment-related “immunity arrest” in China.

Our study showed relatively a low seroreversion rate (6.7%) of the anti-Pgp3 antibody during the 4 years of follow-up. This finding was in agreement with a previous study in which 96.5% of CT seropositive women and 83.9% of men maintained seropositivity for 12 years (Horner et al., 2016). We also found that the titers of the anti-Pgp3 antibody remain stable with a half-life time of around 8 years. A similar decay rate of anti-C. trachomatis IgG antibody has been described by Puolakkainen et al. (1986) who reported that anti-C. trachomatis EB antibodies remain stable in 3–6 years. Interestingly, for the CT incident samples, we observed a rapid antibody reduction and a high seroreversion rate of 35.3% within the first year post-infection. Workowski et al. (1993) also reported a seroreversion rate of 31.6% within 5 months after C. trachomatis NAAT-positive test. Öhman et al. (2020) found that anti-C. trachomatis MOMP IgG antibodies were only detected in 65.5% of women within 3 months of CT diagnosis. However, different from our results, several studies reported a stable IgG level for at least 6 months (Cunningham, 1995; Geisler et al., 2012) or very low seroreversion rate (1.4–6%) within 1 year (Komoda, 2007; Bakshi et al., 2017) after a positive C. trachomatis NAAT test. Therefore, the dynamics of CT-specific IgG antibodies after C. trachomatis new infection needs to be further investigated. We believe that for the CT reinfection, pre-existent C. trachomatis antibody may be an important factor to affect the antibody dynamics since the reinfected individuals experienced a much lower seroreversion rate (10%) in our study. Therefore, we can further infer that slow anti-CT antibody decay may be a proxy of C. trachomatis reinfection, which can only be identified by serological assays because the NAAT test can only identify current C. trachomatis infection, but cannot distinguish primary infection from reinfection of C. trachomatis.

There are several limitations in our study. First, C. trachomatis infection refers to CT-specific antibody-inducing infection in our study. However, a lack of CT-specific antibody response has been reported in 5.5–10.2% of C. trachomatis-infected subjects (Wang and Grayston, 1974; Komoda, 2007; Geisler et al., 2012; Bakshi et al., 2017). Considering seroreversion over time, especially the rapid decay of CT-specific antibody in C. trachomatis newly infected individuals, serological investigation could underestimate the real burden of C. trachomatis infection. Second, lack of data about the STI history of participants as well as personal behavior and sex attitudes prevented further analysis of the risk factors associated with C. trachomatis infection. Third, previous studies in Europe and the United States indicate the 25-year-old females as a high-risk population of C. trachomatis infection. However, only 4% of the participants were 18–25 years old in our study, and no conclusion can be made about their risk or burden of C. trachomatis infection in our study. Finally, although there were no significant differences observed between the subjects included and excluded in this study with respect to age, gender, ethnicity, marital status, and education levels, we cannot ensure that the characteristics, such as participants’ STI history, personal behavior, and sex attitudes, were equal between these two groups due to the non-random sampling strategy. The 744 people with longitudinal records and serum samples may not represent the underlying population of the community. Therefore, we must be cautious in interpreting the current results, and further study is needed. Additionally, the community we investigated is special since the participants are all oil field employees and their family members. The results obtained in this community cannot be expanded to the entire Chinese population, although our study provided the first community-based seroepidemiology and antibody dynamics of C. trachomatis in China.

In this community-based longitudinal cohort study, we presented an overall CT seropositive frequency of 28.1% for Chinese adults and up to 48.6% frequency for a 60–65-year-old population. Although CT incidence and reinfection rate were low, all age groups are susceptible to CT infection. Comprehensive C. trachomatis prevention strategies are thus needed, especially for the prevention of adverse reproductive outcomes in women.

The dataset presented in this study can be found in online repository at https://www.ncbi.nlm.nih.gov/genbank/ with an accession number of OK148602.

The studies involving human participants were reviewed and approved by the Ethical Committee of the Jidong Oil-field Hospital of China National Petroleum Corporation. The patients/participants provided their written informed consent to participate in this study.

ST and YD conceived and designed the experiments. DX, JS, and CA collected the clinical sample and data. JS, JZ, HL, and YL performed the experiments. JS, DX, YD, and ST interpreted the results. JS made the tables and figures. JS and ST wrote the manuscript. All authors reviewed, revised, and approved the final report.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.