Article Text

PDF

Original article
Chlamydia trachomatis IgG seroprevalence in the general population of the Netherlands in 1996 and in 2007: differential changes by gender and age
  1. F van Aar1,
  2. M de Moraes1,
  3. S A Morré2,3,
  4. J E A M van Bergen1,4,5,
  5. F R M van der Klis6,
  6. J A Land7,
  7. M A B van der Sande1,8,
  8. I V F van den Broek1
  1. 1Epidemiology & Surveillance Department, Centre for Infectious Disease Control, National Institute of Public Health and the Environment (RIVM), Bilthoven, The Netherlands
  2. 2Laboratory of Immunogenetics, Department of Medical Microbiology and Infection Control, VU University Medical Centre, Amsterdam, The Netherlands
  3. 3Institute of Public Health Genomics, Department of Genetics and Cell Biology, Research Institute GROW, Faculty of Health, Medicine & Life Sciences, University of Maastricht, Maastricht, the Netherlands
  4. 4STI AIDS Netherlands (SOA AIDS Nederland), Amsterdam, The Netherlands
  5. 5Department of General Practice, AMC-UVA, Amsterdam, The Netherlands
  6. 6Laboratory for Infectious Disease and Screening, Centre for Infectious Disease Control, National Institute of Public Health and the Environment (RIVM), Bilthoven, The Netherlands
  7. 7Department of Obstetrics and Gynaecology, University Medical Centre Groningen, University of Groningen, Groningen, The Netherlands
  8. 8Julius Centre, UMCU, Utrecht, The Netherlands
  1. Correspondence to F van Aar, National Institute of Public Health and the Environment (RIVM), P O Box 1 (nr 75), Bilthoven 3720 BA, The Netherlands; fleur.van.aar{at}rivm.nl

Abstract

Objectives Chlamydia trachomatis (CT) reporting rates from sexually transmitted infection clinics and general practitioners have shown a rising trend in the Netherlands. It is unknown to what extent this reflects increased CT transmission or improved case finding. To achieve more insight into the CT epidemic, we explored the CT IgG seroprevalence (a marker of past CT infection) in the general population of the Netherlands in 1996 and in 2007.

Methods From two population-based studies in 1996 and 2007, serum samples, demographic and sexual behaviour outcomes were examined, including 1246 men and 1930 women aged 15–39 years. Serum CT IgG antibodies were analysed using the Medac CT IgG ELISA test. Multivariate logistic regression analyses explored the seroprevalence and determinants over time.

Results The CT IgG seroprevalence was higher in women than in men (10% vs 6%). Among women aged 25–39 years the seroprevalence was lower in 2007 (9%) than in 1996 (14%; adjusted OR (aOR) 0.6, 95% CI 0.4 to 0.8). There was no statistical evidence of a difference in seroprevalence within birth cohorts. Factors associated with seropositivity were male gender (aOR 0.4, 95% CI 0.3 to 0.7), a self-reported history of CT infection (aOR 5.1, 95% CI 2.6 to 10.0), age 25–39 years (aOR 1.7, 95% CI 1.1 to 2.7), non-Western ethnicity (aOR 2.2, 95% CI 1.4 to 3.3) and ≥2 recent sexual partners (aOR 2.2, 95% CI 1.3 to 3.5).

Conclusions Between 1996 and 2007 the proportion of individuals in the general population with CT IgG antibodies was lower among women aged 25–39 years, but remained similar among younger women and men.

  • CHLAMYDIA TRACHOMATIS
  • SEROLOGY
  • SURVEILLANCE
  • EPIDEMIOLOGY (GENERAL)

Statistics from Altmetric.com

Introduction

Chlamydia trachomatis (CT), an obligate intracellular bacterium, is the cause of the most commonly reported bacterial sexually transmitted infection (STI) in the Netherlands.1 ,2 A population-based study in 2002/2003 reported a CT prevalence of 2.5% and 1.5% in 16–29-year-old women and men, respectively.3 In 2012, 14 721 cases were reported at the Dutch STI centres and an estimated 25 380 by general practitioners (GPs). The highest positivity rates for CT observed in STI clinics were among women aged 15–19 years (19.0%) and among (heterosexual) men aged 15–24 years (19.0%).1

The course of CT infections is often asymptomatic, and the proportion of subclinical infections has been estimated at 50% in men and 70% in women.4 ,5 This enables the infection to persist and spread unknowingly. In men, the most common manifestation of persistent CT infections is urethritis, which may lead to epididymitis.4 ,5 In women, persistent infections may ascend to the upper genital tract and lead to pelvic inflammatory disease (PID)4 ,5 which, in turn, may have major sequelae such as tubal factor infertility, ectopic pregnancies and chronic pelvic pain.6

Undetected and untreated infections do not necessarily lead to complications; spontaneous clearance of the CT infection may occur, with a reported 95% clearance rate in three years.7 ,8 Little is known about antibody profiles during acute and chronic CT infections. Specific CT antibodies (such as IgG and IgA) can be detected in serum during infection, but the proportion and characteristics of individuals in the general population who will have a persistent seroresponse is not well established.9 ,10 In retrospective studies in infertile women, the detection of CT IgG antibodies in serum has been associated with tubal pathology.10 ,11 CT IgG antibodies are therefore considered as an intermediate marker for tubal pathology in infertile women, and this knowledge is applied in the decision tree for further diagnostic procedures in the subfertility investigation.

CT control programmes aim to interrupt the CT transmission cycle, as early detection and treatment of (asymptomatic) CT infections can reduce the risk of PID and its sequelae on an individual level and may reduce transmission in the population. CT reporting rates from STI clinics and GPs have shown a rising trend in the past two decades in the Netherlands.1 ,2 This trend is difficult to interpret, as increased intensity of CT transmission may have occurred, but also the increased availability of DNA amplification testing and the introduction of proactive screening programmes since 1995 might be major contributors to the increasing CT positivity rates. We investigated whether the CT IgG seroprevalence has changed in the general population between 1996 and 2007, using repeated cross-sectional estimates of the CT IgG seroprevalence.

Methods

Study design

Serum samples from two consecutive cross-sectional population-based serosurveillance studies were used, which had been conducted in 1995–1996 (Pienter 1, referred to as ‘1996’) and 2006–2007 (Pienter 2, referred to as ‘2007’). These samples were from established large serum banks and were appropriate for the current study since they covered a time period in which enhanced CT prevention efforts were gradually implemented in the Netherlands.

The primary objective of the Pienter studies was to estimate age-specific seroprevalence of antibodies against diseases targeted by the national immunisation programme. The designs of the two Pienter studies, previously described in detail, were both approved by a medical ethical committee.12 ,13 Briefly, from randomly selected municipalities in the Netherlands, an age-stratified selection of individuals (classes 0, 1–4, 5–9, … 75–79) were invited to donate blood and to complete a questionnaire. Pienter 2 aimed to assess the seroprevalence also in migrants and low immunisation communities. Therefore, oversampling of these specific subpopulations was performed. Participants signed an informed consent form that stated, inter alia, that the serum samples would be tested for antibodies against infectious diseases and that the data and samples would be stored for the purpose of later public health research related to infectious diseases.

All samples that were available from the targeted age group (15–39 years) in 2007 were included. From the samples collected in 1996, a random selection was made matched for age and sex. In total, 1930 women were included (958 in 1996 and 972 in 2007) and 1246 men (621 in 1996 and 625 in 2007). Assuming a baseline prevalence in 1996 of around 10% (range 5–15%, based on a study by Lyytikäinen et al14), we calculated that the specified number of samples would enable detection of a 3% change (increase or decrease) from 1996 to 2007 at the level of 80% power (β) and 5% significance level (α).

Laboratory methods

Serum CT IgG antibody testing was performed using the Medac quantitative CT IgG ELISA test (Wedel, Germany). This Chlamydia antibody test (CAT) uses a synthetic peptide from the immunodominant region of the major outer membrane protein (MOMP) as an antigen, ensuring minimum cross-reactivity with Chlamydia pneumoniae.15–17 Performance comparison of the ELISA test with the microimmunofluorescence assay found the sensitivity and specificity of the CAT test to be 71.4 and 97.3, respectively.16 CAT results are reported on a continuous scale (arbitrary units per millilitre (AU/mL), which are classified into three groups for analyses: positive (≥28 AU/mL as standard), negative (≤22 AU/mL) and equivocal (22–28 AU/mL).

Statistical analyses

Equivocal CAT results (n=19, 0.6%) were excluded from all analyses creating two groups: positive (≥28 AU/mL) and negative (≤22 AU/mL). Seroprevalence estimates representative of the general Dutch population were achieved by assigning sampling weights (incorporating gender and age) to each sample. The χ2 test was used to compare the characteristics and seroprevalence of the populations between 1996 and 2007. Analyses of the seroprevalence were stratified by gender and age group. Two different age variables were used: (1) split into two categories, 15–24 vs 25–39 years, because people <25 years of age are considered a high-risk group and therefore they are offered free testing at STI clinics1; (2) using 5-year age bands equal to the age-stratified selection of the study population (χ2 test for trends). To explore the seroprevalence within birth cohorts 1967–1971, 1972–1976 and 1977–1981, the seroprevalence among people aged 15–19, 20–24 and 25–29 years in 1996 was compared with those aged 25–29, 30–34 and 35–39 years in 2007, respectively.

We explored the interaction between gender and study period and between age and study period by assessing the association between the seroprevalence and study period including an interaction term (gender×year or age×year). We considered a value of p<0.05 as strong evidence for an interaction and a value between p<0.10 and p≥0.05 as suggestive evidence for an interaction. Logistic regression analyses were conducted: first, to explore changes in seroprevalence between 1996 and 2007 stratified by gender and by age group (15–24 vs 25–39 years) and, second, to identify determinants of seropositivity for 1996 and 2007 separately and for combined data. Determinants reported in the literature (gender, age group, level of urbanisation, level of education, ethnicity, number of partners and a self-reported history of CT) were a priori included into the multivariable model.3 ORs with 95% CI were calculated. All analyses were performed using SPSS V.19.0.

Results

Characteristics of the study populations

The study populations of 1996 and 2007 had similar proportions by gender, age group and sexual preference but not for level of urbanisation and education, past CT infection and ethnicity (table 1). The majority of both study populations were of Western ethnicity, among which Dutch origin was the most common (1996: 93.4%, n=1394; 2007: 91.6%, n=1325). The non-Western group mainly consisted of first-generation Moroccan or Turkish people in 1996 (41.2%, n=35) and first-generation non-Western people from other countries in 2007 (33.1%, n=50; p<0.01).

Table 1

Numbers and proportions by sociodemographic characteristics and sexual risk factors of the study population in 1996 and 2007

Chlamydia trachomatis IgG seroprevalence in 1996 and 2007

Analyses of the combined data from 1996 and 2007 showed differences in seroprevalence between women and men and between age groups (table 2). The overall seroprevalence among individuals aged 15–24 years did not differ between women and men (6.7% and 5.6%, respectively; p=0.44). The seroprevalence among women was higher in 25–39-year-olds (11.4%) than in 15–24-year-olds (p<0.01), whereas among men the seroprevalence was similar in both age groups (p=0.99). Accordingly, a higher seroprevalence among women aged 25–39 years than in men aged 25–39 years was observed (11.4% vs 5.6%, p<0.01).

Table 2

Weighted numbers and percentages by gender and age group, weighted CT IgG seroprevalence (N CAT+), and corresponding CT IgG seroprevalence rates (% CAT+) with 95% CI in the Netherlands in 1996 and 2007

Comparison of the seroprevalence between 1996 and 2007 showed differences by gender and by age group (15–24 vs 25–39-year-olds, table 2). There was suggestive evidence for an interaction between gender and study period (p=0.07). The evidence for an interaction between age group and study period was strong among women (p<0.05) but not among men (p=0.61). The overall seroprevalence among women was lower in 2007 than in 1996 (8.5% vs 11.1%; adjusted OR (aOR) 0.7, 95% CI 0.5 to 0.9, p=0.02). The seroprevalence among women aged 15–24 years was 6.1% in 1996 and 7.3% in 2007 (aOR 1.3, 95% CI 0.7 to 2.4, p=0.49). Among women aged 25–39 years, the seroprevalence was lower in 2007 than in 1996 (9.1% vs 13.8%, aOR 0.6, 95% CI 0.4 to 0.8, p<0.01). Among men the overall seroprevalence was 5.2% in 1996 and 6.2% in 2007 (aOR 1.1, 95% CI 0.7 to 1.9, p=0.68). In men aged 15–24 years and those aged 25–39 years the seroprevalence was similar in 1996 and 2007 (1996: aOR 1.0, 95% CI 0.4 to 2.3, p=0.93; 2007: aOR 1.2, 95% CI 0.6 to 2.3, p=0.59).

In 1996 the seroprevalence increased by age group among women from 3.2% among 15–20-year-olds to 15.3% among 34–39-year-olds (p<0.01; figure 1A). In 2007 there was no strong statistical evidence for an association between age and seroprevalence among women (p=0.07). Among men there was neither an increased nor decreased seroprevalence by age group in either 1996 (p=0.89) or 2007 (p=0.61; figure 1B).

Figure 1

(A,C) Chlamydia trachomatis (CT) IgG seroprevalence by study period (dark grey bars: 1996; light grey bars: 2007) and by age group for women (A) and men (C). (B,D) CT IgG seroprevalence by birth cohort for women (B) and men (D) in 1996 and 2007. Please note that the bars shown in (B) and (D) are similar to the bars shown in (A) and (C). For example, the seroprevalences shown for birth cohort 1967–1971 in (B) are similar to the dark grey bar for 25–29-year-olds in 1996 and the light grey bar for 35–39-year-olds in 2007 in (A).

Analysis of the seroprevalence within birth cohorts over time showed a small increase within the 1977–1981 female birth cohort, but there was no strong statistical evidence (p=0.18; figure 1C). There was also no statistical evidence for changes in seroprevalence within the other female birth cohorts (1967–1971: p=0.65; 1972–1976: p=0.92) or within the male birth cohorts (1967–1971: p=0.15; 1972–1976: p=0.94; 1977–1981: p=0.79; figure 1D).

Determinants of Chlamydia trachomatis seropositivity

Multivariable analyses of combined data from 1996 and 2007 indicated that gender (aOR 0.4, 95% CI 0.3 to 0.7 for men) was associated with seropositivity (table 3). Multivariable analyses including data from 1996 showed that the odds of CT seropositivity was 0.5 (95% CI 0.3 to 0.7, p<0.01) times lower among men than women, but there was no statistical evidence for this in 2007 (aOR 0.8, 95% CI 0.5 to 1.2, p=0.24). Furthermore, individuals aged 25–39 years, those of non-Western ethnicity, a self-reported history of CT infection and those who had ≥2 sexual partners in the previous 6–12 months had higher odds of being seropositive (aOR 1.7, 95% CI 1.1 to 2.7; aOR 2.2, 95% CI 1.4 to 3.3; aOR 5.1, 95% CI 2.6 to 10.0; and aOR 2.2, 95% CI 1.3 to 3.5; respectively).

Table 3

Chlamydia trachomatis (CT) weighted IgG seroprevalence and determinants of CT IgG seropositivity

Discussion

In the Netherlands it was found that the seroprevalence of CT IgG antibodies was lower among women aged 25–39 years in 2007 (9%) than among women of the same age in 1996 (14%), while no such difference was observed among women aged 15–24 years. Correspondingly, the seroprevalence among women increased with age in 1996 but not in 2007. We observed a higher overall CT IgG seroprevalence in women (10%) than in men (6%). For men, we did not find a difference in the seroprevalence between 2007 and 1996 for any of the age groups nor an age-related increase. In both men and women the seroprevalence within birth cohorts remained similar over time. Identified determinants of seropositivity were related to sociodemographics and were similar to the determinants of CT infection and STI in the Netherlands.1 For example, a higher number of recent sexual partners, which has been considered one of the most relevant determinants of CT infection, PID and its sequelae, was found to be associated with seropositivity.3 ,18 ,19

Data from the national STI surveillance system provide information about detected infections only and are focused on the high-risk population that visits STI clinics rather than the general population. Since population-based studies give the least biased results, a major strength of the current study is that the results reflect changes in the prevalence of CT IgG antibodies among both men and women in the general population.10 ,11 A limitation of the present study was the availability of only two separate cross-sectional cohorts, which makes it impossible to assess trends. Another limitation was the relatively small number of seropositive cases in the cohorts, which resulted in insufficient power to detect differences between subgroups. Therefore, wide age bands were used in the multivariable analyses, resulting in a cumulative measure of (past) infection, which does not reflect the actual CT prevalence such as when using a narrow young age band.20 Lastly, (differences in) sexual risk behaviour could not be fully assessed in our study population due to limited data, particularly in 1996.

The seroprevalence among women aged 17–24 years in the general population was recently explored in England, showing an increase in the frequency of Chlamydia antibodies between 1993 and 2002.21 Between 2007 and 2010, a decline in the seroprevalence was reported which, according to the authors, was concurrent with increased rates of Chlamydia screening. Consistent with our data from 1996, they found higher seroprevalences in the older age groups, which is in line with a plausible cumulative prevalence as long as acquisition is faster than waning. However, the seroprevalence among 15–24-year-olds in our study is almost three times lower than the seroprevalence found among 17–24-year-olds in England (6% vs 17% in 1996 and 7% vs 19% in 2007). This might be explained by differences in study population (random sample of the general population vs people accessing healthcare in England) and/or the use of a different CT IgG antibody test (ELISA based on the antigen MOMP vs Pgp3 in England17). The results of another population-based study in which Finnish pregnant women were included and serum IgG antibodies to CT were analysed by a peptide-based enzyme immunoassay test were partly consistent with our results.14 They also found a decreasing CT IgG seroprevalence among women aged 23–28 years between 1983 and 2003, but they also found a decreasing trend in seroprevalence among women aged ≤23 years.

The interpretation of Chlamydia seropositivity is not straightforward due to limited insight into the course of antibody titres during and after CT infection. In addition, the effects of antibiotic treatment, repeated infections and progression to sequelae on seroconversion and serological titres over time are not yet well established.9 ,20 Studies assessing the course of IgG antibodies are difficult to interpret because it is not possible to determine precisely the time span between the moment of (often asymptomatic) CT infection and testing. Misclassification of chronically infected individuals as recently infected may therefore occur. Moreover, most of the published studies on the course of CT IgG antibodies were based on selected populations either visiting fertility clinics (ie, at high risk for late sequelae) or STI clinics because of symptoms or recent risk of exposure (having a high probability for a recent infection).20 ,22–24 Knowledge of the presence and the course of IgG antibodies in individuals with an asymptomatic infection in a general population, presumably with far fewer recent infections, is not available.9 Another complication in interpreting seroprevalence data is that the proportion of individuals in the general population who will seroconvert and be persistently seropositive is not well understood.9 ,10 The seroprevalence in our study will therefore be an underestimation of the proportion of the population who ever had a Chlamydia infection.

A recent study showed that the majority of individuals visiting a genitourinary medicine clinic tested seropositive after acute infection and that the seroprevalence decreased in the first six months after infection, but they did not find a further reduction in seropositivity thereafter.20 Other studies also indicated that CT IgG antibodies can persist for years and, while antibody titres may decline over time, they can remain detectable even after treatment.22–24 This might explain the higher seroprevalence rate among older women than among young women in 1996, as the number of women who become sexually active and the number of lifetime partners increases with age. Similarly, one would expect an increasing seroprevalence within birth cohorts. Interestingly, there was no statistical evidence of a difference in seroprevalence over time for any of the birth cohorts. This finding could be partially explained by the lower seroprevalence observed among women aged 25–39 years in 2007 compared with women of the same age in 1996. Among men, we also did not find an age-related increase in seroprevalence within birth cohorts. Consistent with previous observations,25 ,26 the seroprevalence was lower among men than among women aged 25–39 years, which may be a result of sex differences in the humoral response to CT infection.27 We hypothesise that antibodies may persist for a longer period in women than in men, which might be associated with a stronger immune response in women with CT infections ascending to the upper genital tract.28

It is difficult to assess the impact of Chlamydia control activities using seroprevalence data because several factors may have had either an increasing or a decreasing impact on the seroprevalence over time, but the contributions of these factors are difficult to disentangle. For example, the number of STI-related episodes reported by GPs increased by 44%, from 664 per 100 000 patients to 955 per 100 000 patients between 2002 and 2007.2 At STI clinics the number of consultations increased from 11 378 in 1996 to 78 062 in 2007.29 Increased testing rates might have resulted in lower seroprevalence rates between 1996 and 2007 for two different reasons. First, the early detection and treatment of Chlamydia infections may interrupt Chlamydia transmission and could lead to a decline in incidence. Second, it has been hypothesised that early treatment may prevent seroconversion or may lead to an aborted immune response.30 Under this assumption, the lower seroprevalence among women aged 25–39 years in 2007, but not among men, might be partially explained by the fact that women visit GPs and STI centres more often than men.1 ,31 In the current study we also found that the proportion of individuals who reported having had a CT infection in the past was higher in 2007 (2.0%) than in 1996 (0.7%), which could reflect either a higher incidence or increased testing for CT. However, the seroprevalence among individuals with a self-reported history of CT decreased from 46% in 1996 to 29% in 2007 (results not shown due to small numbers). This could be in line with earlier treatment resulting in lower and less persistent seropositivity. Besides control activities, changes in sexual behaviour over time might also have had an impact on the seroprevalence. For example, between 1995 and 2006 the percentage of those sexually active in the age group 12–18 years increased from 24% to 31%,32 which might have had an increasing effect on the seroprevalence between the two studies.

To conclude, our results suggest that the proportion of individuals in the general population who had IgG antibodies against CT declined among women aged 25–39 years and remained similar among young women and men between 1996 and 2007. Changes over time differed by gender and by age group among women, which may reflect changes in CT epidemiology. Our data generate hypotheses for further study, as a better understanding of the path from infection, seroconversion and duration of seropositivity to CT complications is required to optimise the use of serology in evaluating control policies at the population level.

Key messages

  • The overall CT IgG seroprevalence did not differ between women (7%) and men (6%) aged 15–24 years.

  • Among women, the seroprevalence was higher in those aged 25–39 years (11%) than in those aged 15–24 years but, among men, the seroprevalence was similar in both age groups.

  • The seroprevalence was lower in 2007 (9%) than in 1996 (14%) among women aged 25–39 years and similar among men and younger women over time.

  • More insight into the course of CT antibody levels is needed to use serology for the evaluation of control policies at the population level.

References

View Abstract

Supplementary materials

  • Abstract in Dutch

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • Handling editor Jackie A Cassell

  • Acknowledgements The authors thank Dr L Mollema (RIVM, CIb) from the Pienter study group for providing the questionnaire data and calculating the sampling weights. The Laboratory of Immunogenetics, VU University Medical Center is acknowledged for analysing the serum samples. We also thank Dr L Nic Lochlainn (RIVM, CIb) for revising the spelling and grammar of the manuscript. In particular, we would like to thank Dr J C M Heijne (RIVM) for her helpful and constructive comments that greatly contributed to improving the final version of the paper.

  • Contributors IVFvdB, SAM, JEAMvB, JAL, FRMvdK, MABvdS developed the study design. MdM analysed and interpreted the data. FvA and IVFvdB supervised the data analyses. FvA drafted the manuscript. All the authors, in particular MdM and IVFvdB, contributed to drafting and revision of the paper. All authors read and approved the final manuscript.

  • Competing interests None.

  • Ethics approval Medical Ethical Committee of Netherlands Organisation for Applied Scientific Research (TNO), Leiden, The Netherlands (1996). Testing Committee of the Foundation of Therapeutic Evaluation of Medicines (METC-STEG) in Almere (2007).

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.