You are here

Article Text

Article menu
Basic science
Original article
Effect of time since exposure to Chlamydia trachomatis on chlamydia antibody detection in women: a cross-sectional study
  1. Patrick J Horner1,2,
  2. Gillian S Wills3,
  3. Rosy Reynolds4,
  4. Anne M Johnson5,
  5. David A Muir6,
  6. Alan Winston3,
  7. Andrew J Broadbent3,
  8. David Parker7,
  9. Myra O McClure3
  1. 1School of Social and Community Medicine, University of Bristol, Bristol, UK
  2. 2Bristol Sexual Health Centre, University Hospitals Bristol NHS Foundation Trust, Bristol, UK
  3. 3Jefferiss Trust Laboratories, Wright-Fleming Institute, Imperial College, London, UK
  4. 4Department of Medical Microbiology, North Bristol NHS Trust, Bristol, UK
  5. 5Research Department of Infection & Population Health, University College, London, UK
  6. 6Department of Diagnostic Virology, St Mary's Hospital, Imperial College Healthcare NHS Trust, London, UK
  7. 7Novel Consulting, Crown House, Home Gardens, Dartford, UK
  1. Correspondence to Profeser Myra O McClure, Jefferiss Research Trust Laboratories, Wright-Fleming Institute, Faculty of Medicine, Imperial College London, St Mary's Campus, Norfolk Place, London W2 1PG, UK; m.mcclure{at}imperial.ac.uk

Abstract

Objectives To investigate what factors influence the detection of Chlamydia trachomatis antibody following genital tract infection.

Methods One hundred and sixty-four women with a previous history of C trachomatis infection contributed to an earlier report on the performance of chlamydia antibody ELISA assays. We undertook further analysis to explore how chlamydia antibody assay sensitivity changes with time since infection.

Results Chlamydia antibody was detected in more women soon after the last detection of chlamydia at the lower genital tract than at later times. This holds true for all tests, but the Anilabsystems IgG EIA, Medac pELISA plus ELISA and the Savyon SeroCT-IgG ELISA were less sensitive than the pgp3 ELISA and the Anilabsystems microimmunofluorescence (MIF) assay at all time points except during current infection. Fall in seropositivity in women generally occurred in the early weeks and months following the last episode of chlamydia infection. There was no clear pattern of further reduction in seropositivity after 6 months. Multiple previous episodes were associated with increased seropositivity in the pgp3 assay (two or more vs one, OR 19, p<0.001) and other tests, but the effect was significantly smaller for the Anilabs, Medac and SeroCT MOMP peptide ELISAs, but not for the MIF assay.

Conclusions Chlamydia antibody detection decreases with time since infection and this is most apparent in the first 6 months. In women who have had more than one infection, antibody remained detectable longer for all tests, but this was more marked for the pgp3 ELISA and MIF assay.

  • Antibodies
  • Chlamydia Trachomatis
  • Chlamydia Serology
  • Seroprevalence

https://doi.org/10.1136/sextrans-2011-050386

Statistics from Altmetric.com

    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.

    Chlamydia trachomatis is the most common bacterial sexually transmitted infection in the developed world.1–3 It has been associated with significant morbidity, including pelvic inflammatory disease (PID), ectopic pregnancy and tubal factor infertility and, as a result, chlamydia control programmes have been widely advocated.1–3 In England, the National Chlamydia Screening Programme (NCSP), set up in 2003 for sexually active 15–24-year-olds, aims to reduce the population prevalence, ongoing transmission, and the development of chronic sequelae.2 This programme has been criticised because no mechanism of demonstrating its effectiveness was ever introduced.4 Although 32 per cent of the 15–24-year-old population was tested for chlamydia in 2010/2011 there is no good evidence to indicate that it has resulted in either a fall in prevalence of chlamydia or a reduction in chronic sequelae that are due to chlamydia.5 ,6 Based on a review by the National Audit Committee, the House of Commons Public Accounts Committee recommended that the NCSP urgently needed to measure the Programme's impact on the prevalence of chlamydia.6 No evaluation framework for this exists. Chlamydia-positivity within the NCSP as assayed by using nucleic acid amplification test (NAATs), cannot be used to monitor changes in population prevalence over time, as a reduction in the numbers found to be positive may reflect a higher proportion of low-risk people being screened over time, rather than the impact of screening itself.7 ,8

    Serology has recently been demonstrated to be able to provide valuable information on the natural history of chlamydia infection9 and could improve our understanding of chlamydia transmission and pathogenesis.10 Used in this way, it has the potential to improve our ability to assess the impact of efforts to reduce the burden of genital C trachomatis infection.10 Successful application of serology for this purpose will require a sensitive and specific test and a better understanding of the persistence of antibody titres following acute infection.10 We have shown that a recently developed ELISA using the pgp3 antigen is more sensitive than three of the most commonly used commercial ELISAs. The specificity of all four ELISAs was >95%, with no cross-reactivity to Chlamydophila pneumoniae.11 We undertook further analysis of data from this study to explore how assay sensitivity changes with time since infection in women.

    Methods

    Patient samples

    One hundred and sixty-four women with a previous history of C trachomatis infection were recruited from two genitourinary medicine (GUM) clinics in Bristol and London between May 2006 and January 2008, as described previously.11 Signed informed consent was obtained and blood taken for subsequent analysis. Bristol has used Gen-Probe APTIMA Combo 2 since 2005 and prior to that PCE EIA (Dako, Ely, Cambridgeshire, UK) with PCR confirmation using Cobas PCR assay (Roche Diagnostics Inc, Branchburg, New Jersey, USA). St Mary's Hospital, London has used the Becton Dickinson ProbeTecET SDA since 2001, and prior to that the Abbott LCx.11 Testing was undertaken if patients presented with a new problem, or if they had been at risk of infection since they were last tested.

    Details of age, ethnicity, number of sexual partners in the previous 3 months, number of previous episodes of C trachomatis, past history of gonorrhoea, past history of chlamydia PID, the test used to detect chlamydia and whether chlamydia had been detected outside the source clinic (we were, therefore, unable to verify if chlamydia had been detected from the person's notes) were obtained from the clinic notes.

    This study was approved by the North Somerset ethics committee ref 05/Q2003/48.

    C trachomatis antibody assays

    The pgp3 ELISA was carried out as previously described.11 Individuals were also tested using three MOMP peptide-based ELISAs: the C trachomatis-IgG-pELISA plus Medac assay (Medac, Germany), the SeroCT-IgG ELISA (Savyondiagnostics, Israel) and the C trachomatis IgG EIA (AniLabsystems, Finland). We also evaluated the C trachomatis Microimmunofluorescence (MIF) assay using sera at a titre of 1 in 16. This assay is contained within the C pneumoniae IgG/IgM lipopolysaccaharide-depleted MIF Test kit (AniLabsystems, Finland). All assays were undertaken as directed by the manufacturer's instructions.

    Statistical analysis was undertaken using Stata 11 (College Station, Texas, USA). Associations were tested by univariate logistic regression; p<0.05 was taken to indicate statistical significance (two-sided) and OR calculated with 95% CI. Robust SEs were used in models including antibody assay method to account for clustering of results within patients. The differential effect of covariates on seropositivity measured by different assays was assessed by the significance of the interaction terms in models including the covariate, assay method and their interaction. To elucidate the form of the influence of time since last episode of detected and treated chlamydia, a number of logistic regression models were fitted (see online supplementary appendix 1), considering different periods of time and their combinations (currently positive; and periods up to 6 months, 12 months and 4 years since the last recorded chlamydia episode). Models were compared by Akaike Information Criterion. All the covariates that were significant in univariate analysis were included in a combined model for pgp3 seropositivity. The combined model was refined by stepwise removal and addition of single covariates; covariates were retained if they contributed to a model fit with p<0.05.

    Results

    Effect of the length of time since last episode of chlamydia infection on chlamydia antibody in serum

    Clinical information was available on 150 women and for 141 of these, it was known when they were last diagnosed with and treated for chlamydia. Table 1 details the effect of time since the last chlamydia-positive diagnosis on the proportion of women who had antibodies detected by each assay. This is also represented graphically in figure 1 for the pgp3 and Ani Labsystems ELISAs (the latter being the most sensitive commercial ELISA) and the chlamydia MIF assay. Seropositivity (see methods) was higher early after the last detection of chlamydia, and this held true for all tests (see table 1). The 3 MOMP peptide-based ELISAs, Anilabs, Medac and Sero-CT, were less sensitive than the pgp3 ELISA at both early (≤6 months) and at later times (>6 months) since chlamydia was detected and treated. Thus, the advantage of the pgp3 ELISA over the 3 commercial MOMP peptide ELISAs did not appear to change with length of time since last infection. However, there was no significant difference in sensitivity between the pgp3 ELISA and MIF assay, either overall or at ≤6 months and at later times. Seropositivity was also significantly higher with the Anilabs, Medac and SeroCT ELISAs, and the MIF assay, in women who were chlamydia NAAT-positive at recruitment, compared with those who were either chlamydia NAAT-negative or not tested. The difference was not significant for the pgp3 ELISA (OR 4.7, 95% CI 0.7 to 209, p=0.19).

    View this table:
    Table 1

    The length of time since chlamydia diagnosis affects chlamydia antibody detection

    Figure 1
    Figure 1

    Shows moving averages for seropositivity against time since Chlamydia trachomatis infection was diagnosed, measured by three assays: the pgp3 and Anilabs ELISAs and microimmunofluorescence assay for which sera were diluted 1/16. The patients were arranged in order of time since last chlamydia diagnosis. The % seropositivity and mean time since last chlamydia detection was calculated for the first 20 patients, then for patients 2–21, 3–22 and so on, up to the last 20 patients. These are the 20-observation moving averages. (Note that currently NAAT-positive patients contribute to the averages only up to 0.22 years (2.6 months)).

    For the Anilabs, SeroCT and MIF-CT16 tests, all the NAAT-positive women were seropositive. In these cases, logistic regression models (see online supplementary appendix 1), including current positivity, could not be compared with others because of this ‘perfect prediction; of the other models considered, that with a single cut-off at 6 months gave the best prediction (whether or not NAAT-positive women were included). For the remaining tests (pgp3 and Medac), where comparison with models including current positivity was possible, the single 6-month cut-off still gave the best prediction in almost all cases whether or not NAAT-positive women were included. The single exception was Medac, after excluding currently positive women, when later reductions in seropositivity also became significant. Therefore, the evidence was quite consistent in showing that the main fall in seropositivity in women happened in the early weeks and months after the last detected chlamydia episode. There was no clear pattern of further reduction in seropositivity after 6 months (table 2 and figure 1). Seropositivity seemed to decline with a similar time-course regardless of the test used.

    View this table:
    Table 2

    Effect of demographic and clinical presentation on chlamydia antibody detection

    Effect of demographic and clinical variables on assay sensitivity

    Table 2 details the effect of age, ethnicity, number of previous episodes of chlamydia, prior infection with gonorrhoea, previous episode(s) of chlamydia PID, number of sexual partners in the previous 3 months, testing method and whether chlamydia had been detected outside the source clinic. The most important factor affecting pgp3 seropositivity in women was the number of previous episodes of chlamydia infection (two or more vs one, OR 19, 95% CI 2.5 to 146), p<0.001), tables 2 and 3 and figure 1b. Multiple previous episodes were also associated with increased seropositivity in other tests, but the effect was significantly smaller for the Anilabs, Medac and SeroCT MOMP peptide ELISAs (OR for effect 2.1, 2.4, 2.8; p for interaction 0.028, 0.045, 0.055, respectively) and insignificantly smaller for the MIF (OR 6.1; p for interaction 0.34) compared with the pgp3 ELISA (table 2 and see online supplementary appendix 1, figure). Seropositivity, surprisingly, appeared lower in women reporting ≥2 sexual partners in the previous 3 months; this difference was just significant for pgp3 (57% vs 75%; p=0.017) but not the other assays. Age, ethnicity, prior infection with gonorrhoea, previous episode(s) of chlamydia PID, and whether chlamydia had been detected outside the source clinic were not significantly associated with differences in seropositivity with any of the assays under test. However, numbers were small and CIs were wide in some cases, so this does not definitively exclude the possibility that these factors may have some influence over seropositivity.

    View this table:
    Table 3

    Logistic regression models for pgp3 seropositivity

    Women at the Bristol GUM clinic were significantly more likely to be pgp3 antibody-positive than women at the St Mary's GUM clinic. In univariate analysis, the OR for Bristol was 2.9 (95% CI 1.4 to 6.2), p=0.004 (see online supplementary appendix table). The association was at least partly explained by differences between the clinics in the proportion of women with multiple previous chlamydia episodes and shorter (≤6 months) time since last chlamydia detection and treatment (see online supplementary appendix table); there was no difference between the clinics in the proportion of women having had multiple partners in the previous 3 months. When these factors were included in the model, the OR for Bristol GUM clinic dropped to 2.3 (95% CI 1.0 to 5.6) and its significance became borderline (p=0.051).

    Significant predictors for pgp3 seropositivity in a combined model were multiple previous chlamydia infections, more than 6 months since the last diagnosed chlamydia episode, and more than one sexual partner in the previous 3 months (table 3). There was also evidence that the antibody response was more durable in patients having had multiple previous infections (figure 1b). It was not possible to investigate this in detail by logistic regression because so few women with multiple prior infections were seronegative and, of them, none was within 6 months of previous infection.

    Discussion

    We have demonstrated that, as time since initial infection increased, the proportion of women with detectable chlamydia antibody decreased. The sensitivity of the pgp3 ELISA antibody was greater than that of the Anilabs, Medac and SeroCT MOMP peptide ELISAs, at both early (≤6 months) (92% vs 89%, 66% and 79%, respectively) and later times (>6 months) (68% vs 50%, 38% and 47%, respectively) since last detection of chlamydia. This was also true for the MIF assay. In addition, if women had experienced more than one diagnosed episode of chlamydia, the proportion of women with detectable antibody increased. This was most notable for the pgp3 and MIF assays, where 97% and 92% of women with more than one episode had detectable antibody, compared with 65% and 65% of those with one episode only. The effect of having more than one episode of chlamydia infection was significantly smaller when the MOMP peptide ELISAs were employed.

    This large study compared four of the most sensitive ELISA assays, all of which are highly specific (>95%), as demonstrated using sera taken from 722 children who had not been previously exposed to chlamydia and had been characterised for presence of previous infection with C pneumoniae status.11 We included a MIF assay in order to compare directly with other studies in which this assay has been used.9 ,12 The specificity of the MIF assay was also >95% when tested using the same aforementioned sera from children (unpublished data).11 This may reflect the fact that the assay is depleted of the genus-specific lipopolysaccharide antigen which is also expressed in C pneumoniae.10 ,12 Nevertheless, MIF assays have the disadvantage of being subjective and are not designed for high throughput.12 ,13 The original study design included recruitment of women at time of treatment with repeat serology at 3 and 6 months to investigate the change in seropositivity with time following infection. However, recruitment to this part of the study proved difficult. There are a number of potential reasons for this. We found young people, in general, reluctant to commit to serial venepuncture specimens over a 6-month period. This may represent general reluctance to reattend a department of GUM following an episode of chlamydia perhaps as a result of stigma.14 ,15 Having repeat venepunctures undertaken, which can be painful, may have also played a role. Venepuncture is routinely taken at the time of chlamydia testing as part of a routine sexual health screen. We were unable, for logistical reasons, to store all sera from women tested for chlamydia, thus, to enrol in this part of the study, many would have required 4 venepunctures. We therefore discontinued this part of the study and focused on recruiting women infected previously in order to compare the performance of the four assays. A cross-sectional design is methodologically weaker than a prospective cohort design as there are a number of potential biases we would be unable to control for. Nevertheless, to our knowledge, this analysis has not been undertaken previously and we believe these results will help inform both the interpretation and design of future studies using chlamydia serology. Only two clinics contributed, and there were clear differences between them, so some caution is advised in generalising these results directly to national level. It is possible that misclassification may have occurred for those patients who reported a chlamydia diagnosis elsewhere. However, we observed no significant difference in assay sensitivity between those in whom a previous chlamydia diagnosis was recorded in the notes compared with those who reported being diagnosed elsewhere. Finally, it is likely 10–20% of individuals will become reinfected within 1 year, and it is probable that not all will have presented for retesting, so this study may have underestimated the number of individuals who had two or more infections.16

    Interpreting previous studies involving chlamydia serology is complex, as assay sensitivity and specificity have not been rigorously defined, and many assays are known to cross-react with C pneumoniae.10 ,12 Few studies have looked at decline in antibody titre over time since infection.10 ,17 The presence of chlamydia PID is associated with high antibody titres, which might be expected to remain detectable as the time since infection lengthens.18 ,19 Although our data do not show that such women are more likely to be antibody-positive, the numbers are small and we cannot exclude this possibility. Another potential explanation is that, as the clinical diagnosis of PID is known to be difficult,1 it is possible that some women diagnosed with PID did not have upper genital tract disease. One possible reason for the apparent smaller increase in seropositivity in women with two or more infections using the MOMP peptide ELISAs is that women may have been reinfected with a different serovar containing MOMP epitopes to which they had not been exposed in the original infection.20 ,21 There is some evidence that MOMP peptide assays may not have the same sensitivity for detecting antibody to all serovars,22 and we have recently demonstrated that while Medac and SeroCT assays have similar sensitivities in a community setting, some seropositive individuals are only detected by a single assay.23 We also observed an apparent greater probability of detecting chlamydia antibody with all assays for women attending Bristol GUM compared with St Mary's GUM departments. The women recruited at these two centres differed significantly in the number of previous infections and the time since last infection, which partially explains this difference. Another factor is that when the study was undertaken, the majority of women attending St Mary's GUM open access service could attend within 48 h. However, only approximately 40% of women attending Bristol, covering Avon, could be seen within 48 h, while over 20% waited longer than 2 weeks.24 This would suggest that individuals at Bristol may have had a longer duration of untreated infection prior to testing, compared with those attending St Mary's Hospital. Early antibiotic therapy may arrest the immune response, and thus, antibody production.25 ,26 This may also explain why women with >2 sexual partners in the previous 3 months are less likely to be antibody-positive, as clinical experience suggests that these women may be more likely to present sooner for testing following infection.

    Our results suggest that the proportion of women with chlamydia antibody following infection is a function of time since exposure and the number of infective episodes. A larger study in sera drawn from a much larger sample of centres would be valuable to confirm our findings and define more precisely the relationship between seropositivity and timing and frequency of prior exposure to chlamydia. Ideally, this should be prospective, involve both known chlamydia-positive and chlamydia-negative individuals, and undertaken in a community setting. The development of sensitive and specific assays which do not require venous blood, for instance, saliva would probably facilitate recruitment and retention in such a study. Nevertheless, our findings clearly demonstrate that age-specific seroprevalence is a potential measure of cumulative population infection. In addition, the more closely the age-specific seroprevalence is to the age of sexual debut the closer it will reflect actual population prevalence. Thus, changes in age-specific seroprevalence are likely to be most informative about chlamydia epidemiology in the early years following sexual debut compared with older women.

    Key messages

    • Chlamydia antibody decreases in the first 6 months following infection then plateaus.

    • Chlamydia antibody is more readily detectable in women with multiple episodes of infection.

    • Prospective population-based studies would be desirable to enable more accurate characterisation of how chlamydia antibody changes with time since infection.

    Acknowledgments

    We wish to thank Shaw Taylor, Angela Taylor, the staff of the Bristol Sexual Health Centre, Bristol, and the staff of the Jefferiss Wing of St Mary's Hospital, London.

    References

    1. ↵
      Centers for Disease Control and Prevention. Sexually Transmitted Diseases Treatment Guidelines, 2010. http://www.cdc.gov/std/treatment/2010/STD-Treatment-2010-RR5912.pdf
    2. ↵
      National Chlamydia Screening Programme. Health Protection Agency. 2011. http://www.chlamydiascreening.nhs.uk/ps/index.asp
    3. ↵
      1. Ward H,
      2. Fredlund H,
      3. Gotz H,
      4. et al
      . ECDC Guidance: Chlamydia control in Europe. 2009. http://www.ecdc.europa.eu/en/publications/Publications/0906_GUI_Chlamydia _Control_in_Europe.pdf
    4. ↵
      1. Low N
      . Screening programmes for chlamydial infection: when will we ever learn? BMJ 2007;334:725–8.
      OpenUrlFREE Full Text
    5. ↵
      Genital Chlamydia trachomatis tests and diagnoses in young adults in England, 2011. Health Protection Report. 1 June 2012:6(22).
    6. ↵
      House of Commons Committee of Public Accounts. Young people's sexual health: the National Chlamydia Screening Programme. 2010. http://www.publications.parliament.uk/pa/cm200910/cmselect/cmpubacc/283/283.pdf
    7. ↵
      1. Miller WC
      . Epidemiology of chlamydial infection: are we losing ground? Sex Transm Infect 2008;84:82–6.
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Low N
      . Caution: chlamydia surveillance data ahead. Sex Transm Infect 2008;84:80–1.
      OpenUrlFREE Full Text
    9. ↵
      1. Price MJP,
      2. Ades AP,
      3. Welton NJP,
      4. et al
      . How much tubal factor infertility is caused by chlamydia? Estimates based on serological evidence corrected for sensitivity and specificity. Sex Transm Dis 2012;39:608–13.
      OpenUrlCrossRefPubMed
    10. ↵
      1. Johnson AM,
      2. Horner P
      . A new role for Chlamydia trachomatis serology? Sex Transm Infect 2008;84:79–80.
    11. ↵
      1. Wills GS,
      2. Horner PJ,
      3. Reynolds R,
      4. et al
      . Pgp3 antibody enzyme-linked immunosorbent assay, a sensitive and specific assay for seroepidemiological analysis of chlamydia trachomatis infection. Clin Vaccine Immunol 2009;16:835–43.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Persson K
      . The role of serology, antibiotic susceptibility testing and serovar determination in genital chlamydial infections. Best Pract Res Clin Obstet Gynaecol 2002;16:801–14.
      OpenUrlCrossRefPubMed
    13. ↵
      1. Peeling RW,
      2. Wang SP,
      3. Grayston JT,
      4. et al
      . Chlamydia pneumoniae serology: interlaboratory variation in microimmunofluorescence assay results. J Infect Dis 2000;181(Suppl 3):S426–9.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Duncan B,
      2. Hart G,
      3. Scoular A,
      4. et al
      . Qualitative analysis of psychosocial impact of diagnosis of Chlamydia trachomatis: implications for screening. BMJ 2001; 322:195–9.
      OpenUrlAbstract/FREE Full Text
    15. ↵
      1. Lorimer K,
      2. Reid ME,
      3. Hart GJ
      . ‘It has to speak to people's everyday life...”: qualitative study of men and women's willingness to participate in a non-medical approach to Chlamydia trachomatis screening. Sex Transm Infect 2009; 85:201–5.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Hosenfeld CBM,
      2. Workowski KAM,
      3. Berman SM,
      4. et al
      . Repeat infection with Chlamydia and gonorrhea among females: a systematic review of the literature. Sex Transm Dis 2009;36:478–89.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Clad A,
      2. Freidank HM,
      3. Kunze M,
      4. et al
      . Detection of seroconversion and persistence of Chlamydia trachomatis antibodies in five different serological tests. Eur Jo Clin Microbiol Infect Dis 2000;19:932–7.
      OpenUrlCrossRef
    18. ↵
      1. Taylor-Robinson D,
      2. Stacey CM,
      3. Jensen JS,
      4. et al
      . Further observations, mainly serological, on a cohort of women with or without pelvic inflammatory disease. Int J STD AIDS 2009;20:712–18.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Svensson L,
      2. Westrom L,
      3. Ripa KT,
      4. et al
      . Differences in some clinical and laboratory parameters in acute salpingitis related to culture and serologic findings. Am J Obstet Gynecol 1980;138:7–21.
      OpenUrl
    20. ↵
      1. Geisler WM,
      2. Suchland RJ,
      3. Whittington WL,
      4. et al
      . The relationship of serovar to clinical manifestations of urogenital Chlamydia trachomatis infection. Sex Transm Dis 2003;30:160–5.
      OpenUrlPubMedWeb of Science
    21. ↵
      1. Eckert LO,
      2. Suchland RJ,
      3. Hawes SE,
      4. et al
      . Quantitative Chlamydia trachomatis cultures: correlation of chlamydial inclusion-forming units with serovar, age, sex, and race. J Infect Dis 2000;182:540–4.
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Weill FX,
      2. Le Hello S,
      3. Clerc M,
      4. et al
      . Serological reactivity and bacterial genotypes in Chlamydia trachomatis urogenital infections in Guadeloupe, French West Indies. Sex Transm Infect 2009;86:101–5.
      OpenUrlPubMedWeb of Science
    23. ↵
      1. Horner P,
      2. Wills G,
      3. Pebody R,
      4. et al
      . A comparative study of Chlamydia trachomatis PGP3 antibody elisa and two C. trachomatis specific momp peptide assays on sera from young people in the community. Chlamydial infections: Proceedings of the Twelfth International Symposium on human Chlamydial Infections; 2010; San Francisco, CA 94110: International Chlamydia Symposium, 2010:453–6.
    24. ↵
      1. Oliver I,
      2. Camm J
      . Access to Genitourinary medicine clinics. South West STI Task Force Quarterly Bulletin 2005. http://www.hpa.org.uk/web/HPAwebFile/HPAweb_C/1194947411400
    25. ↵
      1. Rekart ML,
      2. Brunham RC
      . Epidemiology of chlamydial infection: are we losing ground? Sex Transm Infect 2008;84:87–91.
      OpenUrlAbstract/FREE Full Text
    26. ↵
      1. Brunham RC,
      2. Pourbohloul B,
      3. Mak S,
      4. et al
      . The unexpected impact of a Chlamydia trachomatis infection control program on susceptibility to reinfection. J Inf Dis 2005;192:1836–44.
      OpenUrlAbstract/FREE Full Text

    Supplementary materials

    • Supplementary Data

      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

    • Contributors PJH, MOMc and AJ conceived the original study; PJH and AW were involved in obtaining samples; GSW and AJB carried out the chlamydia antibody assays, RR and PJH undertook the data analysis; PJH, RW, MOMcC and DP were involved in the initial data interpretation; PH prepared the first draft of the paper and all authors commented and contributed to subsequent revisions.

    • Funding This work was funded by the Medical Research Council, UK (Grant no.s G050004 and G0500152). We are also grateful for support from the NIHR Biomedical Research Centre funding scheme.

    • Competing interests PJH and MOMcC hold a patent through Imperial College London for the use of pgp3 antibody alone or in combination with other antigens to determine whether an individual is at risk of chronic sequelae following Chlamydia trachomatis infection. Patent number 1111673.8. This patent does not cover the use of pgp3 antibody to detect chlamydia antibody as a marker of previous infection.

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