OVER THE LAST DECADE, SEXUALLY transmitted diseases (STDs) are again increasing in many European countries.1 Seen mainly among men having sex with men (MSM),2–5 the rise is probably explained by the introduction of highly active antiretroviral therapy in 1996, which has decreased the perceived threat of HIV and led to increased sexual risk behavior.5,6 Similar processes among heterosexuals might be inferred. In The Netherlands, the total number of chlamydia infections increased considerably among heterosexuals as registered between 2002 and 2004 in the Dutch surveillance system.7 Also, in the United Kingdom, the number of new chlamydia and gonorrhea diagnoses increased since 1996, particularly among women and men aged under 25 years.8 In addition, since 2000, the total number of HIV diagnoses in heterosexuals rose to the same level as in MSM in The Netherlands.7 This, combined with an increasing tendency for unsafe sex, especially among young heterosexuals,9,10 likewise suggests increasing STD incidence.
To evaluate whether STDs are rising among heterosexuals, like among MSM, we explored STD and HIV time trends among heterosexuals attending the outpatient STD clinic of the Health Service of Amsterdam between 1994 and 2005. In addition, we aimed to define high-risk STD core groups by assessing risk factors of STD infections to enable targeting of prevention messages among heterosexuals in The Netherlands.
Materials and Methods
The outpatient STD clinic of the Health Service of Amsterdam offers free-of-charge examination and treatment for STDs. It diagnoses approximately 50% of all STDs recorded in The Netherlands7 and takes approximately 20,000 new consultations per year. Most are by heterosexual attendees (defined as having sex exclusively with someone of the opposite sex in the past 6 months), and data from each new consultation made by a heterosexual attendee between 1994 and 2005 were included in this study. We defined a new consultation as a first-time clinic visitor or a visitor already known but seeking care for new symptoms or seeking asymptomatic STD testing.
Procedure and Laboratory Testing
All attendees are routinely screened for Chlamydia trachomatis (CT), Neisseria gonorrhea (NG), and Treponema pallidum (syphilis) at every new consultation. Until 2004, all male attendees were screened for urethral CT and NG, and all female attendees for cervical and/or urethral CT and NG plus pharyngeal and anal NG during an extensive physical examination. If women reported receptive anal intercourse, they were also screened for anal CT. Culture tests (GC lect; Becton & Dickinson Diagnostics, Franklin Lakes, NY) were used for NG diagnosis. For CT diagnosis, ligase chain reaction (Abbott Diagnostics Division, Abbott Park, IL) has been used since July 1995 and polymerase chain reaction (PCR; Cobas Amplicor; Hoffman-La Roche, Basel, Switzerland) since July 2002. Before 1995, cultures were used for CT diagnosis. Since April 2004, a distinction between high- and low-risk attendees is made by a prioritizing system based on six questions that describe recent sexual behavior, STD-related complaints, and previous STD episodes. High-risk attendees are assigned to a standard screening protocol, which has been the routine screening before 2004 as described previously. Low-risk attendees are assigned to a short screening protocol, which includes self-collection of urine (males) or a vaginal swab (females), which are tested for CT and NG using PCR (Cobas Amplicor CT and NG; Roche Diagnostics NV, Basel, Switzerland). For NG, confirmation is performed by an in-house-developed real-time PCR11,12 because the results of the PCR are often false-positive.13 Until 2004, all female attendees were tested for Trichomonas vaginalis (TV) (Trichosel; Tritium Microbiologie BV, Veldhoven, The Netherlands). After April 2004, only females in the standard—or high-risk—screening are tested for TV. Syphilis testing is done for all attendees using the Treponema pallidum particle agglutination assay (Fujirebio, Tokyo, Japan); when reactive, the RPR-nosticon II (rapid plasma reagin; Biomérieux, Marcy l'Etoile, France) and FTA-absorption test (Trepo-spot IF, Biomérieux), and the Venereal Disease Research Laboratory test (Wellcome, Dartford, U.K.) are performed to confirm and classify syphilis infection. Human immunodeficiency virus (HIV) antibody testing (Axsym; Abbott Laboratories, North Chicago, IL; Abbott Determine HIV 1/2 since 2004 in the standard—or high-risk—screening) is not routinely performed but is offered to all attendees since 1999. Since 2004, reactive samples are confirmed by immunoblot (Line Immuno Assay, Inno-Lia HIV I/II Score; Innogenetics NV, Ghent, Belgium). Before that date, reactive samples were confirmed by Western blot. All STD diagnoses and subsequent treatment are recorded in an electronic patient database along with patient characteristics such as gender, age, nationality, and ethnicity (defined as the population group with which someone identifies) and information on sexual orientation, injecting drug use, being a commercial sex worker (CSW), or visiting a CSW in the past 6 months.
Besides information collected during routine STD consultations, we used data collected in half-yearly cross-sectional anonymous HIV prevalence surveys at the clinic to assess HIV prevalence.14 In addition, these surveys provided information on the number of partners, which was the only information available on changes in sexual behavior over time. During each survey, approximately 1,000 attendees, representing a representative sample of clinic attendees, are enrolled and interviewed on HIV risk behaviors and HIV testing after informed consent. Interviews are followed by routine physical examination, including STD and optional nonanonymous (i.e., named) HIV testing. Additional blood is taken for anonymous HIV antibodies testing using commercially available enzyme-linked immunosorbent assay with positive results confirmed by Western blot.14
We used routinely tested STDs to evaluate trends over the years 1994 to 2004. These included anogenital NG, anogenital CT (since 1996), infectious syphilis (primary, secondary, and early latent), and TV among women (until 2004). Infections at multiple anatomic sites at the same visit were counted as one case, but a single person could contribute to multiple cases if a repeated infection was diagnosed at more than one consultation. For symptomatic infections such as syphilis and NG, changes in total numbers of infections might represent changes in incident infections and therefore probably changes in incidence in Amsterdam.15 For asymptomatic infections such as CT and HIV, total numbers do not necessarily represent incident cases and increases might by confounded by increases in screening, making changes in prevalences more informative than total numbers for generalization to the Amsterdam population.16 To aid interpretation, we therefore analyzed time trends in both prevalence (i.e., number of infections divided by number of visits) and total number of infections. We analyzed prevalence time trends univariately using logistic regression. For total numbers of infections, we used generalized linear models with a negative binomial distribution assuming a stable population over calendar time. We chose a negative binomial distribution because we wanted to adjust for overdispersion to correct for unmeasured variables. In both logistic and negative binomial models, we modeled year as a continuous variable and, in case of nonlinearity, year was modeled by a piecewise linear effect based on percentiles. Nonlinearity was tested by comparing the fit of the linear and the piecewise linear model. To evaluate whether STD prevalence trends differed among demographic subgroups, we added interaction terms between calendar time and gender, age, and ethnicity in bivariate logistic analysis. Interaction, however, could only be evaluated for STD trends on prevalence because for the total number of infections, we would need to know the size of each risk group, which was unknown. For HIV trends, prevalence as found in the anonymous HIV surveys was used to evaluate trends (available until 2004) by logistic regression as described previously. Trends in sexual risk behavior, defined as only one partner versus more than one partner, were analyzed using univariate logistic regression with year as a continuous variable or, in case of nonlinearity, year was modeled by a piecewise linear effect based on percentiles.
To assess STD risk factors (i.e., gender, age, ethnicity, drug use, HIV status, and being a CSW or visiting a CSW), we performed univariate and multivariate logistic regression analysis. Backward selection was used to obtain a multivariate model that included statistically significant risk factors only. Significance for individual parameters was tested using the univariate and multivariate Wald test. Estimates and standard errors were corrected for intraindividual correlation among visits of the same person by using generalized estimating equations, assuming an exchangeable correlation matrix. To correct for demographic changes over time, we also adjusted for year of visit. We did not test for interactions when we assessed risk factors, being interested only in basic demography. We considered a P value <0.05 as statistically significant.
The annual number of new consultations by heterosexuals almost doubled between 1994 and 2005, giving a total of 161,510 new consultations during the study period (Fig. 1). Of the heterosexual attendees, 47% was male and 60% was of Dutch origin; 12% was of Surinamese or Dutch Antillean descent. The median age was 28 years (interquartile range: 23–35 years). The demographics changed over time, with the proportion of women, the proportion of persons aged younger than 30, and the proportion of persons with Dutch ethnicity increasing steadily (from 48–53%, from 47–66%, and from 50–67%, respectively). Since 1994, diagnoses included 4,055 NG infections (529 anal NG), 346 syphilis infections, and 2,236 TV infections (until 2004); and since 1996, 14,331 CT infections (620 anal CT). In total, 17,093 attendees participated in the anonymous HIV surveys from 1994 to 2004; 108 were found to be HIV-positive.
Trends in Prevalence
In univariate analysis, the CT prevalence showed a small increase in the period 1999 through 2002 (odds ratio [OR] = 1.05 per year, 95% confidence interval [CI] = 1.03–1.07). However, this increase was not sustained and the prevalence has declined since then (OR = 0.98 per year, 95% CI = 0.96–0.998). There were no significant differences in prevalence time trends between men and women (Fig. 2, p interaction [pi]sex = 0.1). However, a significant interaction between ethnicity and year was found (pi ethnicity <0.001); for persons with a Surinamese or Dutch Antillean ethnicity, time trends differed from the overall trend described previously because in this group, an increase in CT prevalence was seen after 2002 (OR = 1.05 per year, 95% CI = 1.001–1.10, Fig. 3). In addition, there was an interaction effect between age and calendar time (pi age = 0.01). In attendees aged 30 years and over, the annual CT prevalence decreased significantly after 2002 (OR = 0.92 per year, 95% CI = 0.88–0.96), whereas in the two other age groups, it showed a nonsignificant decrease or remained stable (Fig. 3). In women, the anal CT prevalence showed a significant linear increase over time (OR = 1.08 per year, 95% CI = 1.05–1.11). Similar to the CT prevalence, the NG prevalence showed a short-term increase over time, which was also not sustained. After an increase in 1998 to 2002 (OR = 1.19 per year, 95% CI = 1.16–1.22), the NG prevalence decreased (OR = 0.82 per year, 95% CI = 0.79–0.85). Prevalence time trends did not differ between men and women and for the various age groups (Figs. 2 and 3, pi sex = 0.4 and pi age = 0.13, respectively). Despite significant interaction for ethnicity (pi ethnicity <0.001), time trends showed similar trends for the various ethnic groups (Fig. 3) with significant declines in all groups after 2002. In women also, the prevalence of anal NG decreased after 2002 (OR = 0.81 per year, 95% CI = 0.73–0.91). In general, the syphilis prevalence showed no significant linear time trend, being 0.11% in 1994, 0.29% in 1996, and 0.20% in 2005 (OR = 1.01 per year, 95% CI = 0.98–1.05, Fig. 4). However, there was a significant increase among attendees aged 30 years and older (OR = 1.09 per year, 95% CI = 1.04–1.14), whereas among younger attendees, the syphilis prevalence appeared to decrease (pi age <0.001, OR for persons aged 20–30 years = 0.95 per year, 95% CI = 0.90–0.998; OR for persons aged <20 = 0.89 per year, 95% CI = 0.76–1.04). The TV prevalence decreased significantly over the years (Fig. 2, P <0.001). However, this decrease was only significant in women of Dutch or Surinamese or Dutch Antillean background (pi ethnicity <0.001). Except for syphilis, controlling for demographics revealed similar time trends (Table 1). The anonymous HIV survey showed a significant linear decline in HIV prevalence among heterosexual STD clinic attendees (Fig. 4, OR = 0.93 per year, 95% CI = 0.87–0.99).
Trends in Total Number of Infections
Univariate analysis of the total numbers of NG and TV infections showed comparable time trends as seen for prevalence (Fig. 2). The total number of TV infections decreased linearly over time (incidence rate ratio [IRR]: 0.94 per year, 95% CI = 0.90–0.98). For NG total infections, similar to the NG prevalence, there was a short-term increase (IRR: 1.21 per year, 95% CI = 1.12–1.32) followed by a decline (IRR: 0.92 per year, 95% CI = 0.85–1.00). For syphilis, the total number of infections increased significantly from 1994 to 1997 (IRR: 1.37 per year, 95% CI = 1.11–1.71) but did not show any significant changes afterward (Fig. 4). However, for CT, we observed differences between trends in prevalence and total number of infections. The number of CT infections increased linearly over time (IRR: 1.04 per year, 95% CI = 1.01–1.08).
Risk Factors for Sexually Transmitted Diseases
Men were at a higher risk of CT, NG, and syphilis infection compared with women (Table 1). Younger age was a significant risk factor for both NG and CT, whereas older age was a risk factor for syphilis and TV. Persons of Surinamese or Dutch Antillean background were at higher risk of any STD compared with persons with a Dutch background. Drug use and HIV positivity were significant risk factors for NG, syphilis, and TV, but not for CT. Importantly, not being tested for HIV was in itself a significant risk factor for an STD diagnosis.
Number of Partners
The percentage of heterosexual attendees participating in the anonymous HIV surveys reporting more than one partner in the 6 months before STD consultation fluctuated over time. There was an increase from 1994 to 1998 (OR = 1.04 per year, 95% CI = 1.0–1.08), a decease from 1998 to 2001 (OR = 0.93 per year, 95% CI = 0.89–0.96), and an increase again from 2001 to 2004 (OR = 1.07 per year, 95% CI = 1.03–1.1), being 61% in 1994, 66% 1997, 61% in 2000, and 67% in 2004.
Despite increases in the number of heterosexual STD clinic attendees, we found no evidence for an increase in STD as seen among MSM. For example, the total number of gonorrhea infections only showed a short-term increase, which was not sustained after 2002, being 227 in 1994 and 190 in 2005 among heterosexual men. However, among MSM attending the STD clinic Amsterdam, the number of gonorrhea infections increased substantially, being 119 in 1994 and 559 in 2005.17 The total number of syphilis infections in heterosexuals increased from 1994 to 1997 but remained stable afterward with approximately 30 infections per year. In MSM clinic attendees, the number of syphilis infections was low in 1998 with only 9 infections but increased to 213 in 2005.17 Besides the only small and short-term increases in syphilis and gonorrhea infections, both the HIV and trichomonas prevalence decreased over time among heterosexuals.
The total number of chlamydia infections, however, showed a continuous long-term increase. For chlamydia, drawing conclusions is problematic, because it is largely asymptomatic and widely distributed in the population.16,18 New diagnoses do not necessarily represent incident cases and increases in total numbers of infections might be the result of prevalent infections found resulting from increased screening. This phenomenon might partially explain the STD increases, particularly in chlamydia, seen in the United Kingdom and in The Netherlands.7,8,19 In contrast, syphilis and gonorrhea infections tend to be confined to high-risk groups and are largely symptomatic.15,18,20 In Amsterdam, syphilis is rarely diagnosed by general practitioners, who refer such cases to the STD clinic. New infections diagnosed at the STD clinic are thus likely to represent incident cases, and trends in total number of infections probably correspond to trends in incidence in Amsterdam.
The number of STD clinic attendees increased considerably since 1994. This could reflect increasing STD prevalence and risky sexual behavior in the community, but also increasing awareness for STD and STD testing. Part of the increase in STD clinic attendees between 2003 and 2005 can be explained by changes in screening practices and subsequently increased screening capacity and increased STD testing acceptability resulting from availability of self-swabs in low-risk heterosexuals. Previously, screening capacity was insufficient to meet visitor demand and persons often had to be sent away without being tested. In April 2004, a prioritizing system was implemented assigning low-risk attendees to a short screening protocol, including self-collection of urine or vaginal swabs, and high-risk attendees to a more extensive screening protocol, including physical examination and counseling (Heijman et al.; In press). Most of the increase in STD clinic attendees was among persons with Dutch ethnicity. Persons of Dutch ethnicity were among the lowest at risk for any STD in our population. Increases in attendance of persons of low risk—or the “worried well”—might result in an underestimation of STD trends and their STD risk. Conversely, the attendance of persons of a non-Dutch ethnic group who come mainly when having symptoms will lead to their underrepresentation and, in turn, an overestimation of the STD risk within their ethnic group. Then again, if risk behavior and STDs would be on the rise in general among heterosexuals, larger increases in attendees at higher risk such as persons of Dutch Antillean or Surinamese ethnicity would be expected. This seems in part true for the increase in attendees aged below 30, with persons aged below 20 at highest risk for gonorrhea and chlamydia.
Changes in diagnostic tests can further affect STD time trends reported. The most important testing alteration was the replacement of culture for a ligase chain reaction-based test for chlamydia diagnostics in 1995, the latter being considerable more sensitive.22,23 To avoid bias, we only described chlamydia trends as from 1996. In 2002, chlamydia testing changed from ligase chain reaction to PCR, but sensitivity for both systems are comparable.13,23 Low-risk heterosexuals attending after April 2004 perform self-swabs for CT and NG diagnostics, still offering satisfactory sensitivity.24–26
Although drawing conclusions based on STD clinic data as mentioned previously is limited by several factors, we do propose that although the number of persons seeking STD testing over time increased, there is currently no evidence that supports an increase in STDs among heterosexuals as seen among MSM.2–5 We base our conclusions on the continuous decrease in the prevalence and total number of trichomonas infections, the decrease in HIV prevalence, and the relatively stable prevalence and total number of syphilis and gonorrhea infections. Both syphilis and gonorrhea showed a short-term increase in both prevalence and total infections. However, this was not sustained after 2002. Syphilis and gonorrhea are more likely markers of unsafe sexual behavior than chlamydia as a result of the more symptomatic nature of these infections. Therefore, most, if not all, of the increase in total number of chlamydia infections might very well be explained by increased number of STD clinic attendees and thus by increased screening. Unfortunately, besides the number of partners, only providing limited information on sexual risk behavior, no data were available on sexual behavior, supporting our conclusions. Additionally, like reported by others, we identified risk groups for STDs in The Netherlands that would benefit from targeted prevention and screening such as persons of Surinamese or Dutch Antillean background and young persons, although persons aged over 30 were at highest risk for syphilis.7,27,28
1. Fenton KA, Lowndes CM. Recent trends in the epidemiology of sexually transmitted infections in the European Union. Sex Transm Infect 2004; 80:255–263.
2. Van der Heyden JH, Catchpole MA, Paget WJ, et al. Trends in gonorrhoea in nine western European countries, 1991–6. European Study Group. Sex Transm Infect 2000; 76:110–116.
3. Fox KK, del Rio C, Holmes KK, et al. Gonorrhea in the HIV era: A reversal in trends among men who have sex with men. Am J Public Health 2001; 91:959–964.
4. Fennema JS, Cairo I, Spaargaren J, et al. Syphilis epidemic and an increase of the number of HIV infections among homosexual men attending the Amsterdam venereal disease clinic [in Dutch]. Ned Tijdschr Geneeskd 2002; 146:633–635.
5. Stolte IG, Dukers NH, de Wit JB, et al. Increase in sexually transmitted infections among homosexual men in Amsterdam in relation to HAART. Sex Transm Infect 2001; 77:184–186.
6. Stolte IG, Dukers NH, Geskus RB, et al. Homosexual men change to risky sex when perceiving less threat of HIV/AIDS since availability of highly active antiretroviral therapy: A longitudinal study. AIDS 2004; 18:303–309.
7. Laar MJW van de, Boer IM, Koedijk FDH de, et al. RIVM Rapport 441100022: HIV and Sexually Transmitted Infections in the Netherlands in 2004. An Update: November 2005. Bilthoven, The Netherlands: RIVM, 2005.
8. The UK Collaborative Group for HIV and STI surveillance. Mapping the Issues. HIV and Other Sexually Transmitted Infections in the United Kingdom: 2005. London, UK: Health Protection Agency of Infections, 2005.
9. Johnson AM, Mercer CH, Erens B, et al. Sexual behaviour in Britain: Partnerships, practices, and HIV risk behaviours. Lancet 2001; 358:1835–1842.
11. Bruisten SM, Noordhoek GT, van den Brule AJ, et al. Multicenter validation of the cppB gene as a PCR target for detection of Neisseria gonorrhoeae
. J Clin Microbiol 2004; 42:4332–4334.
12. Geraats-Peters CW, Brouwers M, Schneeberger PM, et al. Specific and sensitive detection of Neisseria gonorrhoeae
in clinical specimens by real-time PCR. J Clin Microbiol 2005; 43:5653–5659.
13. van Doornum GJ, Schouls LM, Pijl A, et al. Comparison between the LCx Probe system and the COBAS AMPLICOR system for detection of Chlamydia trachomatis
and Neisseria gonorrhoeae
infections in patients attending a clinic for treatment of sexually transmitted diseases in Amsterdam, The Netherlands. J Clin Microbiol 2001; 39:829–835.
14. Fennema JS, van Ameijden EJ, Coutinho RA, et al. HIV surveillance among sexually transmitted disease clinic attenders in Amsterdam, 1991–1996. AIDS 1998; 12:931–938.
15. Lowndes CM, Fenton KA. Surveillance systems for STIs in the European Union: Facing a changing epidemiology. Sex Transm Infect 2004; 80:264–271.
16. Catchpole MA. The role of epidemiology and surveillance systems in the control of sexually transmitted diseases. Genitourin Med 1996; 72:321–329.
17. Annual Reports of the Sexually Transmitted Disease Outpatient Clinic. Amsterdam, The Netherlands: Health Service of Amsterdam, 1994–2005.
18. van Duynhoven YT, van de Laar MJ, Schop WA, et al. Different demographic and sexual correlates for chlamydial infection and gonorrhoea in Rotterdam. Int J Epidemiol 1997; 26:1373–1385.
19. Brown AE, Sadler KE, Tomkins SE, et al. Recent trends in HIV and other STIs in the United Kingdom: Data to the end of 2002. Sex Transm Infect 2004; 80:159–166.
20. Mehta SD, Erbelding EJ, Zenilman JM, et al. Gonorrhoea reinfection in heterosexual STD clinic attendees: Longitudinal analysis of risks for first reinfection. Sex Transm Infect 2003; 79:124–128.
21. Heijman RLJ, Van der Bij AK, de Vries HJC, et al. Effectiveness of a risk-based visitor-prioritizing system at an STI out-patient clinic. Sex Trans Dis 2007; In press.
22. Buimer M, van Doornum GJ, Ching S, et al. Detection of Chlamydia trachomatis
and Neisseria gonorrhoeae
by ligase chain reaction-based assays with clinical specimens from various sites: Implications for diagnostic testing and screening. J Clin Microbiol 1996; 34:2395–2400.
23. Puolakkainen M, Hiltunen-Back E, Reunala T, et al. Comparison of performances of two commercially available tests, a PCR assay and a ligase chain reaction test, in detection of urogenital Chlamydia trachomatis
infection. J Clin Microbiol 1998; 36:1489–1493.
24. Schachter J, Chernesky MA, Willis DE, et al. Vaginal swabs are the specimens of choice when screening for Chlamydia trachomatis
and Neisseria gonorrhoeae
: Results from a multicenter evaluation of the APTIMA assays for both infections. Sex Transm Dis 2005; 32:725–728.
25. Schachter J, McCormack WM, Chernesky MA, et al. Vaginal swabs are appropriate specimens for diagnosis of genital tract infection with Chlamydia trachomatis
. J Clin Microbiol 2003; 41:3784–3789.
26. Knox J, Tabrizi SN, Miller P, et al. Evaluation of self-collected samples in contrast to practitioner-collected samples for detection of Chlamydia trachomatis, Neisseria gonorrhoeae
, and Trichomonas vaginalis
by polymerase chain reaction among women living in remote areas. Sex Transm Dis 2002; 29:647–654.
27. Gotz HM, van Bergen JE, Veldhuijzen IK, et al. A prediction rule for selective screening of Chlamydia trachomatis
infection. Sex Transm Infect 2005; 81:24–30.
© Copyright 2007 American Sexually Transmitted Diseases Association
28. van Bergen J, Gotz HM, Richardus JH, et al. Prevalence of urogenital Chlamydia trachomatis
increases significantly with level of urbanisation and suggests targeted screening approaches: Results from the first national population based study in the Netherlands. Sex Transm Infect 2005; 81:17–23.