GENITAL INFECTIONS CAUSED BY Chlamydia trachomatis are common among young, sexually active women in European countries1 and in the United States.2,3 These infections are most often asymptomatic but have potential long-term effects on fecundity.4
The link between genital C. trachomatis infections and ectopic pregnancy has mainly been established through seroepidemiologic cross-sectional studies in which selected groups of patients with ectopic pregnancy have been compared with women with normal pregnancies5–11 or with nonpregnant controls.12–14
Two previous prospective registry-based studies from the United States15 and Denmark16 report ectopic pregnancy after a diagnosed C. trachomatis infection. In the U.S. study, women with multiple infections were found to have a higher risk for ectopic pregnancy compared with women with a single infection. In the Danish study, however, women with a positive test result were found to be at lower risk for subsequent ectopic pregnancy.
The aim of the present population-based, nested case–control study was to assess C. trachomatis as a risk factor for ectopic pregnancy.
Materials and Methods
The study was carried out in Sør-Trøndelag County, central Norway. Of the county's 270,000 inhabitants, approximately 150,000 are living in the major city, Trondheim.
A unique 11-digit personal identification number is given to all Norwegians at birth and to new residents of Norway. In the current study, we used 2 databases, both containing the personal identifier and data on residency.
In Sør-Trøndelag County, a single laboratory is responsible for all C. trachomatis diagnostics.17 The C. trachomatis database contains information on all tests (date, diagnostic method, and test outcome) in Sør-Trøndelag from November 1990 to December 2003 with person as the unit of analysis. The database comprised data from 90,309 women, of whom 61,283 were residents of Sør-Trøndelag. Data on specimens collected within 60 days from a previous test were excluded from analysis.
The database on ectopic pregnancy covers all women hospitalized with this diagnosis in 1970 to 2004 at the 2 hospitals in the county (Orkdal Hospital and Trondheim University Hospital). Diagnoses of ectopic pregnancy were identified through hospital in- and outpatient registries. All medical records were reviewed, and only women with a verified diagnosis of ectopic pregnancy were eligible as cases.18
Several methods were used for C. trachomatis detection throughout the study period.17 Briefly, the IDEIA Chlamydia Test (Celltech Diagnostics/Novo BioLabs/DAKO) was replaced by PACE 2 Gen-probe in 1992, whereas Amplicor (Roche Molecular Systems) became the routine detection method in 1999.
Case and Control Definition
From November 1, 1990, to December 31, 2004, 792 women had their first diagnosis of ectopic pregnancy. We linked the ectopic pregnancy database to the C. trachomatis database by means of the personal identifier.
The reference year was the year of the case's ectopic pregnancy. When at least one prior C. trachomatis test was registered before the ectopic pregnancy date, we selected 3 controls from the C. trachomatis database. Women residing in Sør-Trøndelag with no diagnosis of ectopic pregnancy were eligible as controls. Controls were matched to cases for year of birth (±1 year), month and year of first registered C. trachomatis test (±1 month), and number of prior C. trachomatis tests (cases: number of tests before the ectopic pregnancy, controls: number of tests by the end of the reference year). Among eligible controls, we chose those with the date of the first C. trachomatis test closest to the date of the first test for the case. The number of prior tests was matched exactly up to 5 tests, whereas 6 or more tests were treated as one category.
The data were analyzed using SPSS for Windows, version 11.0 (SPSS, Chicago, IL). We chose the nested case–control design to be able to match for testing pattern (age at first test and number of tests before reference year). Odds ratios for matched data were calculated by using conditional logistic regression. χ2 tests were used for linear trend. For all analyses, P values <0.05 were considered significant.
The study was approved by the Regional Committee for Medical Research Ethics, Central Norway, and the Norwegian Data Inspectorate. Authorization for the use of laboratory data and data retrieved from medical records was obtained from the Norwegian Directorate for Health and Social Affairs.
Among the 792 women presenting with their first ectopic pregnancy from December 1990 to December 2004, 616 (78%) had had at least one prior C. trachomatis test. The proportion of patients with previous tests for C. trachomatis and thus included as cases increased for the younger cohorts (1950–1959 cohorts: 72 of 129 [56%]; 1960–1969 cohorts: 351 of 451 [78%]; 1970–1984 cohorts: 193 of 212 [91%]). The number of tests for each case also increased for the younger cohorts, with just one registered test for 44% of cases born 1950–1959, decreasing to 24% of cases born 1970–1984 (Table 1).
The success of the matching procedure is displayed in Table 2. Cases and controls were similar in age at the end of the study, age at first test, number of prior tests, and time from first test to end of the study. Furthermore, there was no difference between cases and controls in time from last test until the end of the study.
Prior Positive Chlamydia trachomatis Tests
Overall, 11.0% (68 of 616) of cases and 8.5% (157 of 1,848) of controls were registered with at least one positive test before the reference date (Table 3). The proportion of cases and controls registered with a prior positive test was higher in the younger cohorts.
Among women born 1950–1959 and 1960–1969, there was no difference in the cumulative prevalence of prior positive tests between cases and controls. Among women born 1970 and later, however, 24.9% (48 of 193) of cases were registered with at least one prior C. trachomatis infection(s) compared with 15.1% (88 of 579) of controls.
In a conditional logistic regression model, women with at least one C. trachomatis infection had an odds ratio of 1.4 (95% confidence interval [CI], 1.0–2.0) for ectopic pregnancy compared with women with no positive tests (Table 3, model 1). Model 2 suggests a dose–response relationship between previous C. trachomatis infections measured as the number of positive tests and subsequent ectopic pregnancy (χ2 for trend, P = 0.03).
Because data on C. trachomatis tests were not available before November 1990, completeness of testing history differed between birth cohorts. Table 3 displays the association between ectopic pregnancy and outcome of chlamydia test(s) for the cohorts born 1950–1959, 1960–1969, and 1970–1984 (Table 3, models 3–6). Among the younger women, a stronger relationship between previous infections and ectopic pregnancy was found (odds ratio, 2.1; 95% CI, 1.3–3.2) (Table 3, model 5). Furthermore, a statistically significant dose–response relationship was observed (Table 3, model 6).
The present population-based, nested case–control study has demonstrated that women with ectopic pregnancy have higher odds for a history of chlamydial infection than women without ectopic pregnancy.
The study includes all women having their first ectopic pregnancy and all procedures for C. trachomatis diagnostics over a 13-year time period within in a well-defined geographic region (county). Loss of ectopic pregnancy patients was minimal for 2 main reasons. First, all ectopic pregnancies in Norway are treated in hospitals,18 ensuring high quality of case ascertainment. Second, the Trondheim University Hospital serves as regional hospital for central Norway, and treatment of ectopic pregnancy patients resident in Sør-Trøndelag in other counties is very rare (data from the national discharge and outpatient registry, not shown).
A single laboratory covers all C. trachomatis analyses within the region,17 providing complete recording of diagnostics over the entire study period. By matching controls to cases for year of birth, age at first test, and number of tests before reference year of case ascertainment, we eliminated potential bias resulting from differential testing.
We have previously shown that 85% of women in Sør-Trøndelag had been tested for C. trachomatis at least once before the age of 25 and that general testing rates are high.17 The patients in the present study had the same testing pattern as the general female population because 91% of patients born 1970 and later were registered with at least one prior C. trachomatis test before their ectopic pregnancy.
We observed a stronger relationship between prior C. trachomatis infections and ectopic pregnancy when we restricted the analyses to women born 1970 and later, whereas no association was found among women born 1950–1959 and 1960–1969. Routine testing for asymptomatic C. trachomatis infections started on a large scale in 1985 but electronic data recording of diagnostic performance was not introduced before November 1990. Thus, the history of C. trachomatis infection is the most complete for the cohorts born 1970 and later. In the youngest cohorts, fewer than one in 10 patients were without previous C. trachomatis tests and had to be excluded from the study as opposed to nearly half the patients born 1950–1959. Another explanation for the lack of an association among women born before 1970 is that infections caused by Neisseria gonorrhoeae are likely to have played an important role in the pathogenesis of ectopic pregnancy in this age group. By the late 1980s, gonorrheal infections were nearly eradicated in Norway,19 and the younger cohorts have therefore not been exposed to such infections.
Because this is a registry study, we did not have access to data on potential confounders such as sexual behavior, the use of oral contraceptives, or douching. Sexual behavioral characteristics and, to a lesser degree, also the use of hormonal contraceptives are most likely reflected in increased risk for sexually transmitted infections as seen in C. trachomatis exposure in the current study. Furthermore, douching is not of interest to our analyses, because douching is rarely practiced by Norwegian women.
We found a significant dose–response relationship between the number of prior infections and the risk for ectopic pregnancy using women with negative tests results as reference. These results confirm the findings from a U.S. study in which women with multiple C. trachomatis infections were at higher risk for ectopic pregnancy compared with women with a single infection.15 Our historical prospective, nested case–control study resembles the study design applied by Hillis15 in that women with previous infections were followed in an historical prospective design for an event of ectopic pregnancy regardless of other pregnancy outcomes. In a recently published Danish registry linkage study,16 women with a positive test were found to be at lower risk for subsequent ectopic pregnancy compared with women with only negative tests. In the Danish study, there are several issues that all may have caused underestimation of the risk of ectopic pregnancy. Only women with proven fertility were included to the analyses, testing histories were incomplete because most women were registered with only one test, the observation period for fertility outcomes extended the time period for registration of chlamydia exposure by 10 years, and antigen tests with low performance were used. In our study, the oldest cohort (1950–1959) had incomplete testing histories and an odds ratio of 0.35 for ectopic pregnancy by history of a positive chlamydia test, which is close to the overall hazard ratio of 0.55 as published by the Danish group.
Over the time period of our study, the incidence of ectopic pregnancy decreased by nearly 50%,18 whereas prevalence of C. trachomatis at first test has been increasing among women 15 to 24 years of age.17 Similarly, an ecologic study from Australia has shown that when the number of diagnosed chlamydial infections increased over the years 1990 to 2001, ectopic pregnancy rates were constant and hospitalizations for pelvic inflammatory disease fell.20 Although women with ectopic pregnancy more frequently had been exposed to prior chlamydial infection in our study, the proportion of women with such infection who subsequently were diagnosed with an ectopic pregnancy was low. That is, for the 1970–1984 cohort, the cumulative incidence of ectopic pregnancy at age 25 was 4.3 (95% CI, 1.6–6.9) per 1,000 among women with one or more positive tests and 3.2 (95% CI, 2.3–4.1) per 1,000 among women with negative tests, only (data not shown). On the other hand, within the 1970–1984 cohort, 75% of the women with ectopic pregnancy had never been diagnosed with chlamydia despite frequent testing.
In conclusion, our data showed that women with prior C. trachomatis infection are at a 2-fold risk for ectopic pregnancy.
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. Sexually Transmitted Disease Surveillance 2003 Supplement, Chlamydia Prevalence Monitoring Project. Atlanta: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, October 2004.
3. Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance 2004. Atlanta: US Department of Health and Human Services, September 2005.
4. Farquhar CM. Ectopic pregnancy. Lancet 2005; 366:583–591.
5. Brunham RC, Binns B, McDowell J, et al. Chlamydia trachomatis
infection in women with ectopic pregnancy. Obstet Gynecol 1986; 67:722–726.
6. Chow JM, Yonekura ML, Richwald GA, et al. The association between Chlamydia trachomatis
and ectopic pregnancy. A matched-pair, case–control study. JAMA 1990; 263:3164–3167.
7. Coste J, Laumon B, Bremond A, et al. Sexually transmitted diseases as major causes of ectopic pregnancy: Results from a large case–control study in France. Fertil Steril 1994; 62:289–295.
8. Ødland JO, Ånestad G, Rasmussen S, et al. Ectopic pregnancy and chlamydial serology. Int J Gynaecol Obstet 1993; 43:271–275.
9. Osser S, Persson K. Chlamydial antibodies and deoxyribonucleic acid in patients with ectopic pregnancy. Fertil Steril 1992; 57:578–582.
10. Phillips RS, Tuomala RE, Feldblum PJ, et al. The effect of cigarette smoking, Chlamydia trachomatis
infection, and vaginal douching on ectopic pregnancy. Obstet Gynecol 1992; 79:85–90.
11. Tuomivaara LM. Ectopic pregnancy and genital infections: A case–control study. Ann Med 1990; 22:21–24.
12. Brunham RC, Peeling R, Maclean I, et al. Chlamydia trachomatis
-associated ectopic pregnancy: Serologic and histologic correlates. J Infect Dis 1992; 165:1076–1081.
13. Sherman KJ, Daling JR, Stergachis A, et al. Sexually transmitted diseases and tubal pregnancy. Sex Transm Dis 1990; 17:115–121.
14. Svensson L, Mårdh PA, Ahlgren M, et al. Ectopic pregnancy and antibodies to Chlamydia trachomatis
. Fertil Steril 1985; 44:313–317.
15. Hillis SD, Owens LM, Marchbanks PA, et al. Recurrent chlamydial infections increase the risks of hospitalization for ectopic pregnancy and pelvic inflammatory disease. Am J Obstet Gynecol 1997; 176:103–107.
16. Andersen B, Østergaard L, Puho E, et al. Ectopic pregnancies and reproductive capacity after Chlamydia trachomatis
positive and negative test results: A historical follow-up study. Sex Transm Dis 2005; 32:377–381.
17. Bakken IJ, Nordbø SA, Skjeldestad FE. Chlamydia trachomatis
testing patterns and prevalence of genital chlamydial infection among young men and women in central Norway 1990–2003: A population-based registry study. Sex Transm Dis 2005; 33:26–30.
18. Bakken IJ, Skjeldestad FE. Time trends in ectopic pregnancies in a Norwegian county 1970–2004: A population-based study. Human Reproduction 2006; In press.
20. Chen MY, Fairley CK, Donovan B. Discordance between trends in chlamydia notifications and hospital admission rates for chlamydia related diseases in New South Wales, Australia. Sex Transm Infect 2005; 81:318–322.