CHLAMYDIA TRACHOMATIS (CT) IS THE MOST common bacterial sexually transmitted infection in the United States. Approximately, 70% to 90% of infections in women and a large percentage of male infections are asymptomatic.1–3 Acute epididymitis/orchitis is the most common CT-related health consequence in men.4–6 Other complications have been linked to CT as well, including chronic prostatitis, urethral stricture,7 and infertility. However, the role that CT infection plays in the development of prostatitis and infertility is still debatable. Though prostatitis has long been linked to CT infection,4,8 quality issues regarding the specimen collection methods used by most studies examining this association brings the validity of their findings into question.9 The link between CT and infertility also remains controversial, as several studies have found an association,10,11 whereas others have not.12–14 To our knowledge, there have not been any large, population-based epidemiologic studies published that assess either the correlation between CT infection and health outcomes in males or the overall burden of these outcomes.
The US Air Force (USAF) population contains a large number of young, sexually active males, making it suitable for conducting a large-scale study of CT infection and related outcomes in men. Among asymptomatic male military populations, previous studies have demonstrated CT prevalence ranging from 4.1% to 5.3%,15–17 which is on par with or even slightly lower than the prevalence observed in their civilian counterparts.18,19 The objectives of this study were to assess the overall burden of reproductive tract health outcomes among USAF males and to compare these events in men with and without recent CT infections.
Materials and Methods
Study Population and Exposure Status Determination
We conducted a record-based retrospective cohort study to evaluate the occurrence of reproductive tract complications among active duty Air Force (ADAF) men with and without prior CT infections. In the USAF, men are generally tested for CT only when presenting with symptoms or high-risk behavior. The Air Force Institute for Operational Health (AFIOH) in San Antonio, TX, houses a reference laboratory that performs CT testing using nucleic acid amplification tests for many USAF clinics (52 of 77, 67.5%). By protocol, all CT specimens are also tested simultaneously for Neisseria gonorrhoeae (GC).
Using existing AFIOH laboratory records we generated the study cohort, which consisted of all ADAF males receiving at least one CT test between January 1, 2001, and December 31, 2002; we considered this timeframe to be the enrollment period. Subsequent laboratory tests occurring through December 31, 2005 were also linked to monitor the occurrence of later CT infections. Within the cohort, two main exposure groups were created: 1) “CT Positive”: tested positive during the enrollment period and possibly again during follow-up; 2) “CT Negative”: never tested positive, either at enrollment or thereafter. For CT positive men, the date of each man’s first positive test was his enrollment date. For CT negative men, enrollment began on the date of the first recorded CT test. Using electronic medical data, we then tracked cohort members through December 31, 2005, to monitor for health outcomes. Institutional Review Board approval was obtained before initiation of this study.
Evaluation of Outcomes
To monitor for health outcomes, we linked laboratory records to centralized outpatient and inpatient electronic medical databases containing information from all USAF clinics. Medical visits with International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM)20 codes indicating a reproductive tract complication were retained for analysis. Outcomes included orchitis/epididymitis (ICD-9-CM code 604), male infertility including azoospermia, oligospermia, infertility due to extratesticular causes, and unspecified male infertility (ICD-9-CM codes 606.0-1 and 606.8-9), urethral stricture (ICD-9-CM code 598), and prostatitis (ICD-9-CM codes 601.1 and 601.8-9). Only the first and second diagnostic codes were considered when determining the occurrence of an outcome to ensure that it was a key reason for the visit. Additionally, diagnostic fields beyond the first and second codes are populated approximately 10% or less of the time, so including this information would not likely impact study outcomes.
Person-time was determined by calculating the time from study enrollment to health outcome, loss to follow-up, or study completion on December 31, 2005. For men who developed a given complication, the first date the event was noted in medical records was the last date of follow-up; subsequent laboratory results were not considered when determining CT exposure status. CT exposure status and person-time were calculated separately for each health outcome to allow for differences in censoring and to ensure that CT test results occurred before each health outcome diagnosis. For those men who did not develop a given complication, either the USAF separation date obtained from military personnel data or the study completion date was used to determine person-time, whichever came first. The last date of any outpatient medical visit was used to calculate follow-up time when separation date information was unavailable. Men without a known length of follow-up were excluded from analysis altogether.
CT infection status was compared over demographic and risk variables including age at enrollment, marital status, USAF rank, race/ethnicity, and GC infection history, using unpaired Student t test for continuous variables and Pearson χ2 tests for categorical variables. Demographic information was obtained from military personnel data, whereas GC test results were obtained from AFIOH laboratory records. The proportion of persons developing each complication was compared among exposure groups using the binomial proportions test; mean ages among those developing each complication and time to diagnosis were compared using Student t test. In addition, life-table estimates of cumulative incidence were calculated and plotted for each health outcome within the CT positive and negative exposure groups; log-rank tests were used to compare incidence curves. Finally, Cox proportional hazard regression models, with 95% confidence intervals and P values, were generated to estimate the association between CT infection and each reproductive tract health outcome. To adjust for potential confounding, all demographic and risk variables listed above were included in each model, except for enrollment age, which was included as a time-varying covariate. Analysis was restricted to men with complete demographic information. Possible variable interactions were assessed using likelihood ratio tests. Proportionality was assessed among individual covariates using the log-rank test for equality and overall using Schoenfeld residuals. Statistical analyses were conducted using Stata software, version 9.2 (StataCorp, College Station, TX).
The AFIOH laboratory tested CT specimens from 30,041 ADAF males between January 1, 2001, and December 31, 2002; of these, 8850 men were excluded because of unknown length of follow-up time and 3427 were excluded because of incomplete demographic data. The final cohort consisted of 17,764 ADAF males and approximately 58,945 person-years. The mean age of the cohort was 23.5 (±5.0) years; ages ranged from 17 to 56 years, and CT-positive men were significantly younger than those never testing positive (Table 1). The mean follow-up time was 3.4 (±1.0) years. Overall, 18.2% of the cohort tested positive for CT at least once during the study timeframe. Among demographic and risk groups, a significantly higher proportion of CT-positive men were unmarried, Hispanic, enlisted, and GC positive, when compared with those never testing positive for CT.
During the study period, 913 (5.14%) men experienced a reproductive tract complication. Orchitis/epididymitis was the most common health outcome among both exposure groups (Table 2), followed by infertility in CT-positive men and prostatitis in CT-negative men. Any outcome and orchitis/epididymitis occurred more frequently in the CT positive exposure group. CT positive men developing any outcome, orchitis/epididymitis, and prostatitis were younger than CT negative men developing these same outcomes (Table 2). For CT positive men the median time between first CT positive test and an adverse health outcome ranged from 3.7 to 4.0 years (Table 2); in comparison, the mean follow-up time for CT negative men was statistically similar. Among ADAF men who tested positive for CT, the cumulative incidence of orchitis/epididymitis, prostatitis, infertility, and urethral stricture was 4.28% (3.61–5.09), 1.41% (1.05–1.90), 1.27% (0.91–1.78), and 0.13% (0.05–0.34), respectively. Among men never testing positive, the cumulative incidence of orchitis/epididymitis, prostatitis, infertility, and urethral stricture was 3.30% (3.01–3.63), 1.45% (1.26–1.66), 1.00 (0.84–1.19), and 0.08 (0.05–0.15), respectively.
Adjusted Hazard Ratios (AHR) indicate that ADAF men who tested positive for CT were more likely to develop orchitis/epididymitis [AHR 1.38 (1.13–1.70)] and any outcome [AHR 1.37 (1.16–1.61)] than those who did not (Fig. 1, Table 3). A positive, but marginal association was observed between CT and infertility [AHR 1.36 (0.93–2.00)]. Men who were married during part or all of the follow-up time were more likely to develop orchitis/epididymitis, infertility, and any complication, compared with those who were unmarried (Table 3). Compared with other racial/ethnic groups, other non-Hispanic men were less likely to develop orchitis/epididymitis [AHR 0.24 (0.10–0.59)] and any outcome [AHR 0.54 (0.34–0.87)] than other race/ethnic groups. GC infection was not associated with an elevated risk of any of the health outcomes; removing the GC variable from the hazard models did not significantly impact AHRs (data not shown).
Overall, the cumulative incidence of reproductive tract complications was low among ADAF males, ranging from 0.1% for urethral stricture to 3.6% for orchitis/epididymitis. However, in a population of this size, even small percentages can translate to large numbers. Among men with a history of CT infection, we found increased incidences of orchitis/epididymitis and any complication compared with those who did not have a CT positive test. Similarly, previous clinic-based studies have observed a positive association between CT and orchitis/epididymitis and have estimated that between 45% and 85% of men with orchitis/epididymitis had previous infections with CT and/or GC.5,21–23 Additionally, men who were married during part or all of the study time were more likely to develop orchitis/epididymitis, infertility, and any outcome than were those who were unmarried. It is unclear whether married men are actually more likely to develop these conditions or to simply realize that they have occurred. For example, with infertility, we anticipate that married men were more likely to start a family and recognize their infertility. Finally, the lower rates of orchitis/epididymitis and any outcome observed among other, non-Hispanic men were anticipated; most of these men were of Asian or Native American descent, groups previously shown to have lower CT rates than other racial/ethnic groups.18,19
We observed a positive, although not statistically significant, association between CT infection and infertility. Perhaps significance may have been achieved if the study size had been larger or if cohort members were followed for a longer period. Many of the young men enrolled in the cohort may not yet have been planning families, a time when fertility problems would most likely be realized. A longer follow-up time may have revealed additional cases among currently unmarried cohort members as more of them eventually married and tried to have children. In addition, only men seeking medical care would have been diagnosed with infertility. Thus, it is possible that the infertility health outcome group in the current study consisted mainly of health-seeking, married men who were trying to conceive. Infertile men who were unmarried and not actively trying to conceive may have been misclassified as noncases, resulting in a biased measurement of the association between CT status and infertility. The direction of the bias is difficult to predict, as the misclassification could have been conceivably differential if men with past Chlamydial infections were more likely to seek health care for infertility. Nonetheless, similar to our findings, previous studies have also observed an epidemiologic link between male infertility and prior CT infection.24–26 A nested case-control study conducted within the Northern Finland Birth Cohort found that, compared with gender-matched controls, the presence of CT antibodies was proportionally higher in men than in women among subfertile couples.25 Idahl et al.26 also found that the presence of CT IgG in male partners significantly predicted subfertility in couples. Additionally, the ability for CT to cause infertility is biologically plausible. In a systematic literature review to epidemiologically examine the association between sexually transmitted urethritis and infertility, Ness et al.6 found evidence suggesting that, just as pelvic inflammatory disease may lead to infertility in women, chronic orchitis/epididymitis caused by chlamydial urethritis in men may result in tubular damage or epididymal occlusion and, eventually, infertility. However, as other studies have found either no association or that CT infection is likely not very important to male infertility,12–14 whether a link exists is still unclear.
In the current study, we found no association between CT infection and prostatitis. As was discussed for infertility, it is possible that a longer follow-up time may be needed to observe the development of this condition as well. To our knowledge, the average time from CT infection to prostatitis manifestation, and thus the proper follow-up time, is unknown. Therefore, it is possible that a true positive association was missed by our study. Whether any association between CT and prostatitis exists at all is still uncertain. Though CT and prostatitis have long been linked,4,8 the quality of CT specimens collected by most studies limits the ability to establish cause and effect.9 Most prostate specimens are collected via expressed prostatic secretion that may become contaminated with urethral microorganisms, where CT infections are initiated, during the collection process. Using perineal biopsy samples from the peripheral prostatic lobes is the only way to prevent urethral contamination.9,27 Two studies28,29 that used perineal biopsy samples to investigate the connection between CT and prostatitis did not isolate CT from any of the total of 72 protatitis cases they examined, though several of the same patients tested positive when using prostatic urethra specimens. This suggests that CT infection does not ascend to the prostate and that specimens collected from previous studies may have been contaminated. Thus, the conclusion that CT causes prostatitis reached by those studies may be inaccurate, and the null association we observed may be true.
The fact that 18.2% of the cohort tested positive for CT during follow-up, though most men in this group were likely symptomatic, lends support to the idea of routinely screening all sexually-active young men. As CT is asymptomatic in most cases,1–3 population-based screening would catch many more positive men and help prevent transmission to women. Among 1192 sexually-active women observed during their first year of military service, 19.9% acquired a CT infection.30 Screening their male counterparts could have prevented many of these infections and also future adverse health outcomes. In a recent study, universal urine nucleic acid amplification tests screening of 450 males, ages 16 to 24 years, entering a National Job Training Program was shown to be cost-effective, and was estimated to have prevented 21 cases of pelvic inflammatory disease in their female partners.31 A study using stochastic simulation models to predict the success of various screening scenarios estimated a reduction in CT prevalence from 4.2% to 1.4% in 10 years if both men and women were screened and if partner referral was routinely used.32 In our study, had screening data been routinely available, our cohort size would have been much larger, potentially allowing stronger associations between CT and subsequent health outcomes to be observed.
The use of a records-based, retrospective methodology, rather than a prospective one, comes with several limitations. First, CT exposure status may have been misclassified in some cohort members. Prestudy CT status was unknown, as we did not include CT tests occurring before January 2001. Additionally, because of the asymptomatic nature of CT, some men with subclinical infections may not have sought medical care and received CT testing. Thus, men with unrecognized CT infections may have erroneously been placed into the “CT negative” exposure group, resulting in an underestimation of health outcome rates in the “CT positive” group.
Second, ICD-9-CM codes retrieved from the electronic medical databases were not verified in individual medical records because of the logistical difficulties of doing so for such a large cohort. Previous reports have shown that using ICD-9-CM codes alone to diagnose patients may be problematic because such data include some level of miscoding and thus may not accurately predict health conditions.33,34 However, one study35 found that military inpatient data are coded with a high degree of specificity.
Third, information on sexual activity (i.e., frequency, contraceptive use) and family planning intentions was not available from study participants. This information may have allowed the creation of a subpopulation of men trying to conceive in which to better examine the infertility health outcome. As stated previously, infertile men who are not planning a family may be less likely to seek fertility-related medical care, and excluding this group of men from the current analysis may have eliminated some misclassification when determining infertility status.
Finally, the medical databases we used only contain information from visits conducted at USAF clinics. Under the Military Health System, ADAF may also use non-USAF health facilities; these visits were not captured in our study. Use of non-USAF facilities differs by base, as the medical capabilities of base clinics and/or hospitals vary widely. Simply because of military healthcare availability, or lack thereof, we expect that some visits involving CT-related health outcomes were missed. However, the resulting underestimation likely occurred in a nondifferential manner regarding exposure and outcome status. It is also possible, because of associated stigmas, that some ADAF members purposely sought outside care for CT-related health issues to maintain privacy. Consequently, some men placed into the “CT negative” group may have tested positive at nonmilitary facilities, and their results were not captured. Such healthcare-seeking behavior likely biased the association between CT and health outcomes toward the null, meaning that actual hazard ratios may be higher than we observed.
The main strengths of our study include the large population size and accessibility to a vast amount of information on cohort members. Use of a military population, whose members are systematically tracked from entrance to separation, may allow for more complete data collection than would otherwise be possible. Because medical records and enrollment data are readily available for nearly all ADAF members, monitoring for health outcomes and calculating person-time are relatively simple tasks. Thus, although conducting a retrospective cohort study can be challenging in general, the availability of standardized data within the military allows such a study to be more easily and accurately done than may be possible in civilian settings. In addition, should we decide to revisit this study in the future to examine longer-term relationships between CT and health outcomes in the same cohort, we will be able to continue the follow-up because of the continuous presence of complete data for ADAF members.
In conclusion, our findings indicate that the burden of the health outcomes we examined is small, although not inconsequential, in the male ADAF population. CT-infected men are an important source of infection in women, for whom the relationships between CT and adverse outcomes are much more established. In addition, we observed a statistically significant association with CT and both orchitis/epididymitis and any outcome. We also found an association between CT and infertility that bordered on significance; with a larger cohort or greater follow-up time we may have been able to attain significance. Longer-term, prospective studies are needed to further explore these associations.
1. Phillips RS, Hanff PA, Holmes MD, et al. Chlamydia trachomatis
cervical infection in women seeking routine gynecologic care: Criteria for selective testing. Am J Med 1989; 86:515–520.
2. Schachter J, Stoner E, Moncada J. Screening for chlamydial infections in women attending family planning clinics. West J Med 1983; 138:375–379.
3. Stamm WE, Holmes KK. Chlamydia trachomatis
infections of the adult. In: Holmes KK, Mardh PA, Sparling PF, et al., eds. Sexually Transmitted Disease. New York, NY: McGraw-Hill, 1990:181–193.
4. Nelson HD, Helfand M. Screening for chlamydial infection: A summary of the evidence. Am J Prev Med 2001; 20(Suppl 3):95–107.
5. Dale AW, Wilson JD, Forster GE, et al. Management of epididymo-orchitis in genitourinary medicine clinics in the United Kindom’s North Thames region 2000. Int J STD AIDS 2001; 12:342–345.
6. Ness RB, Markovic N, Carlson CL. Do men become infertile after having sexually transmitted urethritis? An epidemiologic examination. Fertil Steril 1997; 68:205–213.
7. McMillan A, Pakianathan M, Mao JH, et al. Urethral stricture and urethritis in men in Scotland. Genitourin Med 1994; 70:403–405.
8. Weidner W, Schiefer HG, Krauss H. Role of Chlamydia trachomatis
and mycoplasmas in chronic prostatitis. A review. Urol Int 1988; 43:167–173.
9. Wagenlehner FME, Weidner W, Naber KG. Chlamydial infections in urology. World J Urol 2006; 24:4–12.
10. Bollman R, Engel S, Sagert D, et al. Investigations on the detection of Chlamydia trachomatis
infections in infertile male outpatients. Andrologia 1998; 30:23–27.
11. Samra Z, Soffer Y, Pansky M. Prevalence of genital chlamydia and mycoplasma infection in couples attending a male infertility clinic. Eur J Epidemiol 1994; 10:69–73.
12. Eggert-Kruse W, Rohr G, Demirakca T, et al. Chlamydial serology in 1303 asymptomatic subfertile couples. Hum Reprod 1997; 12:1464–1475.
13. Habermann B, Krause W. Altered sperm function or sperm antibodies are not associated with chlamydial antibodies in infertile men with leucocytospermia. J Eur Acad Dermatol Venereol 1999; 12:25–29.
14. Eggert-Kruse W, Buhlinger-Göpfarth N, Rohr G, et al. Antibodies to Chlamydia trachomatis
in semen and relationship with parameters of male fertility. Hum Reprod 1996; 11:1408–1417.
15. Arcari CM, Gaydos JC, Howell MR, et al. Feasibility and short-term impact of linked education and urine screening interventions for Chlamydia and gonorrhea in male army recruits. Sex Transm Dis 2004; 31:433–437.
16. Shafer MA, Boyer CB, Shaffer RA, et al. Correlates of sexually transmitted diseases in a young male deployed military population. Mil Med 2002; 167:496–500.
17. Cecil JA, Howell MR, Tawes JJ, et al. Features of Chlamydia trachomatis
and Neisseria gonorrhoeae
infection in male Army recruits. J Infect Dis 2001; 184:1216–1219.
18. Schillinger JA, Dunne EF, Chapin JB, et al. Prevalence of Chlamydia trachomatis
infection among men screened in 4 U.S. cities. Sex Transm Dis 2005; 32:74–7.
19. LaMontagne DS, Fine DN, Marrazzo JM. Chlamydia trachomatis infection in asymptomatic men. Am J Prev Med 2003; 24:36–42.
21. Melekos MD, Asbach HW. The role of chlamydiae in epididymitis. Int Urol Nephrol 1988; 20:293–297.
22. Mulcahy FM, Bignell CJ, Rajakumar R, et al. Prevalence of chlamydial infection in acute epididymo-orchitis. Genitourin Med 1987; 63:16–18.
23. Berger RE, Alexander ER, Monda GD, et al. Chlamydia trachomatis as a cause of acute “idiopathic” epididymitis. N Engl J Med 1978; 298:301–304.
24. Penna Videau S, Cermeno Vivas J, Salazar N. IgA antibodies to Chlamydia trachomatis
and seminal parameters in asymptomatic infertile males. Arch Androl 2001; 46:189–195.
25. Karinen L, Pouta A, Hartikainen AL, et al. Association between Chlamydia trachomatis
antibodies and subfertility in the Northern Finland Birth Cohort 1966 (NFBC 1966), at the age of 31 years. Epidemiol Infect 2004; 132:977–984.
26. Idahl A, Boman J, Kumlin U, et al. Demonstration of Chlamydia trachomatis
IgG antibodies in the male partner of the infertile couple is correlated with a reduced likelihood of achieving pregnancy. Hum Reprod 2004; 19:1121–6.
27. Weidner W, Diemer T, Huwe P, et al. The role of Chlamydia trachomatis
in prostatitis. Int J Antimicrob Agents 2002; 19:466–70.
28. Doble A, Thomas BJ, Walker MM, et al. The role of Chlamydia trachomatis
in chronic abacterial prostatitis: A study using ultrasound guided biopsy. J Urol 1989; 141:332–333.
29. Weidner W, Schiefer H-G, Krauss H, et al. Chronic prostatitis a thorough search for etiologically involved microorganisms in 1461 patients. Infection 1991; 19(Suppl 3):119–125.
30. Shafer MA, Boyer CB, Pollack LM, et al. Acquisition of Chlamydia trachomatis
by young women during their first year of military service. Sex Transm Dis 2008; 35:255–259.
31. Blake DR, Quinn TC, Gaydos CA. Should asymptomatic men be included in chlamydia screening programs? Cost-effectiveness of Chlamydia screening among male and female entrants to a National Job Training Program. Sex Transm Dis 2008; 35:91–101.
32. Kretzschmar M, Welte R, van den Hoek A, et al. Comparative model-based analysis of screening programs for Chlamydia trachomatis
infections. Am J Epidemiol 2001; 153:90–101.
33. Ratelle S, Yokoe D, Blejan C, et al. Predictive value of clinical diagnostic codes for the CDC case definition of pelvic inflammatory disease (PID): Implications for surveillance. Sex Transm Dis 2003; 30:866–870.
34. Geller SE, Ahmed S, Brown ML, et al. International Classification of Diseases—9th
revision coding for preeclampsia: How accurate is it? Am J Obstet Gynecol 2004; 190:1629–1633.
© Copyright 2008 American Sexually Transmitted Diseases Association
35. Meyer GS, Krakauer H. Validity of the Department of Defense Standard Inpatient Data Record for quality management and health services research. Mil Med 1998; 163:461–465.