Recent efforts to control and prevent the epidemic of sexually transmitted infections (STIs) have focused on universal screening of young women. Rates of STIs are highest in women aged 15–24 years.1 The majority of treatable STIs, including Chlamydia trachomatis, Neisseria gonorrhoeae, and Trichomonas vaginalis, are asymptomatic in women and therefore can remain unrecognized. Sexually transmitted infections that are unrecognized and untreated can lead to long-term sequelae such as pelvic inflammatory disease, ectopic pregnancy, infertility, and enhanced acquisition and transmission of human immunodeficiency virus (HIV).2–4 Consequently, early recognition of STIs in young women by universal STI screening has been a recent focus of prevention efforts. For example, several national organizations recommend routine annual screening for C. trachomatis in all sexually active women 25 years of age and younger.5–8 The Centers for Disease Control and Prevention (CDC) also recommend screening for other common STIs in all sexually active adolescents. Programs for universal STI screening have been implemented for adolescents and young adults in both clinic and nonclinic settings.9–13
However, little is known about patients' behaviors after universal STI screening. Although a few studies have documented high rates of successful treatment of infection found on universal screening,14–17 it remains unclear whether treated patients subsequently engage in behaviors that reduce or increase their risk for repeat exposure and reinfection with STIs. Successful efforts to decrease STI rates will require a better understanding of patients' behaviors after they experience universal screening and treatment. Our study of young women enrolled in recruit training for the U.S. Marine Corps examined a nationally-derived group of healthy young women who presented for job training and not specifically for STI screening or clinical services. Military training also provides an ideal research setting by offering precisely defined windows of opportunity for both risky and protective behaviors.
The aims of this prospective study were to estimate whether women who test STI positive at baseline: 1) subsequently exhibit lower rates of sexual activity during vacation than do women who test STI negative; 2) subsequently exhibit lower rates of risky sexual behaviors (defined as inconsistent condom use, nonmonogamous sexual partnerships, “casual” sexual partnerships, and multiple sexual partnerships) during vacation than do women who test STI negative; and 3) are more likely to test STI positive after vacation.
MATERIALS AND METHODS
All women who enrolled in recruit training for the U.S. Marine Corps between June 1999 and June 2000 were invited to participate in the study. Entry to recruit training was a year-round rolling process for each 13-week training cycle. Women from across the United States and territories traveled to a single training facility for all U.S. Marine Corps female recruits. Female civilian research assistants obtained written informed consent from all study participants. No military personnel were present during study enrollment, and participation was strictly voluntary. The institutional review boards for the University of California, San Francisco, and the Naval Hospital Beaufort, Beaufort, SC approved the study. All study participants were given the Human Subjects Bill of Rights statement in accordance with institutional review board guidelines.
The “T1” data set (time point 1, or baseline) consisted of participants' answers to a self-administered questionnaire and laboratory results from universal STI screening. The questionnaire was completed immediately after study enrollment and included demographic information and sexual history. In addition, within the 10-day period after study enrollment, each participant received a routine reproductive health examination required by the U.S. Marine Corps. Universal STI screening was performed during participants' routine examinations regardless of sexual risk history and consisted of laboratory testing for Chlamydia trachomatis, Neisseria gonorrhoeae, and Trichomonas vaginalis. Each woman who tested positive for any of these three STIs was treated and counseled by trained U.S. Navy preventive medicine staff according to clinical guidelines published by the CDC.18 For example, the routine standard-of-care provided by the U.S. Marines was to administer the medications recommended by the CDC and to provide individualized counseling that included information about STIs, discussion of safer sex practices, and recommendation for partner notification and referral. For research purposes, we confirmed that every infected woman was treated by reviewing the medication logbook routinely maintained by the preventive medicine staff.
Recruit training consisted of 13 weeks of structured training that precluded any social activities or contact with male peers. During this time, study participants also participated in a group-level randomized controlled intervention trial to either prevent STIs and unintended pregnancies or to improve nutrition and reduce physical injury. The larger study is described in detail elsewhere.19 Recruit training was followed by unrestricted vacation leave for a median duration of 10 days (range 0–90 days; 72% of participants received between 7 and 14 days of vacation; and 99% received 30 days or less.) Women then returned for a second phase of military training for another three weeks, again precluding any social activities or contact with male peers.
The “T2” data set (time point 2, or vacation) consisted of participants' answers to a second self-administered questionnaire and laboratory results from a second universal STI screening. This questionnaire was completed at the end of the second phase of military training (3 weeks after return from vacation) and assessed sexual behaviors during vacation. The universal STI screening again consisted of laboratory testing for C. trachomatis, N. gonorrhoeae, and T. vaginalis, and each participant who tested positive for any of these STIs was treated and counseled according to clinical guidelines published by the CDC.
Laboratory procedures were similar at T1 and T2. At T1, the screening for C. trachomatis and N. gonorrhoeae was performed by ligase chain reaction (LCx, Abbott Laboratories, Abbott Park, IL—this product is now discontinued) using samples from three sites: endocervix, first void urine, and vagina.20 Endocervical samples were collected by clinicians during the routine reproductive health examination using swabs from the LCx kits. These samples were transported to the Naval Hospital laboratory for processing within 6 hours of collection. First void urine samples (first 20 mL in a marked urine cup) and vaginal samples (vaginal swab inserted 1 to 2 inches and rotated 3 times) were collected by participants after instruction by trained research assistants. These urine and vaginal samples were frozen at –70°C within 24 hours of collection, and then transported to the university-based research laboratory for processing. All endocervical, urine, and vaginal samples were processed using LCx according to manufacturer's directions. Universal screening for T. vaginalis was performed by culture (InPouch TV, Biomed Diagnostics Laboratories, San Jose, CA) using a second participant-collected vaginal sample. This sample was immediately inoculated into the InPouch TV according to manufacturer's instructions, incubated at 37°C, and read for the presence of T. vaginalis at 2 and 5 days after inoculation. At T2, samples were collected and processed in the same manner, except endocervical samples were not collected.
The independent variable, “STI positive diagnosis” at T1 (yes or no), was defined as a positive diagnosis for one or more of the three STIs screened: C. trachomatis, N. gonorrhoeae, or T. vaginalis. A total of six dependent variables were examined. “Sexual activity” during vacation (yes or no) was defined as vaginal or anal intercourse or both. Among women who were sexually active during vacation, four risky sexual behaviors during vacation were examined: “inconsistent condom use” (less than 100% or 100%); “non-monogamous sexual partnerships,” defined as someone who “had or may have had sex with someone other than yourself” during recruit training or vacation (yes or no); “casual sexual partnerships,” defined as “someone you knew for a short period of time with whom you did not have an ongoing relationship” (at least one or none); and “multiple sexual partnerships” (at least two partners or one). The dichotomous categories for each risky sexual behavior represented higher compared with lower risk. STI positive diagnosis at T2 (yes or no) was another dependent variable examined.
Several sociodemographic and background variables were evaluated as possible moderators of the effect of T1 STI-positive diagnosis. These included duration of vacation (number of weeks), age (17–18 years compared with 19–21 years compared with 22 years and older), highest education level completed (high school or GED compared with any college, vocational, or technical), marital status (not married compared with married), geographic location of residence (metropolitan compared with nonmetropolitan, as defined by the United States Department of Agriculture), and region of residence (Northeast compared with Midwest compared with South compared with West, as defined by the CDC).
Data analyses of the associations between T1 STI-positive diagnosis and each of the six dependent variables were performed using both χ2 statistics (univariable) and logistic regression models (multivariable). A separate logistic regression model was constructed for each dependent variable. T1 STI-positive diagnosis and the sociodemographic variables as described above were entered into each model. Two-way interactions between T1 STI-positive diagnosis and each sociodemographic variable were examined, and significant terms (P<.05) were retained. All models met the assumption of adequate fit to a logit model, which was operationalized as P>.020 on the Hosmer-Lemeshow goodness-of-fit test. Intervention assignment in the larger randomized controlled trail was not entered into the regression models because intervention assignment was neither expected nor found to be associated with the dependent variables of interest. Specifically, univariable analyses (χ2 statistics) detected no significant associations between intervention assignment (STI and unintended pregnancy intervention compared with nutrition and fitness intervention) and each of the six dependent variables.
Two different subsets of study participants were analyzed. To examine the dependent variable “sexual activity” during vacation, the subset of study participants who reported sexual activity before T1 was analyzed. This subset was selected to reflect women who would typically be eligible for nonmilitary universal STI screening programs (nonresearch). To examine all other dependent variables (four risky sexual behaviors at vacation, and T2 STI-positive diagnosis), a second subset of women who reported sexual activity during vacation was analyzed. The statistical software package SPSS 11.5 (SPSS Inc., Chicago, IL) was used for all data analyses.
Of the 2,288 women invited for study participation, 2,157 (94%) voluntarily enrolled, and 1,712 completed the behavioral and biological assessments at both T1 and T2 (Table 1). Of the 445 women who did not participate in T2, the majority (85%) left the study for reasons not related to the study itself: 60% had been discharged from the military service before T2, 14% were not available for T2 data collection, and 11% were Reservists and were therefore not on site for T2 data collection. The remaining women who did not participate in T2 consisted of 8% who declined to continue study participation and 7% who were missing either the questionnaire or laboratory results and therefore excluded from the analysis. In total, the 445 women who did not participate in T2 were older (P<.01), more likely to be white and less likely to be Latina (P<.01), more likely to have college or vocational education beyond high school (P<.01), and more likely to be married (P=.02) compared with women who participated in T2. However, most importantly, participants and nonparticipants did not differ in their T1 STI diagnosis. They also did not differ in their baseline geographic location or region of residence or intervention assignment.
Our final sample consisted of 1,712 women who completed both T1 and T2 assessments. An STI-positive diagnosis at T1 was found in 204 women (12%), which was composed of C. trachomatis in 174 (10%), N. gonorrhoeae in 33 (2%), and T. vaginalis in 24 (1%). A subset of 1,452 women (85% of 1,712) reported sexual activity before T1, and a subset of 1,038 women (61% of 1712) reported sexual activity during vacation. Of these 1,038 women, an STI-positive diagnosis at T2 was found in 65 women (6%), which was composed of C. trachomatis in 51 (5%), N. gonorrhoeae in 9 (1%), and T. vaginalis in 7 (1%).
Among the 1,452 women who reported sexual activity before military service, there was no significant univariable relationship between T1 STI-positive diagnosis and sexual activity during vacation (P=.11). Sexual activity was reported by 75% of women who had tested STI positive at T1 and by 69% of women who had tested STI negative at T1. However, multivariable analyses revealed a significant relationship that was moderated by geographic region of residence (Table 2). Specifically, among the women who resided in the Southern region of the United States before T1, women who tested T1 STI positive were more likely than those who tested T1 STI negative to be sexually active during vacation (adjusted odds ratio [AOR] 2.24, 95% confidence interval [CI] 1.26–3.99). Among the women who resided in the Western region before T1, women who tested T1 STI positive were less likely than those who tested T1 STI negative to be sexually active during vacation (AOR 0.47, CI 0.23–0.97). There were no significant associations between T1 STI-positive diagnosis and sexual activity during vacation detected among women who resided in the Northeastern and Midwestern regions.
Similarly, among the 1,038 women who reported sexual activity during vacation leave, there was no significant univariable relationship between T1 STI-positive diagnosis and casual sexual partnerships during vacation (Table 3), but multivariable analysis revealed a significant relationship that was moderated by age (Table 4). Specifically, among women in the middle age group (19–21 years old), those who tested T1 STI positive were more likely than those who tested T1 STI negative to have casual sexual partners during vacation (AOR 2.48, CI 1.12–5.50). There were no significant associations detected between STI-positive diagnosis and casual sexual partnerships among the younger (17–18 years old) and older (22 years old or older) age groups. In addition, by multivariable analyses, T1 STI-positive diagnosis was not significantly associated with inconsistent condom use (AOR 0.75, CI 0.51–1.11), nonmonogamous partnerships (AOR 1.08, CI 0.74–1.58), or multiple sexual partnerships during vacation (AOR 1.30, CI 0.79–2.13). However, women who tested T1 STI positive were more likely to test T2 STI positive, as indicated by both univariable (Table 3) and multivariable analyses (Table 4).
This study found that healthy young women who enrolled in recruit training for the U.S. Marine Corps and underwent a routine universal STI screen subsequently exhibited high rates of sexual activity and risky sexual behaviors (inconsistent condom use, nonmonogamous partnerships, casual partnerships, and multiple partnerships) during their first brief vacation. These rates were substantial, regardless of STI diagnosis at entry to military training. Furthermore, after adjusting for sociodemographic characteristics, an STI-positive diagnosis had no effect on most of the behavioral outcomes examined. Nevertheless, despite this lack of an effect on behaviors, women who tested STI positive at baseline were much more likely to subsequently test STI positive at the repeat universal screen after vacation.
A few other studies in nonmilitary settings have examined patient behaviors after STI diagnosis.21–23 For example, one study found that those who tested STI positive at baseline demonstrated favorably decreased rates of both sexual activity and inconsistent condom use at 90-day follow-up, but increased rates of having multiple sexual partners.21 Another study of younger adolescents found that those with a past history of STI diagnosis demonstrated higher current rates of several risky sexual behaviors.22 Despite these variations in reported sexual activity and risky sexual behaviors, our study was consistent with these prior studies in observing that an STI-positive diagnosis was associated with a subsequent STI-positive diagnosis. The observation that STI-positive women were at increased risk for repeat STI acquisition regardless of their behaviors lends support to other current research that has highlighted the role of factors beyond the level of the individual patient, such as sexual networks, social networks, and geographic risk spaces.24–28 Although the contribution of network-level factors was beyond the scope of our study, knowledge of our study participants' socioenvironmental contexts may have helped to explain the increased STI acquisition experienced by some women when their behaviors seemed similar to others' behaviors. In addition, individual women may demonstrate variable degrees of biologic susceptibility to STI acquisition. Thus, work toward understanding individual biologic vulnerability may also help explain the increased STI acquisition experienced by some women.
As mentioned previously, the setting of military training provided many strengths for the purposes of research: a nationally-derived sample of healthy young women, routine universal STI screening, universal treatment in a timely fashion for all women who test STI positive, defined windows of opportunity for sexual behaviors, and an ideal time frame to allow repeat STI screening. The main limitation of our study was the lack of access to information regarding sexual partners and vacation details. For example, women who tested STI positive at baseline were routinely counseled to notify their sexual partners and refer them for treatment and evaluation. However, partners were located throughout the U.S., and it was not known whether partner notification was successful or whether partners successfully received treatment and evaluation. It was not known whether women returned to their prior pretraining social environments and their prior sexual partners or traveled to new locations and established new partnerships. We were also unable to assess whether women reported all of their sexual encounters that may have contributed to STI acquisition.
Another possible limitation of our study is the question of generalizability to nonmilitary women. Our study participants self-selected for enlistment in the U.S. Marine Corps. Our findings from focus groups in preparation for this study were consistent with other research that indicates that young women enlist for a wide variety of reasons, including job training and security, financial support for higher education, sense of personal achievement, respect for self and others, and desire to serve others.29,30 These women were working and living in a highly structured program that included a precisely defined block of “free time” at vacation. The setting of military training is unique. However, we postulate that there are likely some similarities to the setting of college that includes vacation and “spring break.” Behaviors during spring break have not been well-studied, but a survey of young adult women conducted by the American Medical Association found that a majority reported that spring break trips are a time of increased sexual activity and “reckless” behavior, during which “being promiscuous is a way to fit in.”31 Epidemiologic data have also suggested the seasonality of sexual activity, unsafe sex, and loss of virginity that increase during summer vacations and holiday times.32,33 As STI screening is encouraged on college campuses,34 sexual behaviors during vacation may have implications for the optimal timing of screening. The settings of military training and colleges may therefore demonstrate some similarities.
Overall, our findings suggest that the standard-of-care approach of universal STI screening followed by one brief treatment and counseling session of STI positive women does not have a sufficient desired dampening effect on subsequent risky sexual behaviors. As universal screening programs are increasingly implemented to identify and treat STIs, a next step toward control of STIs may be the development of services to specifically decrease risk behaviors after successful screening and treatment. Beyond the typical model of universal screening, treatment, and counseling by clinicians may be a system that offers more detailed risk-reduction counseling tailored to the individual. This counseling regarding behaviors on the individual level can also address factors on the larger network level by including a discussion of the woman's social network and choice in sexual partners. For example, in our military setting, counseling could include not only discussion of a woman's specific risky behaviors, but also discussion of the woman's travel plans and potential reexposure to a network in which she had already acquired an STI. In addition, universal STI screening programs will likely need to be more specific to the population setting. For example, repeat universal STI screening in military settings may need to be offered specifically after periods of vacation and during later phases of service training when there are more frequent opportunities for sexual activity and risky behaviors. Finally, our findings suggest that STI acquisition cannot be reliably predicted by a woman's history of risky sexual behaviors. Universal STI screening must continue to be made increasingly accessible to healthy young women.
1. Centers for Disease Control and Prevention (CDC). Sexually Transmitted Disease Surveillance, 2004. Atlanta (GA): U.S. Department of Health and Human Services; 2005.
2. Cates W Jr, Wasserheit JN, Marchbanks PA. Pelvic inflammatory disease and tubal infertility: the preventable conditions. Ann N Y Acad Sci 1994;709:179–95.
3. Westrom L, Joesoef R, Reynolds G, Hagdu A, Thompson SE. Pelvic inflammatory disease and fertility. A cohort study of 1,844 women with laparoscopically verified disease and 657 control women with normal laparoscopic results. Sex Transm Dis 1992;19:185–92.
4. Rottingen JA, Cameron DW, Garnett GP. A systematic review of the epidemiologic interactions between classic sexually transmitted diseases and HIV: how much really is known? Sex Transm Dis 2001;28:579–97.
5. Sexually transmitted diseases treatment guidelines 2002. Centers for Disease Control and Prevention. MMWR Recomm Rep 2002;51:1–78.
6. U.S. Preventive Services Task Force. Chlamydial infection: screening 2001. Guide to clinical preventive services. Alexandria (VA): International Medical Publishing Company; 2001. p. 325–32.
7. Guidelines for Women's Health Care. 2nd ed. Washington (DC):American College of Obstetricians and Gynecologists; 2002.
8. Institute of Medicine Committee on Leading Health Indicators. Healthy People 2010. Washington (DC): National Academy Press; 1999.
9. Mehta SD, Bishai D, Howell MR, Rothman RE, Quinn TC, Zenilman JM. Cost-effectiveness of five strategies for gonorrhea and chlamydia control among female and male emergency department patients. Sex Transm Dis 2002;29:83–91.
10. Shafer MA, Tebb KP, Pantell RH, Wibbelsman CJ, Neuhaus JM, Tipton AC, et al. Effect of a clinical practice improvement intervention on Chlamydial screening among adolescent girls. JAMA 2002;288:2846–52.
11. Kent CK, Branzuela A, Fischer L, Bascom T, Klausner JD. Chlamydia and gonorrhea screening in San Francisco high schools. Sex Transm Dis 2002;29:373–5.
12. Lifson AR, Halcon LL, Hannan P, St Louis ME, Hayman CR. Screening for sexually transmitted infections among economically disadvantaged youth in a national job training program. J Adolesc Health 2001;28:190–6.
13. Cohen DA, Nsuami M, Martin DH, Farley TA. Repeated school-based screening for sexually transmitted diseases: a feasible strategy for reaching adolescents. Pediatrics 1999;104:1281–5.
14. Hwang LY, Tebb KP, Shafer MA, Pantell RH. Examination of the treatment and follow-up care for adolescents who test positive for Chlamydia trachomatis infection. Arch Pediatr Adolesc Med 2005;159:1162–6.
15. Magid DJ, Stiffman M, Anderson LA, Irwin K, Lyons EE. Adherence to CDC STD guideline recommendations for the treatment of Chlamydia trachomatis infection in two managed care organizations. Sex Transm Dis 2003;30:30–2.
16. Levitt MA, Johnson S, Engelstad L, Montana R, Stewart S. Clinical management of chlamydia and gonorrhea infection in a county teaching emergency department—concerns in overtreatment, undertreatment, and follow-up treatment success. J Emerg Med 2003;25:7–11.
17. Bachmann LH, Richey CM, Waites K, Schwebke JR, Hook EW 3rd. Patterns of Chlamydia trachomatis testing and follow-up at a University Hospital Medical Center. Sex Transm Dis 1999;26:496–9.
18. 1998 guidelines for treatment of sexually transmitted diseases. Centers for Disease Control and Prevention. MMWR Recomm Rep 1998;47:1–111.
19. Boyer CB, Shafer MA, Shaffer RA, Brodine SK, Pollack LM, Betsinger K, et al. Evaluation of a cognitive-behavioral, group, randomized controlled intervention trial to prevent sexually transmitted infections and unintended pregnancies in young women. Prev Med 2005;40:420–31.
20. Carroll KC, Aldeen WE, Morrison M, Anderson R, Lee D, Mottice S. Evaluation of the Abbott LCx ligase chain reaction assay for detection of Chlamydia trachomatis and Neisseria gonorrhoeae in urine and genital swab specimens from a sexually transmitted disease clinic population. J Clin Microbiol 1998;36:1630–3.
21. Crosby RA, DiClemente RJ, Wingood GM, Salazar LF, Rose E, Levine D, et al. Associations between sexually transmitted disease diagnosis and subsequent sexual risk and sexually transmitted disease incidence among adolescents. Sex Transm Dis 2004;31:205–8.
22. Diclemente RJ, Wingood GM, Sionean C, Crosby R, Harrington K, Davies S, et al. Association of adolescents' history of sexually transmitted disease (STD) and their current high-risk behavior and STD status: a case for intensifying clinic-based prevention efforts. Sex Transm Dis 2002;29:503–9.
23. Fortenberry JD, Brizendine EJ, Katz BP, Orr DP. Post-treatment sexual and prevention behaviours of adolescents with sexually transmitted infections. Sex Transm Infect 2002;78:365–8.
24. Ellen JM, Gaydos C, Chung SE, Willard N, Lloyd LV, Rietmeijer CA. Sex partner selection, social networks, and repeat sexually transmitted infections in young men: a preliminary report. Sex Transm Dis 2006;33:18–21.
25. Ellen JM, Brown BA, Chung SE, Potterat JJ, Muth SQ, Valente TW, et al. Impact of sexual networks on risk for gonorrhea and chlamydia among low-income urban African American adolescents. J Pediatr 2005;146:518–22.
26. Jennings J, Glass B, Parham P, Adler N, Ellen JM. Sex partner concurrency, geographic context, and adolescent sexually transmitted infections. Sex Transm Dis 2004;31:734–9.
27. Bettinger JA, Adler NE, Curriero FC, Ellen JM. Risk perceptions, condom use, and sexually transmitted diseases among adolescent females according to social network position. Sex Transm Dis 2004;31:575–9.
28. Fichtenberg CM, Ellen JM. Moving from core groups to risk spaces [editorial response]. Sex Transm Dis 2003;30:825–6.
29. Eighmey J. Why do youth enlist? Identification of underlying themes. Armed Forces Soc 2006;32:307–28.
30. Bachman JG, Freedman-Doan P, O'Malley PM. Should U.S. military recruiters write off the college-bound? Armed Forces Soc 2001;27:461–76.
32. Wellings K, Macdowall W, Catchpole M, Goodrich J. Seasonal variations in sexual activity and their implications for sexual health promotion. J R Soc Med 1999;92:60–4.
33. Rodgers JL, Harris DF, Vickers KB. Seasonality of first coitus in the United States. Soc Biol 1992;39:1–14.
34. Sipkin DL, Gillam A, Grady LB. Risk factors for Chlamydia trachomatis infection in a California collegiate population [published erratum appears in J Am Coll Health 2004;52:236]. J Am Coll Health 2003;52:65–71.
© 2007 The American College of Obstetricians and Gynecologists
Figure. No Caption Available.