Although access to hormonal contraception (HC) reduces unwanted pregnancy and maternal morbidity and mortality, a body of evidence from recent systematic reviews, meta-analyses and in vivo and in vitro studies suggest that the progestin injectable depo-medroxyprogesterone acetate (DMPA) increases risk of human immunodeficiency virus (HIV) acquisition.1–5 Comparatively less emphasis, however, has focused on the potential association of DMPA and other HC and other sexually transmitted infections (STIs).
Several biological mechanisms by which HC use may facilitate STI acquisition have been proposed including through changes in the protective cervicovaginal epithelial barrier from hypoestrogenism induced by progestin-only methods.6,7 A second mechanism is through weakening of immune defense.8 For example, DMPA is known to bind to glucocorticoid receptors, which generally results in immune modulation.5,9 Third, hypoestrogenism induced by progestin-only methods could lead to changes in the vaginal microbiota composition, leading to vaginal dysbiosis and inflammation,10 which in turn could lead to epithelial breaches and mucus degradation.11,12 At a behavioral level, HC use may result in decreased condom use, thereby increasing risk of STI exposure.13,14
Two prior systematic reviews have examined the association between HCs and STI acquisition; evidence has not been synthesized since 2008.15,16 Both reviews found that OCP and DMPA users had a possible increased risk of chlamydia but concluded there was inconclusive evidence for gonorrhea, herpes simplex virus type 2 (HSV-2), trichomoniasis, syphilis and human papillomavirus (HPV). Given the magnitude of women using HC globally and the negative health repercussions of many STIs, we conducted an updated systematic review to incorporate literature from longitudinal studies published between 2009 and 2017 on the association between HC use and non-HIV STI acquisition; systematic reviews on HIV acquisition have been updated regularly.1–4
MATERIALS AND METHODS
The protocol was registered a priori with PROSPERO [record 42017069357] and follows PRISMA guidelines (Supplemental Table 1 https://links.lww.com/OLQ/A339). Articles were identified using key term searches of 2 electronic databases: PubMed and EMBASE (Supplemental Fig. 1 https://links.lww.com/OLQ/A340).
Included articles were peer reviewed, published in English, Spanish or French between January 1, 2009, and June 30, 2017, and measured incident/recurrent cases of cervicovaginal HPV, HSV-2, chlamydia, gonorrhea, syphilis, and trichomoniasis, with laboratory diagnostic tests, among HC users compared with nonusers or users of nonhormonal methods. All HC methods were included except for emergency contraception, since it is typically used in combination with other contraceptive methods.17 We also reviewed articles identified from 2 earlier systematic reviews15,16; articles from these reviews which met our criteria are also included.
We excluded cross-sectional studies, review articles, studies which relied on clinical examination or self-reported STIs, and studies which did not control for potential confounding variables. We also excluded studies of HCs and HIV and bacterial vaginosis (BV), as both have been recently reviewed.18,19 Two independent reviewers (K.J.M. and H.E.J. or E.L.G.) screened each abstract or article using Covidence software; a third reviewer who had not previously reviewed the study (H.E.J. or E.L.G.) resolved the discrepancies.
One reviewer (K.J.M.) extracted data, with independent review for accuracy (H.E.J. or E.L.G.). Extracted information included: participant characteristics, geographic location, sample size, sampling method, contraceptive method, duration of use, comparison group, STI, whether infection was incident or recurrent, STI diagnostic test, confounders in adjusted estimates, type of statistical analysis, treatment of missing data, length of time between exposure and outcome assessment, and the effect estimate, variance and significance level.
Risk of bias was assessed using the Newcastle-Ottawa Quality Assessment Scale,20 adapted to reflect challenges identified previously for assessment of the relationship between HC use and STIs/HIV3,15 (Supplemental Tables 2 and 3 https://links.lww.com/OLQ/A341 and https://links.lww.com/OLQ/A342). Two reviewers (K.J.M. and H.E.J. or E.L.G.) independently rated study quality; discrepancies were resolved by discussion among all 3 reviewers.
Our primary outcome is incident STI. We examined findings by HC method used (e.g., OCP, DMPA, levonorgestrel IUD, norethisterone enanthate [Net-En], Norplant) and type of STI. Forest plots were constructed using the forestplot package in R Studio (Version 1.1.383, Vienna, Austria).
Our key term search resulted in 1477 unique articles, 1284 articles were excluded during abstract screening; 24 required full-text review of which 2 were excluded (Fig. 1). An additional 13 prospective studies identified in the previous 2 systematic reviews were considered for inclusion. Of these 8 met our study inclusion criteria and are included,21–28 5 did not meet our criteria.29,30,31s–33s The 30 reviewed studies were all prospective and observational in design, most were longitudinal cohort studies (N = 25), 4 were secondary analysis of a randomized control trial23,24,34s,35s and 1 used a nested case-control design.36s
The majority of articles assessed the incidence or recurrence of HPV (n = 13),26–28,36s–45s followed by trichomoniasis (n = 9),21,24,34s,37s,46s–50s chlamydia (n = 9),21–23,37s,47s,48s,50s–52s gonorrhea (n = 7),21–23,37s,47s,48s,50s HSV-2 (n = 4),35s,37s,53s,54s and syphilis (n = 3)24,37s,47s (not mutually exclusive). Two studies combined incident chlamydia and gonorrhea.25,55s Ten studies included women ages 18–50 years,23,27,36,37s,41s,42s,48s,51s,53s,55s 10 studies included adolescents (<age 18 years),21,25,26,28,35s,44s,46s,47s,50s,52s 4 included women older than age 49 years31s34s,45s,47s and 7 did not report age range, but the majority of participants were of reproductive age.22,38s–40s,43s,49s,54s One third of the studies enrolled populations considered at increased STI risk: women reporting transactional sex (n = 6),21,22,37s,51s,53s,54s injection drug use (n = 1), 43s lower genital tract infection/partner with diagnosed STI (n = 1),52s or living with HIV (n = 2).22,55s Three additional studies included women living with and without HIV.43s,46s,54s
Half of the studies (15 of 30) compared HC users to nonhormonal method users (e.g., condom), twelve studies compared 2 or more types of HCs23,24,26–28,36s,39s,40s,43s,44s,52s,54s 3 compared HC use to women not using any method and women who were sterilized.21,22,46s
Study Quality Assessment and Risk of Bias
Most studies were considered high (n = 8) or medium (n = 19) quality (Supplemental Tables 2 and 3 https://links.lww.com/OLQ/A341 and https://links.lww.com/OLQ/A342). Low-quality studies (n = 3) are presented in the data tables, but not included in forest plots or discussed.40s,45s,47s
Despite medium/high quality, a number of methodological challenges remained. Nearly all studies relied on self-reported HC exposure, despite known limitations.9 Most studies (20 of 25) did not distinguish between combined or progestin only OCPs, and some did not distinguish between DMPA and Net-En injectable (3 of 19). The reference group of nonusers of a given HC was not defined consistently and sometimes included users of other forms of contraception. Most studies used empirically driven rather than theoretical adjustment for confounding. Nonsignificant estimates were not always presented, prohibiting information on the direction of association. For some studies, incidence rates were low, suggesting limited power.
Eleven studies evaluated the risk of HC on incident HPV infection and provide inconclusive evidence of association (Supplemental Table 4 https://links.lww.com/OLQ/A359, Fig. 2). All diagnostic tests were DNA-based and 5 assessed 1 or more high-risk HPV (HR-HPV) types, one assessed 1 or more low-risk HPV (LR-HPV), while 8 considered any HPV type; 2 disaggregated results more than 1 way. Four studies assessed the influence of injectables; 2 found that incidence of HR (one study) or any HPV (one study) was lower but not significantly lower compared to non-HC users.37s,41s A third study found recent DMPA users had increased incident HR-HPV (used in past 6 months adjusted odds ratio [aOR], 1.6; 95% confidence interval [CI], 0.7–3.7) and long-term users (≥1 year of use aOR, 4.7; 95% CI, 1.4–15.8) relative to nonusers of DMPA.36s Findings were in the same direction but not statistically significant among recent term and former users. The fourth study found nonsignificant results in mixed directions, depending on HPV type:44 DMPA use was associated with lower incidence of HR and increased risk of LR-HPV.
Ten studies evaluated OCP use. Three reported OCP use to be associated with increased HPV risk27,42s,44s 2 found nonsignificant increased risk,37s,41s 1 found significant decreased risk,26 2 reported nonsignificant decreased risk,28,36s 1 found no effect38s and 1 did not report the effect estimate for nonsignificant findings.43s Only 2 studies specified combined OCP use (COC), both documented a nonsignificant association.36s,41s Of the studies which documented evidence of increased risk, 1 was among OCP users versus non-OCP users in the last 3 months among LR-HPV (adjusted hazard ratio [aHR], 2.73; 95% CI, 1.52–4.90) and all-HPV types (aHR, 2.0; 95% CI, 1.28–3.15), but not HR-HPV types . Another study which also assessed OCP users vs. non-OCP users on all-HPV types found a lower magnitude of increased risk (aHR, 1.40; 95% CI, 1.01–1.80).27 The final significant finding of increased risk was documented in the longest exposure group only (7+ years) (aOR, 1.66; 95% CI, 1.17–2.35), with attenuated evidence of marginal risk in lower exposure groups (5–6 year and 3–4 year groups) and null effects among users <2 years relative to nonusers of HC.42s The 1 study that found significant decreased risk was among OCP using US women attending a family planning clinic relative to noncurrent OCP users (aHR, 0.49; 95% CI; 95% CI, 0.28–0.86).26 Overall, inconsistent exposure groups (current versus ever user), reference group (noncurrent versus never user) and differences in HPV subtype may contribute to disparate findings.
Only 1 study assessed the risk of hormonal IUD use on incident HPV infection. This retrospective record review compared levonorgestrel IUD users to copper IUD users and documented a 4-fold higher risk of HR-HPV among the former39s This effect was marginally significant and based on few incident cases.
Studies examining HSV-2 acquisition provide some evidence that injectable use increases risk35s,37s,54s and inconclusive evidence regarding OCPs35s,37s,53s,54s (Supplemental Table 5 https://links.lww.com/OLQ/A360).
Three studies examined the risk of injectable use on HSV-2 incidence. Two studies reported evidence of a significantly increased risk after injectable use (one specifies DMPA, the other is unspecified).35s,54s The remaining study reports evidence of nonsignificant increased risk (injectable type unspecified).37s The 2 studies that did not record the injectable type reported that DMPA was most common. Of the 2 studies that documented a significant effect, 1 study among HIV-negative women in Uganda reported increased risk (aOR, 2.26; 95% CI, 1.09–4.69) among consistent DMPA users, but not those who discontinued use, relative to non-HC users.35s The other study found DMPA use relative to non-DMPA use was strongly associated with HSV-2 acquisition among women both living with and without HIV (aHR, 4.43; 95% CI, 1.90–10.35), and when restricted to women living without HIV (aHR, 3.97; 95% CI, 1.64–9.60).54s The third study documented nonsignificant evidence of increased risk among HIV-negative women who engaged in sex work and used DMPA (aOR, 6.34; 95% CI, 0.25–158.5) compared to non-HC users,37s based on only 5 incident cases among DMPA users.
Three studies examined HSV-2 acquisition among OCP users: 2 documented a nonsignificant reduced risk of HSV-2 among OCP users relative to non-HC users.35s,53s The remaining study was in the harmful direction but was based on only 2 incident cases among OCP users.37
Seven studies provide inconclusive evidence of increased risk of chlamydia among injectable users21,22,37s,48s,50s–52sand 3 provide inconclusive evidence regarding OCP use23,37s,51s (Supplemental Table 6 https://links.lww.com/OLQ/A361, Fig. 3).
Of the 7 studies among injectable users, 3 documented a significant increased risk of acquisition among DMPA users.21,22,51s The magnitude of increased risk ranged between 1.6-fold (95% CI, 1.1–2.4) among DMPA users relative to women who were sterilized or using no contraception21 to 3.1-fold (95% CI, 1.0–9.4) among women living with HIV-1 who used DMPA compared with those who were sterilized or used IUD.22 The latter effect was marginally significant (P = 0.05). Four studies found a nonsignificant increased risk of acquisition among DMPA users relative to non-HC users:37s,48s,50s the direction of effect varies by the reporting period in 1 study but remains nonsignificant.50s An additional study documented a hazard ratio close to 1 among women who reported DMPA at any fourth month visit relative to non-DMPA users.52s Only 1 study compared Net-En users to non-HC users, and found a nonsignificant reduced risk of infection.48s
Six studies examined the incidence of chlamydia among OCP users,21–23,37s,51s,52s only 1 study specified combined or progestin-only pill use.23 Three studies documented significant evidence of increased risk.21,23,37s One study among HIV-negative women engaging in sex work in Rwanda compared OCP users with non-HC users (aOR, 6.13; 95% CI, 1.5–23.8).37s Results from this study are based on few incident cases. The 2 other studies documented significant increased risk of similar magnitude. One study compared OCP users with women who were sterilized or using no contraceptive (aHR, 1.80; 95% CI, 1.10–2.90),21 the other compared OCP users with women who were sterilized or using IUD (aHR, 1.73; 95% CI, 1.08–2.77).23 Three studies reported null findings. One study did not report the effect coefficient,52s and the other found nonsignificant reduced risk (aHR, 0.2; 95% CI, 0.0–1.7), among OCP users relative to non-HC users.51s
We found no significant prospective evidence that injectable use (5 studies)21,22,37s,48s,50s was associated with risk of gonorrhea. Of the 3 studies which compared injectable users to non-HC users, 2 studies found nonsignificant evidence of increased risk among DMPA users48s,50s 1 found nonsignificant evidence of reduced risk among Net-En users,48s and 1 study found nonsignificant evidence of reduced risk (injectable type unspecified).37s Two additional studies which examined DMPA use relative to women who were sterilized or used no contraception found an association close to the null.21,22 One of these was among women who were living with HIV-1.22 Only 1 of 4 studies of OCP use showed significant increased risk of gonorrhea21–23,37s (Supplemental Table 7 https://links.lww.com/OLQ/A362, Fig. 4). Information from the 1 study which found increased risk of gonorrhea after OCP use found nearly double risk (aHR, 1.7; 95% CI, 1.05–2.76) among COC users relative to women who used an IUD or were sterilized.23 This was the only study to assess pill formulation and found that a higher ratio of progestin in COC had a nonsignificant, but positive correlation with the risk of gonorrhea acquisition. The other 3 studies evaluating OCP use found results in mixed directions and did not specify pill type.
Two studies evaluated a combined group of women who tested positive for either C. trachomatis or N. gonorrhoeae due to small sample sizes (Supplemental Table 8 https://links.lww.com/OLQ/A363).25,55s A study among American STI patients found significant increased risk among DMPA users (aHR, 3.6; 95% CI, 1.6–8.5), and nonsignificant increased risk among COC users (aHR, 1.5; 95% CI, 0.6–3.5) relative to non-HC users.25 The second study was among HIV-1–positive women on antiretroviral therapy and was unable to evaluate OCP use due to no incident infections among users. However, women who used DMPA had more than 5 times the incident risk of N. gonorrhoeae or C. trachomatis (combined) (aOR, 5.83; 95% CI, 0.90–37.7), relative to non-HC users.55s
Two studies assessed HC use on syphilis incidence (Supplemental Table 9 https://links.lww.com/OLQ/A364), both which found nonsignificant results. One study found nonsignificant evidence of decreased risk among Kenyan women who engaged in commercial sex work and used OCPs (aHR, 0.40; 95% CI, 0.10–1.50) and DMPA (aHR, 0.50; 95% CI, 0.20–1.40), relative to women who used no contraception or were sterilized.21 The other study found nonsignificant evidence of increased risk among HIV-negative sex workers in Rwanda who used any injectable relative to non-HC users (aOR, 1.43; 95% CI, 0.11–19.1).37s The finding, however, is based on only 4 incident cases.
Studies of HC use on risk of trichomoniasis suggest injectables and OCPs are associated with reduced risk while findings are mixed regarding implant use (Supplemental Table 10 https://links.lww.com/OLQ/A365, Fig. 5).21,24,34s,37s,46s,48s–50s
All 7 studies that measured incident trichomoniasis suggest that injectable use reduced incidence by a magnitude ranging from 0.35 (95% CI, 0.12–1.01) to 0.70 (95% CI, 0.50–1.0), though some results were not statistically significant. Three studies found significant reduced risk after injectable use (2 specified DMPA and 1 was unspecified)21,34s,46s and 2 documented reduced risk that approached significance (1 specified DMPA, 1 was unspecified but DMPA use was most common).37s,48s Two of the studies which documented significant evidence of reduced risk compared HIV-1–negative injectable users (type unspecified) to non-HC users (aHR, 0.60; 95% CI, 0.47–0.78), and DMPA users (aHR, 0.60; 95% CI, 0.4–1.0; P = 0.04) to women who were sterilized or did not use contraception.21,34s [21,34]. The third study found women in Uganda who reported DMPA use in the past 12 months were at decreased risk compared to women who used neither HC nor condoms (adjusted incidence rate ratio [aIRR], 0.54; 95% CI, 0.30–0.98).46s Notably, the same study found nonsignificant findings of a similar magnitude among women who reported consistently using only DMPA at baseline and follow-up (aIRR, 0.59; 95% CI, 0.28–1.26). Only 1 study reported results for Net-En relative to non-HC use and found nonsignificant reduced risk.48s
Six of 7 studies that assessed OCP use and trichomoniasis documented reduced risk, although only 2 were significant. One significant finding was reported in a study among OCP users in 5 countries (Malawi, South Africa, the United States, Zambia, and Zimbabwe) who were significantly less likely to acquire Trichomonas vaginalis relative to non-HC users (aHR, 0.64; 95% CI, 0.47–0.89).34s The other was among OCP using women attending a STI clinic in the United States relative to those who used IUD or were sterilized (aHR, 0.56; 95% CI, 0.39–0.81).24 Only 1 study specified COC use.46s This study documented null findings among women in Uganda who reported COC use in the past 12 months (aIRR, 1.02; 95% CI, 0.40–2.59), or consistently using COCs in the past 12 months (aIRR, 1.07; 95% CI, 0.25–4.56) relative to no method (neither hormonal nor condom).
One of 3 studies which assessed implant use on incident trichomoniasis found a 3-fold increased risk of trichomoniasis (aIRR, 3.01; 95% CI, 1.07–8.49) among Norplant users relative to women who used no contraception method (hormonal or condoms) and slightly higher risk among consistent users of Norplant for 12 months (aIRR, 3.13; 95% CI, 1.08–9.07).46s The 2 remaining studies found no relationship between implant use (type unspecified) and trichomoniasis.34s,49s
Among studies of sufficient quality, DMPA use is consistently associated with a reduced risk of T. vaginalis acquisition, with evidence of substantial (2 times or higher) increased risk of HSV-2 incidence from a smaller number of studies. The results for HPV, chlamydia, gonorrhea and syphilis were inconclusive. Norethisterone enanthate was only assessed in 1 study.48s Data on OCP use suggest reduced incidence of trichomoniasis, with inconclusive findings for HPV, HSV-2, chlamydia, gonorrhea and syphilis. Implant use was less studied (n = 3), and only 1 specified type (Norplant). This study documented increased risk of trichomoniasis, but did not assess other STIs.46s Only 1 study assessed the levonorgestrel IUD and found a higher risk of HR-HPV incidence compared with the copper IUD; however, findings were marginally significant.39s
Findings from our study differ somewhat from 2 previous systematic reviews, which found inconclusive results for DMPA and OCPs on incident trichomoniasis, and increased risk of incident chlamydia.15,16 However, 1 previous review primarily synthesized cross-sectional research.16 In the second review, half of the studies (2 of 4 for trichomoniasis; 3 of 6 for chlamydia) did not include statistical adjustment for confounding.15 Those studies that reported adjusted T. vaginalis analyses also found decreased risk.21,24 Prior prospective evidence of incident HPV from 4 studies26–28,31s also suggest mixed results regarding the influence of OCPs and DMPA,26 without clear trends by HPV type or exposure time.
This review provides limited evidence that DMPA is associated with increased risk of HSV-2; we identified no prior review of HC use on incident HSV-2. Notably, our findings are based on a small number of studies. However, findings correspond with studies in mice which show heightened susceptibility to HSV-2 after prolonged (>2 weeks) treatment with DMPA.56s,57s These findings align with the 1 study that examined multiple exposure periods to DMPA and found a 2-fold increased risk of HSV-2 in consistent DMPA users relative to non-HC users but not among those who initiated, or discontinued use.35s A recent study in mice demonstrated that both DMPA and levonorgestrel, another progestin, increase mucosal epithelial permeability by acting on epithelial cell junction proteins (DSG1α), enhancing access of inflammatory and infectious viral molecules to the genital tissue, a possible biological mechanism.7 Given substantial evidence that HSV-2 increases risk of HIV infection,58s if the finding that DMPA increases the risk of HSV-2 is substantiated, this could be a mechanism for the association between DMPA use and HIV acquisition.
Further prospective research is warranted in several areas. Very few studies have explored the prospective association between HC use and syphilis (n = 3) or HSV-2 (n = 4) incidence. Similarly, few prospective studies have explored the potential risk of Net-En (n = 1), levonorgestrel IUD (n = 1) or implants on STIs (n = 3), while use of these methods is increasing.59s No reviewed studies evaluated Sayana Press, the Nuva Ring, or patch. Current large-scale prospective studies of HIV risk among women should incorporate well measured contraceptive use and STI outcomes to help address these gaps. Further, many of the studies of OCPs did not differentiate between combined or progestin-only OCPs, and similarly some injectable studies did not differentiate between Net-En and DMPA. Given that biological responses to HC differ by class of drug and drug formulations,5 future research needs to distinguish between HC formulations when estimating risk of STI/HIV acquisition.
This updated systematic review of prospective evidence published between 2009 and 2017 suggests that DMPA and OCP use are associated with a reduced risk of incident trichomoniasis, with evidence of increased substantial risk of HSV-2 acquisition with DMPA use from a small number of studies. Our review findings are tempered by notable methodological limitations. Prospective evidence regarding the STI risk of hormonal contraceptive methods are extremely limited or nonexistent, highlighting an urgent research need.
1. Ralph LJ, McCoy SI, Shiu K, et al. Hormonal contraceptive use and women's risk of HIV acquisition: A meta-analysis of observational studies. Lancet Infect Dis 2015; 15:181–189.
2. Morrison CS, Chen PL, Kwok C, et al. Hormonal contraception and the risk of HIV acquisition: An individual participant data meta-analysis. PLoS Med Public Library of Science 2015; 12:e1001778.
3. Polis CB, Phillips SJ, Curtis KM, et al. Hormonal contraceptive methods and risk of HIV acquisition in women: A systematic review of epidemiological evidence. Contraception 2014; 90:360–390.
4. Polis CB, Curtis KM, Hannaford PC, et al. An updated systematic review of epidemiological evidence on hormonal contraceptive methods and HIV acquisition in women. AIDS 2016; 30:2665–2683.
5. Hapgood JP, Kaushic C, Hel Z. Hormonal contraception and HIV-1 acquisition: Biological mechanisms. Endocr Rev 2018; 39:36–78.
6. Miller L, Patton DL, Meier A, et al. Depomedroxyprogesterone-induced hypoestrogenism and changes in vaginal flora and epithelium. Obstet Gynecol 2000; 96:431–439.
7. Quispe Calla NE, Vicetti Miguel RD, Boyaka PN, et al. Medroxyprogesterone acetate and levonorgestrel increase genital mucosal permeability and enhance susceptibility to genital herpes simplex virus type 2 infection. Mucosal Immunol 2016; 9:1571–1583.
8. Wira CR, Fahey JV, Ghosh M, et al. Sex hormone regulation of innate immunity in the female reproductive tract: The role of epithelial cells in balancing reproductive potential with protection against sexually transmitted pathogens. Am J Reprod Immunol 2010; 63:544–565.
9. Achilles SL, Hillier SL. The complexity of contraceptives: Understanding their impact on genital immune cells and vaginal microbiota. AIDS 2013; 27(Suppl 1):S5–S15.
10. Jespers V, Kyongo J, Joseph S, et al. A longitudinal analysis of the vaginal microbiota and vaginal immune mediators in women from sub-Saharan Africa. Sci Rep 2017; 7:11974.
11. Straub RH. The complex role of Estrogens in inflammation. Endocr Rev 2007; 28:521–574.
12. Ghanem KG, Shah N, Klein RS, et al. Influence of sex hormones, HIV status, and concomitant sexually transmitted infection on cervicovaginal inflammation. J Infect Dis 2005; 191:358–366.
13. Ott MA, Adler NE, Millstein SG, et al. The trade-off between hormonal contraceptives and condoms among adolescents. Perspect Sex Reprod Health 34:6–14.
14. Mullinax M, Sanders S, Dennis B, et al. How condom discontinuation occurs: Interviews with emerging adult women. J Sex Res 2017; 54:642–650.
15. Morrison CS, Turner AN, Jones LB. Highly effective contraception and acquisition of HIV and other sexually transmitted infections. Best Pract Res Clin Obstet Gynaecol 2009; 23:263–284.
16. Mohllajee AP, Curtis KM, Martins SL, et al. Hormonal contraceptive use and risk of sexually transmitted infections: A systematic review. Contraception 2006; 73:154–165.
17. Jones RK, Darroch JE, Henshaw SK. Contraceptive use among U.S. women having abortions in 2000-2001. Perspect Sex Reprod Health 2002; 34:294–303.
18. van de Wijgert JH, Verwijs MC, Turner AN, et al. Hormonal contraception decreases bacterial vaginosis but oral contraception may increase candidiasis. AIDS 2013; 27:2141–2153.
19. Vodstrcil LA, Hocking JS, Law M, et al. Hormonal contraception is associated with a reduced risk of bacterial vaginosis: A systematic review and meta-analysis. PLoS One 2013; 8:e73055.
20. Wells G, Shea B, O'Connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. In: Ottawa Hospital Research Institute [Internet]. [cited 30 Nov 2017]. Available: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp
21. Baeten JM, Nyange PM, Richardson BA, et al. Hormonal contraception and risk of sexually transmitted disease acquisition: Results from a prospective study. Am J Obstet Gynecol 2001; 185:380–385.
22. Lavreys L, Chohan V, Overbaugh J, et al. Hormonal contraception and risk of cervical infections among HIV-1-seropositive Kenyan women. AIDS 2004; 18:2179–2184.
23. Louv WC, Austin H, Perlman J, et al. Oral contraceptive use and the risk of chlamydial and gonococcal infections. Am J Obstet Gynecol 1989; 160:396–402.
24. Barbone F, Austin H, Louv WC, et al. A follow-up study of methods of contraception, sexual activity, and rates of trichomoniasis, candidiasis, and bacterial vaginosis. Am J Obstet Gynecol 1990; 163:510–514.
25. Morrison CS, Bright P, Wong EL, et al. Hormonal contraceptive use, cervical ectopy, and the acquisition of cervical infections. Sex Transm Dis 2004; 31:561–567.
26. Moscicki AB, Hills N, Shiboski S, et al. Risks for incident human papillomavirus infection and low-grade squamous intraepithelial lesion development in young females. JAMA 2001; 285:2995–3002.
27. Winer RL, Lee S-K, Hughes JP, et al. Genital human papillomavirus infection: Incidence and risk factors in a cohort of female university students. Am J Epidemiol 2003; 157:218–226.
28. Sellors JW, Karwalajtys TL, Kaczorowski J, et al. Incidence, clearance and predictors of human papillomavirus infection in women. CMAJ 2003; 168:421–425.
29. Avonts D, Sercu M, Heyerick P, et al. Incidence of uncomplicated genital infections in women using oral contraception or an intrauterine device: A prospective study. Sex Transm Dis 1990; 17:23–29.
30. Rahm VA, Odlind V, Pettersson R. Chlamydia trachomatis
in sexually active teenage girls. Factors related to genital chlamydial infection: A prospective study. Genitourin Med 1991; 67:317–321.