Intrauterine device use, sexually transmitted infections, and fertility: a prospective cohort study

Abstract

In the 1970s, numerous medical reports, media coverage, and litigation around the Dalkon Shield intrauterine device led to a perception that all intrauterine devices cause upper genital tract infection and infertility. This study aimed to assess the association between intrauterine device use and time to conception. The Fertility After Contraceptive Termination Study is a multicenter, prospective cohort study of women stopping their contraceptive method to attempt conception. We recruited participants between 2011 and 2017. Participants were a convenience sample of women recruited from academic centers in Philadelphia, PA; Los Angeles, CA; St. Louis, MO; Indianapolis, IN; Aurora, CO; and Salt Lake City, UT. Women were eligible if they stopped their contraceptive method within the past 120 days before enrollment, were between 18 and 35 years of age, had no history of infertility or sterilization, and had at least 6 months of follow-up. Baseline data included demographic and reproductive characteristics, past contraceptive use, nucleic acid amplification testing for sexually transmitted infections, and serology for past infection with Chlamydia trachomatis, Trichomonas vaginalis, and Mycoplasma genitalium. The primary exposure was intrauterine device use (ever); the primary outcome was time to conception. All participants were observed longitudinally for up to 24 months. We used piecewise exponential proportional hazards models with multiple imputation to provide hazard ratios and their respective 95% confidence intervals. Of the 461 participants, mean age was 28.2 years, 178 (38.7%) were Black, 157 (34.1%) were considered as low socioeconomic status, and 275 (59.7%) had a history of intrauterine device use. Without adjusting for any covariates, the median time to conception was shorter for participants who had a history of intrauterine device use (5.1 months) than participants who never used an intrauterine device (7.5 months). After controlling for potential confounders, the association of past intrauterine device use with time to conception was not statistically significant (adjusted hazard ratio, 1.25; 95% confidence interval, 0.99-1.58). In our multivariable model, age, nulligravidity, Black race, low socioeconomic status, and past Mycoplasma genitalium infection were associated with longer times to conception (hazard ratio, 0.76; 95% confidence interval, 0.58-0.99). Conception by 12 months was lower in participants with past Mycoplasma genitalium infection (68% vs 80% without past infection; P=.019). We found no impairment of fertility with ever use of an intrauterine device. Serologic evidence of past Mycoplasma genitalium infection was associated with longer times to conception and higher rates of infertility. Mycoplasma genitalium infection is a potential modifiable cause of infertility.