Female sexual function and distress and time-to-pregnancy in a prospective preconception cohort.

In this North American preconception cohort study, we did not find an association between FSD and TTP using a previously-validated cut point. 21 However, we did find an association between quartiles representing lower sexual function and delayed conception. We also found an association of sexual distress and frequent painful intercourse with delayed conception. Exploratory analyses suggested some evidence of an association of lower levels of lubrication and orgasm frequency and delayed conception. Our results are consistent with another study that evaluated preconception female sexual function and TTP. 17 That study reported a FR of 0.73 for participants with lower sexual function compared with those having higher function. They defined ‘lower sexual function’ as the median value of their sample; our quartile-based analysis suggested that participants in the highest quartile of sexual function scores had the shortest TTP. We report an association between sexual distress and prolonged TTP. Those reporting sexual distress also reported less frequent intercourse, so intercourse frequency may play a role in this association. Another potential mechanistic pathway is stress. Prior research has reported associations between high stress levels and longer TTP. 28 , 43 Our observed association persisted despite adjustment for general stress, as measured by the PSS-10. 25 Conception attempts may exacerbate distressing sexual health issues, causing a stress response that affects reproductive function. 45 It is also possible that low intercourse frequency causes sexual distress and is not a mediator. We cannot determine the temporality of intercourse frequency and distress in our data. We found that frequent painful intercourse was associated with longer TTP but occasional pain was not. Inadequate lubrication is associated with painful sex; 46 therefore, participants who occasionally experience painful sex may preferentially time intercourse. Because cervical mucus and vaginal lubrication are increased during the fertile window 47 – 50 and hormonal fluctuations around ovulation are associated with increased sexual desire, 51 pain may decrease as conception chances increase. Relatedly, it is possible that frequent pain interferes with the frequency and timing of intercourse; the small sample size precluded a formal mediation analysis. Painful intercourse may be caused by upstream factors (e.g., hormones) that also influence TTP. 52 Providers may wish to ask patients about painful intercourse during preconception counseling, as personal lubricant use may ameliorate pain 53 and is not associated with TTP. 54 Exploratory analyses suggested that the association of sexual function and TTP differed by domain. Three domains (desire, satisfaction, and arousal) had little or no relationship to TTP, while lower lubrication and orgasm frequency had an association with longer TTP. Physiologic sexual function is regulated by hormones that also play a role in reproduction (i.e., estrogen) and accompanied by increased blood flow to vulvar tissues, which cause lubrication. 55 Thus, it is possible that hormonal imbalances or vascular issues that impair sexual function also play a role in conception. Additionally, there is a wealth of literature on the evolutionary origins of the human female orgasm, some of which hypothesizes that it is an adaptive mechanism to aid conception, 56 , 57 though to our knowledge this is the first study to directly evaluate orgasm in relation to TTP. Our study provides evidence that preconception sexual function may play a role in TTP, which warrants further investigation. Our research suggests potential limitations of the FSFI-6 in etiologic research. Use of the full 19-item Female Sexual Function Index 22 provides a more detailed assessment of sexual function and may offer researchers more comprehensive insight. Validated scales of domain-specific sexual function may also be beneficial for future research. Strengths of our study include the preconception assessment of sexual function in a prospective cohort study and the use of multiple validated instruments to measure female sexual function and distress. Limitations include the use of a different lubrication question than the original FSFI-6 21 to increase specificity. Though the responses to both questions were highly correlated, direct comparisons with other studies are more challenging. In addition, the FSFI-6 was designed for clinical assessment, not etiologic research. 21 Indeed, our findings suggest that important domain-specific information may be lost when pooling responses into a single measure. Our domain-specific findings were exploratory because domain-specific questions have not been validated independently. Our sample was primarily non-Hispanic White with high levels of education and income, and were a self-selected population of pregnancy planners, which may influence the distribution of potential confounders (i.e., depression or anxiety) and overall generalizability. Our results are vulnerable to selection bias due to left truncation. 42 Though our sensitivity analysis with more restrictive criteria for inclusion (i.e., ≤3 cycles of attempts) did not substantially alter our findings, left truncation bias remains a potential limitation. The small sample size in some exploratory analyses limited the precision of the estimates. Finally, not every participant completed the SHQ. IPW suggested little influence of participation bias on our results, but our model may not have included all relevant factors. Our findings indicate an association of higher levels of sexual dysfunction and distress and frequent painful intercourse with delayed conception. Exploratory analyses suggested associations between more difficulty of lubrication, and lower frequency of orgasm with longer TTP. Because individuals rarely discuss sexual health with healthcare providers, 18 , 58 – 63 sexual function issues may be an underappreciated contributor to prolonged conception attempts. Clinicians should consider discussing sexual function in preconception counseling visits.

We used data from Pregnancy Study Online (PRESTO), an ongoing web-based prospective cohort study. 20 PRESTO enrolls female-identified participants attempting pregnancy with a male partner who reside in the United States (US) or Canada, are not using fertility treatments or contraception, and are aged 21–45 years. Recruitment occurs via online advertising, posted flyers, and word of mouth. Participants complete online questionnaires at enrollment and every 8 weeks for up to 12 months or until pregnancy occurs, whichever comes first. In March 2021, we added the Sexual Health and Well-Being Questionnaire (SHQ) to the PRESTO protocol. This optional questionnaire asks about participants’ sexual health and feelings about their sex life. All former and already-enrolled participants received an email inviting them to complete the questionnaire. Participants who enrolled after SHQ launch were invited via email 30 days after enrollment, though completion was possible any time after enrollment. Approximately 60% of prospectively-enrolled participants completed the SHQ. Research suggests that declines in sexual function begin after 6 months of conception attempts; 18 thus, we restricted our sample to those who completed the SHQ no later than 6 months after their self-reported conception attempt initiation to reduce the risk of reverse causation (see Figure 1 ). The study protocol was approved by the Boston University Medical Campus Institutional Review Board, and participants provided informed consent online. Sexual dysfunction and distress were ascertained using two validated scales included in the SHQ. Participants answered a modified version of the 6-item Female Sexual Function Index (FSFI-6) scale, a clinically validated instrument designed to identify FSD. 21 The FSFI-6 is a reduced version of the 19-item Female Sexual Function Index 22 which asks participants to assess their sexual function over the past 4 weeks across six domains: interest, arousal, lubrication, orgasm frequency, pain, and satisfaction. Responses to all questions were collected via a Likert scale ranging from 1 (lowest function) to 5 (highest function). We replaced the original FSFI-6 lubrication question (i.e., how often did you become aroused during sexual activity?) with the question ‘Over the past 4 weeks, how difficult was it to become lubricated (“wet”) during sexual activity or intercourse (only consider lubricant your vagina produces, not use of bottled lubrication)?,’ drawn from the full Female Sexual Function Index, 22 which we perceived as more specific. In a small validation study of 291 participants who completed both questions, the Pearson correlation coefficient was 0.79. Our replacement question (lubrication difficulty) had slightly higher mean responses than the original question (lubrication frequency) (4.11 vs 4.00). Participants who reported no sexual activity (n=23) or no sexual activity or intercourse (n=28) in the past four weeks received a score of zero on questions requiring activity, as previously described. 21 We summed scores (possible scores ranging from 2 to 30) and assigned those with a score ≤19 as having FSD, consistent with prior research. 21 We also defined quartiles based on the distribution of scores in the sample. To reduce the potential for false positives due to transient pain, we asked participants if they had experienced an infection (e.g., yeast infection) causing vaginal pain in the previous 4 weeks. We reassigned those reporting an infection and meeting the threshold for FSD as not having FSD (resulting in the reclassification of N=15, 0.6%). We assessed sexual distress using the Female Sexual Distress Scale (FSDS), 23 a 12-item measure that assesses feelings related to sex in the past 4 weeks. Responses were on a Likert scale that ranged from “never” (assigned a value of 0) to “always” (value of 4). We summed responses (possible range of 0–48), used a score ≥20 as the cutoff for clinically relevant sexual distress, 23 and calculated quartiles. We used a single question from the FSFI-6 to assess painful intercourse: “Over the past 4 weeks, how often did you experience discomfort or pain during vaginal penetration (intercourse)?” Response options included “Did not attempt vaginal penetration,” “Almost never or never,” “A few times,” “Sometimes,” “Most times,” or “Almost always or always.” For these analyses, we excluded participants who did not attempt vaginal penetration in the past 4 weeks (n=56) and those who reported the presence of an infection which caused pain (n=79). We categorized those who reported that they experienced pain “A few times” or more frequently as having pain with intercourse. We additionally created a three-level variable for experiences of painful intercourse in which the categories were “Always, almost always, or most times,” “Sometimes,” and “Never, almost never, or a few times.” We calculated TTP using data from baseline and follow-up questionnaires. At baseline, participants reported the date of their last menstrual period (LMP), usual menstrual cycle length, and the number of menstrual periods they had since they initiated conception attempts. At follow-up, participants again reported LMP date and whether they had conceived since the previous questionnaire. Usual cycle length for participants reporting irregular cycles was calculated using baseline LMP and consecutive LMPs reported on follow-up questionnaires. We calculated total pregnancy attempt time based on the total discrete cycles at risk of pregnancy using the following formula: cycles of attempt at study entry + [(LMP date from most recent follow-up questionnaire – date of baseline questionnaire completion)/usual cycle length] + 1. Covariates were ascertained via-self report on the baseline questionnaire. They included history of diagnosed medical conditions (endometriosis, polycystic ovarian syndrome, diabetes, depression, anxiety, post-traumatic stress disorder, and a history of any sexually transmitted infection or fibroids), age (<25, 25–29, 30–34, 35–39, ≥40 years), marital status (married, unmarried), body mass index (BMI) (<18.5, 18.5–24, 25–29, ≥30 kg/m 2 ), measures of current socioeconomic status: education (≤high school, some college, college, graduate school), annual household income (<$50,000, 50,000–99,000, 100,000–149,000, ≥150,000 USD), health insurance coverage (private insurer, other), race/ethnicity (Non-Hispanic Asian, non-Hispanic Black, Hispanic/Latina, Mixed/other race, non-Hispanic White), employment status (employed, unemployed), employment factors including average hours worked per week and past-month rotating shifts or work at night, behavioral factors including average alcoholic drinks consumed per week, cigarette smoking status (never, former, current), any cannabis use in the past 4 weeks, and sleeping on average <7 hours per night in the past 4 weeks. We estimated physical activity levels by asking participants to report the average hours per week spent engaging in a number of different activities in the past year which were then used to calculate the average metabolic equivalent (MET) hours per week and categorized as <10, 10–19, 20–39 and ≥40. 24 We assessed general stress using the 10-item version of the Perceived Stress Scale, 25 categorized as <10, 10–19, 20–29, and ≥30. Reproductive factors included a history of live birth, and whether participants had ever tried unsuccessfully to conceive for 12 months (history of infertility). Participants reported on intercourse frequency via the question “In the past month, about how often did you have sexual intercourse with your partner (your partner’s penis inserted into your vagina)?” and categorized as <1, 1, 2–3 or ≥4 times per week. Intercourse frequency was identified as a potential mediator, and thus was not included in adjusted models. Some covariate information was collected on an optional supplemental questionnaire, which participants were invited to complete 30 days after enrollment. Respondents answered to a question from the Brief Trauma Questionnaire: 26 “Has anyone ever made or pressured you into having some type of unwanted sexual contact?” with the responses ‘Yes,’ ‘No,’ or ‘Don’t know.’ Responses for who responded ‘Don’t know’ were imputed. Participants also responded to two questions related to their perception of their relationship in the past year, specifically “To what extent did your partner show you love and affection?” and “To what extent could you count on your partner to provide you with emotional support?” Responses were categorized as ‘never/rarely’, ‘some of the time,’ ‘most of the time,’ ‘all of the time.’ Fifty-nine participants in our analytic dataset did not complete this supplemental questionnaire; their responses were multiply imputed. We calculated age-standardized baseline characteristics stratified by sexual dysfunction, sexual distress, and painful intercourse. We used proportional probabilities regression models to estimate fecundability ratios (FRs) and 95% confidence intervals (CIs) for associations between each exposure and fecundability. 27 The FR compares the average per-cycle probability of conception in exposed and unexposed participants. Participants contributed at-risk cycles from cohort entry until pregnancy or a censoring event (i.e., initiation of fertility treatment, cessation of pregnancy attempts, study withdrawal, loss to follow up, or 12 cycles of unsuccessful conception attempts), whichever occurred first. Participants who entered the study with ≥1 cycle of attempt time contributed cycles at risk beginning with their first cycle under observation (i.e., the first cycle at risk contributed to the analysis for those with 2 cycles of attempt time at study entry was cycle 3). We selected potential confounders using a directed acyclic graph (DAG) 28 (see Supplemental Figure 1 ) informed by published literature. 18 , 29 – 39 A DAG is a visual depiction of the relationship under study, including factors associated with the exposure and outcome, that informs the selection of adjustment covariates. 28 For the primary analysis, we calculated crude FRs, age-adjusted FRs, and FRs adjusted for all prespecified confounders (listed above). For pain-related exposures, we calculated additional models where we omitted psychosocial confounders potentially less relevant to physiological causes of pain (current employment, hours worked per week, shift work, work at night, short sleep duration, marital status, and physical activity). Missing data were minimal (<5%). We imputed missing values for questions in the FSFI-6 and FSDS and covariates using fully conditional specification methods to create 20 imputed datasets. 40 We assigned those without any follow-up information (2.3%) one cycle of follow-up and multiply imputed their pregnancy status (yes vs no). Missingness for all variables is displayed in Supplementary Table 1 . Participation bias may have occurred because not every PRESTO participant completed the SHQ. 41 We used inverse probability of treatment weighting (IPW) to account for this bias, adjusting for factors related to response and TTP. For more details, please see Appendix I . To explore the shape of the association between sexual function and distress with TTP, we fit restricted cubic splines for the continuous modified FSFI-6 and FSDS scores adjusting for all prespecified confounders. We also created three-level variables for the five individual sexual function domains besides pain (interest, arousal, lubrication, orgasm, and satisfaction) in which we grouped the highest two and lowest two categories, respectively, into a ‘high’ and ‘low’ function group and evaluated the association between each individual domain and TTP. We conducted an exploratory analysis to investigate the role of partner sexual function (see details in Appendix II ). To reduce the potential for reverse causation and selection bias due to left truncation, 42 we repeated the primary analysis restricting the sample to those who had ≤3 cycles of conception attempts at the time of SHQ completion. As recommended by the American Statistical Association, we interpreted our findings based on the magnitude, direction, and precision of the effect estimates, as opposed to binary significance testing,. 43

Age-standardized baseline characteristics for the sample (N=2,500) are displayed in Table 1 . Participants enrolled between 2021 and 2024. Means and percentages of evaluated characteristics are standardized to the sample age distribution at baseline (in categories), to facilitate a qualitative assessment of differences in characteristics across exposures after accounting for differences in age. Most of the sample was aged 25–34 years, married, and nulliparous. Over 80% of participants had at least a college degree and over half of the participants reported an annual household income of at least US$100,000. Over 80% of the sample identified as non-Hispanic White. Sexual function issues were more common among those with a history of depression, anxiety, a history of infertility, and greater self-reported stress. Those with sexual function issues reported less frequent intercourse ( Supplementary Figure 2 ). Exposure prevalence was 20.1% for FSD (modified FSFI-6 score ≤19), 8.8% for distress (FSDS ≥20), and 29.6% for any pain with intercourse. We did not see evidence of an association between FSD and TTP when using the previously-established cutpoint for FSD 21 (fully-adjusted FR 1.01, 95% CI 0.90–1.14). However, we did observe a reduction in fecundability among those in the first (score range 0–20), second (score range 21–22), and third (score range 23–25) quartiles of sexual function scores compared with those in the fourth quartile (highest function, score range 26–30). We found reduced fecundability among those with sexual distress as determined by the published cut point 23 (fully-adjusted FR 0.82, 95% CI 69–0.98) and among those in the fourth (score range 12–48), third (score range 6–11), and second (score range 2–5) quartiles, compared with the first (lowest distress, score range 0–1) ( Table 2 ). Spline results indicated that FRs declined steadily with decreasing scores on the modified FSFI-6, plateauing around 22. There was a steep decrease in fecundability with increasing FSDS ( Figure 2 ). ‘Any pain with intercourse’ was not appreciably associated with reduced fecundability (fully-adjusted FR 1.03, 95% CI 0.93, 1.14) but reporting painful intercourse ‘most times, almost always or always’ was associated with a longer TTP compared with ‘almost never or never’ (fully-adjusted FR 0.81, 9% CI 0.62, 1.06) ( Table 3 ). Accounting for participation bias using IPW did not meaningfully change the results ( Tables 2 and 3 ). Neither restricting the sample to those with ≤3 cycles of conception attempts nor to those without a partner issue meaningfully changed the results ( Supplementary Tables 2 & 3 ) and Supplementary Table 4 . Distributions of individual domain responses are presented in Supplementary Figure 3 . We observed little to no association between reported level of sexual desire, arousal, or satisfaction in the past four weeks and TTP. Lower self-reported orgasm frequency and lubrication were associated with longer TTP ( Table 4 ). The strongest effect estimate was a U-shaped association for orgasm frequency, with those who reported orgasming ‘sometimes’ and ‘never, almost never, or a few times’ having adjusted FRs of 0.74 (95% CI 0.65, 0.85) and 0.93 (95% CI 0.83, 1.03), respectively, compared with those who report orgasming ‘most times, almost always, or always.’

Female sexual dysfunction (FSD) is characterized by an impaired ability to participate in or enjoy sexual activity accompanied by personal distress. 1 FSD manifests across different domains, including sexual arousal/desire, orgasm, lubrication, and pain. 1 Sexual dysfunction is associated with reduced quality of life 2 – 4 and bidirectionally associated with poorer mental health 5 and lower relationship and life satisfaction. 6 , 7 Sexual function is relevant for conception among mixed-sex couples not using fertility treatment because frequency and timing of sexual intercourse affect time-to-pregnancy (TTP). 8 While infertility has been shown to impair sexual function, 9 – 16 little research has evaluated the role of sexual function in TTP. In a Singaporean study of pregnancy planners, lower female sexual function was associated with prolonged TTP, though the authors did not use an externally-validated definition of sexual dysfunction. 17 Because sexual function issues (e.g., painful intercourse) are prevalent in the preconception period 18 and healthcare providers do not routinely discuss sexual function with patients, 18 , 19 sexual dysfunction may be an under-recognized cause of prolonged TTP. Our objective was to assess the effects of sexual dysfunction and distress on TTP in a North American cohort of pregnancy planners. We also separately evaluated individual domains of female sexual function in relation to TTP.