Many pregnant women (and some physicians) believe that sexual intercourse at term will hasten the onset of labor. One recent study showed that 73.6% of pregnant women were aware of this concept, and 46.1% believed it to be true.1 Several identified a physician or health care worker as the source of the information. The willingness to accept this belief is understandable, given that there is some biological plausibility regarding the effect of intercourse on initiating labor. Semen contains prostaglandins, and prostaglandin concentrations in the cervical mucus of pregnant women have been demonstrated to be 10- to 50-fold higher than normal 2–4 hours after intercourse.2 Sexual intercourse may also have an effect on uterine contractility. Contraction monitoring of pregnant women during intercourse in late pregnancy has revealed increased uterine activity after coitus, although it is unclear from this study whether this was an effect of maternal or paternal orgasm.3 Maternal orgasm alone, achieved by genital manipulation without intercourse, has been associated with increased uterine contractions.4
Despite the possibility of an effect on labor, there is little evidence to suggest that coitus can initiate labor. Sexual activity in low-risk pregnancies is not associated with preterm birth, premature rupture of membranes, or low birth weight.5 Although one report suggests an increased risk of preterm rupture of membranes after sexual activity, this association was not present in term pregnancies.6 There are few data available to adequately counsel pregnant women at term about the potential effect of coitus on labor and delivery.
The objective of this study was to determine whether women who are sexually active at term are more likely to enter labor than those women who are not. The study also investigates the effect of term intercourse on cervical change. A secondary objective was to provide data on the safety of engaging in sexual intercourse at term.
Provided that prostaglandins may be used as a cervical ripening agent, the study also examines the effect of term intercourse on cervical change on a weekly basis. The hypothesis would be that, even if labor does not ensue earlier after intercourse occurs, the cervix may become more favorable in the presence of seminal prostaglandins.
MATERIALS AND METHODS
Over the 1-year period from July 2004 to July 2005, women receiving prenatal care at term were approached for enrollment into the study. All subjects were patients of the author in his practice at the Ohio State University Medical Center. All women asked to participate were at least 18 years of age and had low-risk singleton pregnancies. Low risk was defined as being free of any chronic medical conditions that would interfere with sexual activity or fetal issues that would require activity restrictions. Patients were excluded from consideration if they had any contraindication to vaginal delivery, such as active genital herpes, breech presentation, or placenta previa, or if they had a cesarean delivery scheduled. Informed consent was obtained, including a discussion of use of medical information, and the study was approved by the Institutional Review Board of the Ohio State University.
Each participant in the study was enrolled at the next office visit after completing her 37th week of pregnancy. At this visit and each subsequent routine visit, the subject was asked whether she had engaged in penile-vaginal intercourse during the preceding week. If so, the subject was also asked how many episodes of intercourse occurred and whether a condom was used. Women who had intercourse at any time during term were assigned to the sexually active group. Cervical examination was performed at each routine weekly visit, and a Bishop score was assigned.7
After delivery, information was obtained from the hospital's computerized database. Gestational age at delivery, mode of delivery, length of labor, and use of induction were noted and compared between the sexually active and inactive groups. In addition, the infant's birth weight, Apgar scores, and need for neonatal intensive care unit (NICU) admission were noted.
To detect a 1-week difference in gestational age at delivery (with an estimated variance of 1 week) between groups at a significance level of .05 with a power of 80%, it was calculated that each group would need to include at least 16 patients. Although estimates of sexual activity at term vary widely, the largest study describing sexual behavior at term identified that 26% of women between 36 and 37 weeks were still sexually active.8 It was therefore calculated that at least 64 women would have to be enrolled to yield a sexually active group of sufficient size.
Two-sample t tests were used to compare continuous measurements between the sexually active and the sexually inactive groups. Comparisons involving categorical variables used χ2 tests. Fisher exact test was used instead when small expected frequencies were present. To evaluate the relationship of the Bishop score to presence or absence of sexual activity in the previous week, a generalized linear model for repeated measures was used.9
A total of 95 women were enrolled in the study over the course of a single year. All women who were approached to participate consented to enrollment. Two women delivered at another institution and were lost to follow-up.
Forty-seven (50.5%) of 93 subjects had intercourse at some point in the final weeks of pregnancy. The range of frequency of intercourse in any given week varied from 0 to 7 times, with the mean number of episodes per week in this group decreasing over time from 1.8 after 37 weeks to 0.5 in those patients still pregnant after 40 weeks. The median sexual frequency over time ranged from 1 at week 37 to 0 at week 40 for the sexually active group.
Table 1 shows the demographic data for the two groups. The two groups did not differ significantly in terms of age, weight, gravidity, or parity. In this population, the number of unmarried gravidas was a substantial majority, but the marital status had no impact on the presence of sexual activity.
The delivery outcomes for the two groups are presented in Table 2. The sexually active group delivered at an average gestational age of 39.9 weeks, compared with 39.3 weeks in the abstinent group (P = .001). The interval from the first term prenatal visit to delivery is also significantly shorter in the group abstaining from sex.
Although fewer total inductions were performed in the abstinent group, there was only one subject in each group induced for postdue date (past 41 completed weeks of pregnancy). Six inductions in the sexually active group and one in the abstinent group were elective. No significant difference existed between the mode of delivery in the two groups. The presence or absence of sexual activity had no relationship to the infant's birth weight, Apgar scores, or likelihood of being admitted to the NICU.
Figure 1 illustrates the average Bishop scores over time in patients reporting sexual activity or no activity in the week before the measurement. There was no significant difference between groups at any given gestational age. Although, as expected, the trend was for average Bishop score to increase as gestational age increased, a decrease could be noted in weeks after subjects with higher Bishop scores delivered. A generalized linear model for repeated measures was used to analyze the data. Because there seemed to be no interaction between time and sexual activity in the prior week (P = .683), the final model included only the main effects. After adjusting for the effect of time (β2 = 1.06, 95% confidence interval [CI] 0.67 to 1.45, P < .001), the estimated effect of engaging in intercourse during the preceding week indicated an average Bishop score that was 0.26 lower (95% CI −1.26 to 0.74, P = .61) than the average of those who abstained.
The effect of frequency of intercourse on Bishop score was also examined. Figure 2 shows a scatterplot of Bishop scores at each gestational age for the number of episodes of intercourse in the preceding week. These graphs demonstrate a high amount of variability in the cervical exams at any gestational age. There is no correlation between the frequency of sexual intercourse and Bishop score. Again, this relationship was evaluated using a generalized linear model. Results indicate that, after adjusting for the effect of time, for each unit increase in frequency of intercourse, the Bishop score decreased on average by 0.16 (95% CI −0.47 to 0.15, P = .3). Only one woman in the study reported that her partner used a condom, precluding any analysis of this issue.
The results of this study indicate that there was no meaningful relationship between sexual intercourse at term and the onset of labor. In contrast, a prolongation of pregnancy in women who are sexually active in the final weeks of pregnancy was observed in this study. This prolongation of 4 days is not clinically significant. Because the variability in gestational age was small, however, the difference between groups achieves statistical significance. The finding that greater cervical change does not occur in sexually active women at term also supports the lack of an effect of sexual activity on inducing labor. Apparently, there is no relationship between coital frequency and onset of labor.
These findings agree with the results of a British retrospective study demonstrating an increased duration of gestation with increasing coital frequency at all gestational ages up through 37 weeks.8 Similarly, low-risk women delivering prematurely in another study were less likely to be sexually active than the full-term controls.10
One possible explanation for these findings is that women may continue to be sexually active if they remain in a state of relative comfort. Women who are experiencing greater abdominal discomfort, pelvic pressure, or increased contractions are less likely to remain sexually active at term. Such symptoms might be markers for earlier delivery.
All of these studies examine sexual behavior in pregnancy retrospectively and thus are subject to several biases. There is only one prospective, randomized study in the literature that examined the effect of intercourse on initiating labor.11 Twenty-eight women at term were randomly assigned to have intercourse for three consecutive nights or abstain. There was no difference between the groups afterward in Bishop scores or number of women in labor. This study limits this intervention to a short course and does not describe the effect of preceding sexual activity.
The personal and highly variable nature of sexual behavior is such that it is not feasible to carry out a prospective study that randomly assigns subjects to a particular diet of sexual activity in pregnancy. Sexual activity may occur for a variety of reasons and in a variety of situations that make it very difficult to identify what confounding factors are present that influence the onset of labor. For example, sexual activity may or may not include breast stimulation, which has been demonstrated to have a positive effect on inducing labor.12 Because condom use was rare in the study population, the effect of paternal ejaculation could not be distinguished from the effect of intercourse itself. Assuming that penile-vaginal intercourse without a condom always involves the introduction of seminal fluid into the vagina, there are still several variables that may impact the effect on the cervix, including volume of ejaculate, concentration of prostaglandin within the ejaculate, or proximity of ejaculate to the cervical os. Likewise, different aspects of the maternal sexual response may impact the results, including degree of arousal and engorgement, presence of orgasm, and possibility of multiple orgasm. The present study did not elicit this information from its subjects; as such, it cannot address what effect the components of each sexual encounter might have on the onset of labor. Clearly, further research is needed to determine what role if any is played by each of these components.
By enrolling in this study, participants may have been subject to a suggestion bias, whereby they may have been inspired to try sexual activity to initiate labor when they might not otherwise have done so if the subject had not been discussed. Those subjects who had intercourse more frequently might then have wanted to shorten the duration of pregnancy for a multitude of other reasons. This may also explain why more elective inductions were performed for those patients who were sexually active because these inductions were generally initiated at the patient's request.
Beyond these inherent biases, the study has other shortcomings. Placing all patients with any sexual activity into a single group may not reflect differences between those with a single act of intercourse and those who regularly engage in intercourse. A larger study would be able to stratify this characteristic. Because the author was not usually present when the patient arrived at the hospital in labor, data were not collected regarding the presence or frequency of sexual activity in the period of time immediately before labor. This information might be more meaningful than data elicited the week before. Because each cervical examination was compared only with the sexual activity in the prior week, this measurement was still felt to provide meaningful data.
Regardless of the motivation to engage in sexual activity, this study has identified a larger proportion of patients at term having sex than in previous studies. Over half of the patients in this population reported sexual activity beyond 37 weeks, despite the large percentage of unmarried individuals. Obstetric caregivers should be aware of this behavior to better counsel their patients. The lack of condom use in this population might be understandable given the absence of a need for contraception, but the role of condoms in preventing sexually transmitted infections should be emphasized in high-risk populations.
Caregivers should take the initiative in talking with their patients about sexuality in pregnancy because it is clearly a concern of a significant proportion of their patients. From the data presented, this discussion should not lead patients to believe that sexual intercourse will initiate labor sooner. Although there does not appear to be any significant morbidity associated with sex at term in patients at low risk for complications, there are no data to support recommending sexual activity. Patients may continue to hear from relatives and other “old wives” that intercourse will hasten labor, but it should not be given credence by the medical community.
1. Schaffir J. Survey of folk beliefs about induction of labor. Birth 2002;29:47–51.
2. Toth M, Rehnstrom J, Fuchs A. Prostaglandins E and F in cervical mucus of pregnant women. Am J Perinatol 1989;6:142–4.
3. Chayen B, Tejani N, Verma UL, Gordon G. Fetal heart rate changes and uterine activity during coitus. Acta Obstet Gynecol Scand 1986;65:853–5.
4. Goodlin RC, Schmidt W, Creevy DC. Uterine tension and fetal heart rate during maternal orgasm. Obstet Gynecol 1972;39:125–7.
5. Von Sydow K. Sexuality during pregnancy and after childbirth: a metacontent analysis of 59 studies. J Psychosom Res 1999;47:27–49.
6. Ekwo EE, Gosselink CA, Woolson R, Moawad A, Long CR. Coitus late in pregnancy: risk of preterm rupture of amniotic sac membranes. Am J Obstet Gynecol 1993;168:22–31.
7. Bishop EH. Pelvic scoring for elective induction. Obstet Gynecol 1964;24:266–8.
8. Klebanoff MA, Nugent RP, Rhoads GG. Coitus during pregnancy: is it safe? Lancet 1984;2:914–7.
9. Diggle PJ, Heagerty PJ, Liang KY, Zeger SL. Analysis of longitudinal data. 2nd ed. New York (NY): Oxford University Press; 2002.
10. Sayle AE, Savitz DA, Thorp JM, Hertz-Picciotto I, Wilcox AJ. Sexual activity during late pregnancy and risk of preterm delivery. Obstet Gynecol 2001;97:283–9.
11. Bendvold E. Coitus and induction of labor [in Norwegian]. Tidsskrift for Jordmodre 1990;96:6–8.
12. Kavanagh J, Kelly AJ, Thomas J. Breast stimulation for cervical ripening and induction of labor (Cochrane Review). In: The Cochrane Library, Issue 3, 2005. Oxford: Update Software.