To investigate the timing of resumption of vaginal sex and assess associations with method of birth, perineal trauma and other obstetric and social factors.
To investigate the timing of resumption of vaginal sex and assess associations with method of birth, perineal trauma and other obstetric and social factors.
Prospective pregnancy cohort study of nulliparous women.
A total of 1507 nulliparous women recruited in early pregnancy (≤24 weeks).
Women were recruited from six public hospitals. Data from hospital records and self-administered questionnaires at recruitment and 3, 6 and 12 months postpartum were analysed using univariable and multivariable logistic regression.
Resumption of vaginal sex.
Sexual activity was resumed earlier than vaginal sex, with 53% resuming sexual activity by 6 weeks postpartum, and 41% attempting vaginal sex. By 8 weeks a majority of women had attempted vaginal sex (65%), increasing to 78% by 12 weeks, and 94% by 6 months. Compared with women who had a spontaneous vaginal birth with an intact perineum, women who had a spontaneous vaginal birth with an episiotomy (adjusted odds ratio 3.43, 95% confidence interval 1.9–6.2) or sutured perineal tear (adjusted odds ratio 3.18, 95% confidence interval 2.1–4.9) were more likely not to have resumed vaginal sex by 6 weeks postpartum. Similarly, women who had an assisted vaginal birth or caesarean section had raised odds of delaying resumption of sex.
Most women having a first birth do not resume vaginal sex until later than 6 weeks postpartum. Women who have an operative vaginal birth, caesarean section or perineal tear or episiotomy appear to delay longer.
The resumption of sex after childbirth is a concern for many women and their partners, and a recommended discussion topic in postnatal care.[1-3] Yet, obstetric and midwifery clinical textbooks generally contain minimal information about the impact of pregnancy and childbirth on resumption of sex. The third edition of a classic obstetric text book published in 1997 suggests that coitus is often resumed within 2–3 weeks with an intact perineum, and at 4–8 weeks or later with an episiotomy or perineal repair. Myles' Textbook for Midwives states that women ‘should be’ pain free and have been able to resume intercourse without pain by 6 weeks postpartum, but acknowledges that this is sometimes not the case. The author of a recent compendium of obstetric evidence-based guidelines concludes that there is insufficient evidence regarding the resumption of sex after childbirth to inform the advice that health professionals give to women about this issue. This is largely because of the limitations of existing studies, most of which are cross-sectional, and therefore subject to recall bias. Few studies distinguish between nulliparous and multiparous women, and most lack sufficient sample size to assess associations with method of birth and degree of perineal trauma.
This paper draws on data collected in the Maternal Health Study, a large multicentre prospective nulliparous cohort study that was designed to assess the natural history of maternal morbidity during pregnancy and after a first birth. Primary outcomes included urinary incontinence, faecal incontinence, persisting perineal pain and other sexual health problems. Data were collected via self-administered questionnaires in early pregnancy, and at 3, 6 and 12 months postpartum, and included questions regarding resumption of sexual activity. The objectives of this paper are: (i) to investigate the timing of resumption of vaginal sex in this prospective nulliparous cohort, and (ii) to assess associations between resumption of vaginal sex and social and obstetric characteristics of study participants, in particular—method of birth and degree of perineal trauma.
Women were recruited to the study between 1 April 2003 and 31 December 2005 from six metropolitan public hospitals in Melbourne, Australia with a mix of high-risk and low-risk perinatal services. Eligibility criteria were: ≥18 years, nulliparous (i.e. no previous live births or pregnancies ending in a stillbirth), sufficient English language fluency to complete self-administered questionnaires and telephone interviews, and estimated gestation of ≤24 weeks at enrolment (according to ultrasound or date of last menstrual period). Following their booking visit, all eligible women were mailed an invitation package by participating hospitals. Study staff also made regular visits to antenatal booking clinics at two participating hospitals, and to childbirth education classes at one of the other study sites, to distribute information packages to women eligible to take part in the study. The invitation package included the baseline questionnaire, an information sheet about the study, a consent form and a separate sheet for providing contact details. The invitation was followed up by a single mailed reminder postcard. We were prevented from following up non-responders by telephone by Australian privacy legislation and the conditions of our ethics approval. Women were invited to return the consent papers, contact information and completed questionnaire in a reply-paid envelope. This paper draws on data collected in the baseline questionnaire and follow-up questionnaires completed at 3, 6 and 12 months postpartum, and on data abstracted from hospital medical records for women giving written consent.
Questionnaires were B5 booklets approximately 40 pages in length, which focused on the assessment of a range of common maternal physical and psychological health problems, including sexual health issues using previously validated standardised questions where possible. Questions regarding resumption of sexual activity, vaginal sex and postnatal sexual health problems drew on the study reported by Barrett et al., and were included in each postpartum questionnaire. Sexual activity was defined as any form of sexual contact which may or may not include vaginal sex.
Data on labour and birth events were abstracted from hospital medical records using a detailed data abstraction protocol, and were also collected in the first postnatal follow-up questionnaire at 3 months postpartum. There was a high level of congruity between medical record data and women's own accounts of method of birth and other obstetric events.[9, 10] Data from questionnaires have been used to categorise method of birth and degree of perineal trauma in the current paper as they provided more complete data for these variables. Medical record data gave more precise information about perineal trauma, and timing of caesarean section in first-stage or second-stage labour. Data on method of birth and degree of perineal trauma were combined to provide a single measure of perineal trauma stratified by method of birth.
Maternal depressive symptoms were assessed at 3 months postpartum using the Edinburgh Postnatal Depression Scale (EPDS), a 10-item scale that has good sensitivity and specificity for identifying probable clinical depression in community samples. Scores ≥13 on the EPDS are indicative of probable major depression. Twelve-month period prevalence of intimate partner abuse was assessed using the short version of the Composite Abuse Scale, a validated instrument comprising 18 items of actions by an intimate partner that constitute emotional or physical abuse.[12, 13]
Tiredness in the first 3 months postpartum was assessed using a symptom checklist that asked about the experience of common maternal physical health problems. Specifically, women were asked ‘Since the birth have you ever experienced “extreme tiredness or exhaustion”?’ Pre-given response categories were: ‘never’, ‘rarely, ‘occasionally’ or ‘often’. Responses were dichotomised into symptoms reported as ‘often’ or ‘occasionally’ versus ‘rarely or ‘never’.
Method of infant feeding was assessed based on responses to a series of questions about whether women had ever breastfed their baby, the type of feeds the baby was having at the time of completing the 3-month questionnaire, and for women who started to breastfeed but had stopped, how old their baby was when they stopped (number of completed weeks). No distinction was made between exclusive, predominant or complimentary breastfeeding. Babies having some formula feeds and some breast-milk feeds were categorised as breastfed for the purposes of analysis.
Data on sociodemographic characteristics such as maternal age, education and relationship status were collected in the baseline questionnaire and the questionnaire at 3 months postpartum.
Data were analysed using STATA version 12. The proportion (and cumulative proportion) of women resuming sexual activity and vaginal sex at given intervals after childbirth were calculated based on the proportions of women reporting resumption of sex divided by the total number of women with data available. Sociodemographic and other factors associated with nonresumption of vaginal sex including maternal depressive symptoms, intimate partner violence in the first 12 months postpartum, tiredness and method of infant feeding were investigated using logistic regression. Multivariable logistic regression was used to examine the association between method of birth and perineal trauma (exposures of main interest) and resumption of sex (primary outcome), taking into account potential confounders, including maternal age, relationship status and site of recruitment to account for possible cluster effects.
A total of 1537 women enrolled in the study. Thirty women were excluded after enrolment due to miscarriage (12), insufficient fluency in English (11), multiparity (5), or termination of pregnancy for fetal abnormality (2), leaving a final sample of 1507 women. We are unable to determine a precise response fraction. Over 6000 information packages were distributed, but some were mailed to women who had already received a questionnaire and information package at a booking visit. It is also likely that some were incorrectly addressed or sent to women who were ineligible (e.g. multiparous). Based on the assumption that 80–90% of invitations were sent to eligible women with a correct mailing address, we conservatively estimate the overall response fraction to be in the range of 28–31%.
The mean gestation of study participants at the time of enrolment was 15.0 weeks (SD 3.1, range 6–24 weeks). Social and reproductive characteristics of participants were compared with routinely collected data for women giving birth at Victorian public hospitals in the study period. Compared with all women ≥18 years giving birth to their first child at public hospitals in Victoria during the 30-month period of recruitment, study participants were representative in relation to method of birth and infant birthweight, but under-represented young women (18–24 years, 15.5% versus 29.9%) and women born overseas from a non-English-speaking background (16.2% versus 21.0%). Further information regarding sociodemographic and reproductive characteristics of the sample and representativeness of study participants compared with other women giving birth at study hospitals and other Victorian public hospitals in the same time period is available in a previous paper.
Table 1 shows the number, proportion and cumulative proportion of women to resume sexual activity and vaginal sex in the first 12 months postpartum. Sexual activity was resumed earlier than vaginal sex, with 53% resuming some form of sexual activity by 6 weeks postpartum, and 41% attempting vaginal sex. By 8 weeks a majority of women had attempted vaginal sex (65%), with this figure increasing to 78% by 12 weeks postpartum, and 94% by 6 months postpartum.
|Timing||Sexual activity||Vaginal sex|
|n (%)||n (%)|
|≤4 weeks||365 (28.0) [28.0]||199 (15.2) [15.2]|
|5–6 weeks||327 (25.0) [53.0]||342 (26.2) [41.4]|
|7–8 weeks||266 (20.4) [73.4]||310 (23.8) [65.2]|
|9–12 weeks||170 (13.0) [86.4]||169 (13.0) [78.2]|
|4 months (13–17 weeks)||74 (5.7) [92.1]||156 (12.0) [90.2]|
|5 months (18–21 weeks)||21 (1.6) [93.7]||36 (2.8) [93.0]|
|6 months (22–26 weeks)||23 (1.8) [95.5]||11 (0.8) [93.8]|
|7–9 months (27–39 weeks)||22 (1.7) [97.2]||27 (2.0) [95.8]|
|10–12 months (40–52 weeks)||8 (0.6) [97.8]||11 (0.8) [96.6]|
|Not resumed by 12 months||29 (2.2) [100.0]||44 (3.4) [100.0]|
Table 2 shows analyses assessing the association between non-resumption of vaginal sex at two time-points (6 weeks and 3 months postpartum) and maternal sociodemographic, obstetric and postnatal characteristics, including scores on the EPDS and the Composite Abuse Scale. The odds ratios shown in the Table are based on the odds of not resuming sex at each time-point. Compared with women aged 30–34 years, younger women (18–24 years) were significantly less likely not to have resumed vaginal sex by 6 weeks postpartum. Women who had a birth assisted with forceps and those who had an episiotomy or sutured tear were more likely not to have resumed vaginal sex compared with women who had a spontaneous vaginal birth and intact perineum, respectively. Breastfeeding and extreme tiredness since the birth were also associated with not resuming vaginal sex by 6 weeks postpartum, as was being single, separated or divorced.
|Total*||Resumed vaginal sex by 6 weeks postpartum||Resumed vaginal sex by 3 months postpartum|
|n||%||OR||95% CI||n||%||OR||95% CI|
|Maternal age at birth (years)|
|30–34||376||150||39.9||1.0 (ref)||305||81.1||1.0 (ref)|
|Married/de facto||1248||525||42.1||1.0 (ref)||991||79.4||1.0 (ref)|
|University degree||622||238||38.3||1.0 (ref)||471||75.7||1.0 (ref)|
|Method of birth|
|Spontaneous vaginal||634||263||41.5||1.0 (ref)||496||78.2||1.0 (ref)|
|Caesarean, no labour||128||61||47.7||0.78||0.5–1.1||96||75.0||1.20||0.8–1.9|
|Intact perineum||539||270||50.1||1.0 (ref)||437||81.1||1.0 (ref)|
|Any breastfeeding at 6 weeks|
|Any breastfeeding at 3 months|
|Extreme tiredness since birth|
|No||432||198||45.8||1.0 (ref)||354||81.9||1.0 (ref)|
|EPDS at 3 months|
|Any type of abuse birth – 12 m postpartum|
|No||1082||451||41.7||1.0 (ref)||843||77.9||1.0 (ref)|
At 3 months postpartum, women were more likely not to have resumed vaginal sex if they were older (≥35 years); single, separated or divorced; had had an episiotomy, or had experienced extreme tiredness since the birth. Compared with women who had degree-level qualifications, women who had not completed year 12 were significantly less likely not to have resumed vaginal sex. There was a weak association between resumption of vaginal sex and depressive symptoms (EPDS ≥13), with data showing a moderate effect bordering on statistical significance. There was no association between exposure to intimate partner abuse as assessed by the Composite Abuse Scale and timing of resumption of vaginal sex.
To investigate the contribution of method of birth and perineal trauma to timing of resumption of vaginal sex, we combined data from these two variables to create a composite variable. Table 3 shows the proportions of women who had resumed vaginal sex by 6 weeks and 3, 6 and 12 months postpartum by degree of perineal trauma (intact perineum/unsutured tear/sutured tear/episiotomy) within strata for method of birth. The pattern of association between degree of perineal trauma and non-resumption of vaginal sex at 6 weeks and 3 months postpartum is similar across unassisted and assisted vaginal delivery methods, with women who had an intact perineum more likely to have resumed sex than women who sustained a perineal tear or episiotomy.
|Method of birth and perineal trauma||Total n||Resumed vaginal sex by (%):|
|6 weeks||3 months||6 months||12 months|
|Spontaneous vaginal birth|
|Vaginal birth forceps|
|Vaginal birth vacuum extraction|
|No labour intact perineum||128||47.7||75.0||88.3||89.1|
|Laboured intact perineum||256||44.1||81.3||94.5||96.9|
|Laboured sutured tear||2||0.0||100.0||100.0||100.0|
To obtain a more precise estimate of the association between method of birth, perineal trauma and resumption of sex, we developed a multivariable logistic regression model with the composite variable for method of birth and degree of perineal trauma as the exposure of main interest and resumption of sex at 6 weeks postpartum as the outcome variable. The small subgroups of women who sustained perineal trauma before caesarean section or who gave birth assisted by forceps or vacuum extraction without sustaining perineal trauma were excluded from the model. Other variables included in the model (shown in Table 4) were maternal age (included for a priori reasons), relationship status, infant feeding method and extreme tiredness in the first 3 months after the birth (all of which were associated with timing of resumption of vaginal sex in univariable analyses). The site of recruitment was included in the model to account for potential cluster effects. Table 4 shows the adjusted odds ratios within strata compared with the reference category of spontaneous vaginal birth with an intact perineum. The results confirm effects associated with degree of perineal trauma within strata of vaginal delivery types. Women who had a spontaneous vaginal birth or birth assisted by forceps or vacuum extraction and sustained a sutured tear or episiotomy were more likely not to resume vaginal sex by 6 weeks postpartum compared with women who experienced a spontaneous vaginal birth with an intact perineum. The small group of women who had an unsutured tear were slightly more likely not to resume vaginal sex by 6 weeks, but the effect was modest and of borderline statistical significance. Women who gave birth by caesarean section were included in the model even though the numbers of women who sustained perineal trauma were too small for inclusion in the modelling analyses. Compared with the reference category of women who had a spontaneous vaginal birth and intact perineum, women who had a caesarean section had raised odds of not resuming vaginal sex by 6 weeks postpartum, irrespective of the timing of caesarean section (before or after commencing labour) that were not explained by maternal age and other variables included in the model.
|Method of birth and perineal trauma||Total||Resumed vaginal sex by 6 weeks postpartum|
|n||%||Adjusted ORa||95% CI|
|Spontaneous vaginal birth|
|Intact perineum||142||86||60.6||1.0 (ref)|
|Vaginal birth with forceps|
|Vaginal birth with vacuum extraction|
|No labour intact perineum||128||61||47.7||1.71||1.0–2.9|
|Laboured intact perineum||256||113||44.1||1.99||1.3–3.1|
|Maternal age (years)|
|University degree||622||238||38.3||1.0 (ref)|
|Any breast feeding at 6 weeks|
|Extreme tiredness since birth|
The most important finding of this study is the wide time interval over which women resume vaginal sex after a first birth. The common assumption that most women will resume sex by 6 weeks postpartum is out of step with the choices made by many women and their partners. Many factors will influence these decisions. It is possible that some couples delay resumption of sex until after the ‘6-week check-up’ on the grounds of waiting to check that everything is ‘back to normal’. This may explain the relatively large number of women who resumed vaginal sex around 7–8 weeks postpartum. The study findings provide evidence that both method of birth and degree of perineal trauma also play a role, with caesarean birth or operative vaginal birth more likely to be associated with a delay in resumption of vaginal sex compared with vaginal birth with an intact perineum. In modern obstetric practice, only a small proportion of women having a first baby will achieve a vaginal birth with no attendant perineal trauma. In this large multicentre study, the proportion was around 10%. Hence, for the vast majority of women and their partners, it is reasonable to anticipate a delay in resuming vaginal sex related to the events of labour and birth.
Young women (<25 years) tended to resume vaginal sex earlier than older women. Women who were single, divorced or separated were more likely to resume sex later than women living with a partner. Apart from maternal age and relationship status, sociodemographic characteristics appeared to play little role in women's decisions about timing of resumption of vaginal sex. Other factors associated at a univariable level with a delay in resumption of vaginal sex were breastfeeding and extreme tiredness. Exhaustion is a common experience in the first year after childbirth,[16-19] and was frequently commented on by study participants as a factor influencing their sex lives in the year after the birth (data not shown). Radestadt et al. have previously reported an association between breastfeeding and delayed resumption of sex citing hormonal changes and lack of vaginal lubrication as potentially contributing to this finding. It is possible that tiredness contributes to the association with breastfeeding, but our results showing an association between breastfeeding and timing of resumption of sex, adjusting for a range of other factors including tiredness, suggest that lactation may make an independent contribution to the timing of resumption of vaginal sex after childbirth.
Women who scored ≥13 on the EPDS at 3 months postpartum did not appear to be any more likely to delay resumption of vaginal sex than women scoring below this cut-off, although the results bordering on statistical significance and modest effect size would suggest a weak association. Other studies have shown associations between maternal depression and sexual health problems and delay in resumption of sex.
Strengths include: prospective data collection with intensive follow up to 12 months postpartum, limited attrition, and very few missing data for primary outcomes and exposures of main interest. Ascertainment of method of birth and degree of perineal trauma drew on data from medical records and on women's own accounts of labour and birth events. As with all studies there are also limitations. Ascertainment of timing of resumption was based on combining data from questionnaires at 3, 6 and 12 months postpartum, which may have introduced some recall bias, although our study involves more frequent assessment than most other studies. Women were recruited from hospitals with a mix of women at high and low risk of obstetric complications, but the sample cannot be considered to be truly population-based. We were able to quantify the extent of selection bias by comparing study participants with women who gave birth in the study period drawing on routinely collected data. These showed that the sample was representative in relation to method of birth and infant birthweight, but that women who are socially disadvantaged, for example younger women, and women born overseas of non-English speaking backgrounds, are under-represented. While this can be expected to influence prevalence estimates, and may for example, have biased estimates for resumption of sex towards a greater delay in resumption than may be true for the total population, recent studies examining exposure–outcome associations in studies with low response fractions suggest that our findings showing associations between obstetric events and delay in resumption of vaginal sex should be considered robust. Assessment of infant feeding method and duration of breastfeeding via questionnaire at 3 months postpartum may have resulted in some misclassification. Similarly, we did not have data on intimate partner violence that was specific to the first 3 months. It is possible that our results may mask effects that were present during this period. Other studies have shown that women experiencing intimate partner violence may suffer major health consequences as a result of being forced to have sex.
The timing of resumption of vaginal sex is only one dimension of sexuality after childbirth, but it is very common for women to want information about when sexual activity may be safely and comfortably resumed, and what to expect in relation to the impact of childbirth on their sexual relationship. Having more reliable information to guide clinical practice is important for a number of reasons. First and foremost, it can help dispel myths about what is normal (or abnormal) in the postnatal period. The fact that most women had not resumed vaginal sex by 6 weeks postpartum means that it is normal for couples to delay resumption of sex until after this time. This is useful information for couples to know before their baby is born, and may help to reduce the feelings of anxiety and guilt about not resuming sexual activity sooner. Second, having more reliable evidence enables clinicians to tailor information to women's circumstances. It may help women who have had a caesarean section, operative vaginal birth and/or an episiotomy or perineal tear to be counselled to anticipate ways in which this might affect their sexual relationship.
The Maternal Health Study is the first large multicentre, prospective pregnancy cohort study to provide robust evidence regarding the association of method of birth, perineal trauma and the timing of resumption of vaginal sex after a first birth. The findings show that most women do not resume vaginal sex until later than 6 weeks postpartum, and that women having an operative vaginal birth, caesarean section and/or perineal tear or episiotomy are likely to delay resumption of vaginal sex for longer. The study provides important new evidence to guide information given to women and their partners about what to expect after childbirth.
EM conducted the analyses and wrote the paper. SB wrote the study protocol, took primary responsibility for the design and conduct of the study, contributed to analysis and interpretation of data and contributed to writing the paper. Both authors have approved the final draft of the paper for publication.
This study was approved by the following human research ethics committees: La Trobe University (2002/38); Royal Womens Hospital, Melbourne (2002/23); Southern Health, Melbourne (2002-099B); Angliss Hospital, Melbourne (2002), Royal Childrens Hospital, Melbourne (27056A).
This research was supported by project grants from the Australian National Health and Medical Research Council (ID191222 and ID433006 Melbourne, Australia); a Vic-Health Public Health Research Fellowship (2002–06) and a National Health and Medical Research Council Career Development Fellowship (ID491205, 2008–11) awarded to SB; and the Victorian Government's Operational Infrastructure Support Programme.
We are grateful to members of the Maternal Health Study research team who have contributed to data collection and coding (Maggie Flood, Deirdre Gartland, Ann Krastev, Renee Paxton, Susan Perlen, Martine Spaull, Hannah Woolhouse), and to Deirdre Gartland and Hannah Woolhouse for reviewing and commenting on the manuscript before submission.