Introduction

Obstetric anal sphincter injury (OASI) is a serious complication of childbirth with potential long-term maternal morbidity, such as anal incontinence and reduced quality of life. Perineal pain and dyspareunia appear to be related to the extent of perineal trauma [1–4]. However, very few studies have addressed sexual activity after OASI in a long-term prospective time span of more than 6 months [5–8]. In general, some studies assessing perineal pain and dyspareunia in relation to different degrees of perineal trauma (but not to OASI specifically) have methodological limitations, such as mixing heterogeneous degrees of perineal lacerations [5, 9–13] or failing to specify the degree of perineal trauma [10, 14, 15]. Systematic reviews are difficult to perform because of the heterogeneous study design, heterogeneous evaluation of exposure, and/or outcome measure and assessment at different postpartum intervals. Consequently, advising women who have suffered an obstetric anal sphincter injury on potential future sexual functioning after delivery is equally difficult.

The aim of the study was to investigate self-reported sexual activity and problems with intercourse 1 year after delivery in relation to perineal trauma, including OASI, and with regard to mode of delivery.

Materials and methods

This study is part of the Perineum Study, which was approved by the Regional Committee for Medical Research Ethics in South-Eastern Norway (ref. S-08810d/20941).

Study population

The Perineum Study population initially consisted of 2,846 pregnant women recruited prospectively during routine ultrasound examination in the second trimester at Oslo University Hospital, Ullevål, from September 2009 to August 2010, as shown in the flowchart (Fig. 1). This routine ultrasound examination is offered to all pregnant women in Norway in gestational weeks 18–20, and 98 % attend. Participants answered a questionnaire in Norwegian (Q1, Fig. 1) concerning urinary and anal incontinence, general state of health, and worries related to pregnancy and delivery. Demographic data, obstetric history, educational level, household income, and country of origin were also collected. Results have been published previously [16]. All participants gave written informed consent, also agreeing to receive further questionnaires after the index delivery.

Fig. 1
figure 1

Flowchart demonstrating the selection of the present study cases (women with obstetric anal sphincter injury, OASI) and controls

Full size image

Of these 2,846 participating pregnant women, 42 subsequently delivered with an obstetric anal sphincter injury (OASI, defined as perineal injury degree 3 or 4; however, all OASI cases in our study were third-degree tears). The OASI incidence rate of 1.5 % in our study population is representative of the current low OASI incidence rate at Oslo University Hospital, Ullevål. These 42 OASI women, together with 840 randomly selected controls (20 controls per case, selected from the initial study population of 2,846), who delivered without OASI, were sent a structured questionnaire by postal mail 1 year after delivery. The questionnaire (Q2, Fig. 1) addressed time of resumption of coitus after delivery, questions on whether there were current problems with intercourse after delivery and specification of such problems. Specifications were prelisted as:

  1. 1.

    Pain at the vaginal orifice during penetration

  2. 2.

    Pain during deep penetration

  3. 3.

    Feeling of having too wide a vaginal introitus

  4. 4.

    Feeling of having too tight or sore a vaginal introitus

  5. 5.

    Coital garulitas

  6. 6.

    Anal incontinence

  7. 7.

    Urinary incontinence

  8. 8.

    Fear of incontinence of any kind

  9. 9.

    Lack of sexual desire

  10. 10.

    Self-reported written specification of any other type of problem

The questionnaire also addressed birth control use and breastfeeding in addition to potential worries concerning socioeconomic status or regarding motherhood, family or partner relationships. Questions concerning these worries were chosen from the validated Cambridge Worry Scale [17], which was modified to address postpartum women who were not pregnant. The Q2 questionnaires were merged with data from the Oslo University Hospital local obstetrical database and with the first questionnaire during pregnancy (Q1).

Definition of perineal trauma

Perineal injury was classified following delivery and registered in the obstetrical database as degree 1 to 4, according to international guidelines [18]. The Oslo University Hospital’s local obstetrical database does not document the type of episiotomy performed. However, local guidelines at Oslo University hospital, Ullevål, recommend the lateral episiotomy technique when episiotomy is clinically indicated. The lateral technique is defined as an incision commencing ≥10 mm from the posterior fourchette and directed toward the ischial tuberosity. The midline episiotomy technique is discouraged in this delivery unit. Mediolateral incisions are also commonly performed, second to the lateral episiotomies [18].

Statistical analysis

Chi-squared test was used in the univariate analysis, and a significance level of 5 % was chosen. Univariate analysis was performed to identify the significant risk factors associated with resumption of sexual activity and with dyspareunia, and variables with p < 0.05 were included in the multivariate analyses. Multivariate regression analysis was used to explore the adjusted OR for postponed resumption of coitus (defined by us as more than 8 weeks postpartum) and self-reported dyspareunia, with 95 % confidence interval. The time period 8 weeks was chosen as a cut-off as all women in Norway are encouraged to have a postpartum check-up approximately 6–7 weeks after delivery. We hypothesized that many women, therefore, might not resume coitus before such a clinical examination. Model 1 depicts the maternal characteristics and delivery method solely. In model 2, delivery characteristics, such as duration of the second stage of labor, epidural analgesia, persistent occiput posterior presentation, episiotomy, and OASI were added. In model 3 we analyzed all degrees of perineal injury and episiotomy separately, and intact perineum was used as reference. PASW (Predictive Analytics SoftWare, SPSS, Chicago, IL, USA) version 22 was used to perform the analyses.

Results

Participants

Of the 882 women (42 cases and 840 controls) recruited to this study, 64 % (n = 561) responded to and returned the questionnaire (Q2). Response rates were similar for the OASI cases (67 %) and the controls (64 %). Table 1 summarizes the clinical characteristics of respondents and nonrespondents. Responders and nonresponders did not differ in age, parity, degree of perineal trauma or mode of delivery (Table 1). The majority of responders were primiparous (62 %), whereas 30 % of responders had given birth to their second child, and the remainders (8 %) were para 3 at the index delivery (Table 1).

Table 1 Clinical characteristics of invited study participants, n = 877 (5 out of 882 women were excluded because they delivered at a different hospital) displayed for responders and nonresponders
Full size table

Responder’s birth outcomes

Most responders (69 %) had delivered spontaneously, 12 % had been delivered by vacuum extraction, 6 % had had a planned cesarean, and 13 % had had an emergency cesarean section (Table 1). Seventeen women had had a preterm delivery before 37 weeks, and 7 women delivered twins. Among the 561 responding women, 34 % (n = 194) delivered with an intact perineum and 21 % (n = 120) underwent an episiotomy only, whereas the remaining group had some degree of perineal tear: 32 % (n = 177) sustained a first-degree perineal tear, 8 % (n = 42) a second-degree tear, whereas 5 % (n = 28) sustained a third-degree tear. Of the 28 OASI cases, 9 were peroperatively classified 3A, 19 were classified 3B, and none was 3C or 4 (data not shown).

Resumption of sexual activity postpartum

By 8 weeks, about half of the responders (51.4 %) had resumed coitus, and by 12 weeks 75.2 % had resumed intercourse, whereas 94.7 % had resumed coitus 1 year postpartum. Six women (1.1 %) reported attempted coitus, but had interrupted their attempt, and 23 women (4 %) had not yet attempted coitus. Two respondents did not answer this particular question and 1 woman had listed “other” as her answer.

Table 2 shows resumption of coitus categorized by 8 weeks, by 9–12 weeks, and after 3 months in relation to degree of perineal trauma and method of birth, showing a significant difference related to degree of perineal trauma. Women with OASI were significantly more likely to postpone coitus till after 3 months compared with any of the groups of women with a less severe degree of perineal trauma or episiotomy (p < 0.05). When comparing second-degree tears with episiotomy per se, we found no significant difference in percentage distribution of coital resumption between these two groups at the three time points studied (data not shown, p = 0.84). Women with a lower educational level were significantly more likely to postpone coital resumption till after 3 months postpartum compared with women with college and university education (Table 2, p < 0.05). We found no significant difference in percentage distribution of coital resumption when comparing different delivery modes, nor was there a significant difference in percentage distribution in relation to parity or breastfeeding (Table 2).

Table 2 Resumption of coitus in relation to the method of birth, perineal trauma, and (answered by 502 out of 561 study responders)
Full size table

Univariate logistic regression analysis is presented in Table 3 and shows that OASI increased the risk of postponed resumption of coitus (defined as after 8 weeks) by 7-fold (OR = 6.97, Table 3) compared with intact perineum.

Table 3 Crude and adjusted OR for postponed sexual resumption (defined as after 8 weeks, n = 244) after the index delivery
Full size table

Multivariate regression analyses for all variables significantly associated with postponed resumption of coitus in the univariate regression analysis are also presented in Table 3. Parity was included, despite not representing a significant risk for postponed sexual intercourse in the crude analyses, as we considered parity a potentially important confounding factor in our study of events associated with labor and delivery. In model 1 parity and maternal age were associated with postponed sexual resumption after delivery, maternal age only slightly associated, with an aOR 1.09.

In model 2 maternal age and OASI were the only significant risk predictors for postponed resumption of sexual intercourse (aOR 1.10 and 4.16). In contrast, episiotomy was no longer significantly associated with postponed resumption of coitus in these adjusted analyses, despite being a significant risk in the unadjusted analysis.

In model 3, as in model 2, maternal age remained statistically significant, but a minor risk factor for postponed sexual resumption (aOR 1.10), and OASI again remained the only strong and statistically significant predictor of postponed resumption of intercourse after delivery (aOR 5.52). Other degrees of perineal trauma were not significantly associated with postponed resumption of coitus (Table 3, model 3).

The models 2 and 3 were also performed after excluding all women who delivered by cesarean section, but this did not change the conclusions (data not shown). We additionally compared women with an episiotomy (n = 120) to women with a second-degree perineal laceration (n = 42) separately in a multivariate analysis, and there was no difference between these two groups in relation to postponed resumption of coitus (p = 0.42, data not shown).

In addition to the variables presented above, we analyzed a large number of maternal, fetal, and obstetric variables to assess whether other factors might contribute to postponed sexual resumption 8 weeks after delivery. Maternal body mass index (BMI), breastfeeding, and vaginal delivery route compared with cesarean were not associated with timing of sexual resumption, nor were obstetric interventions such as labor induction, epidural use or oxytocin augmentation. Newborn characteristics, such as a large baby (birth weight, head circumference), gestational age (prematurity), twins or low Apgar scores, were not associated with postponed resumption of sexual activity after birth (data not shown).

Women who reported that they had not resumed coitus by 12 months were asked to state the reason for nonresumption. Seven of these 23 women reported that they were not in a relationship. Ten women (44 %) reported the reason to be a lack of sexual desire (2 were OASI deliveries, 2 had delivered with an intact perineum, 3 with a first-degree tear, 1 with a second-degree tear, and 2 with an episiotomy only). Two women stated the reason for nonresumption to be that they were afraid of coital pain (1 of which had a spontaneous second-degree tear whereas the other had delivered with a first-degree perineal tear only). The remaining 4 women did not give a reason for nonresumption.

Difficulty with coitus/dyspareunia 1 year postpartum

The majority of participants (545 out of 561) stated that they were currently in a relationship 1 year postpartum, whereas 19 women stated that they were not, and 6 women did not answer this particular question. When asked whether currently experiencing difficulty with coitus after delivery, 530 out of 561 women responded, 164 of whom reported having current coital problems, whereas 366 women stated that they had none. At 1 year postpartum, significantly more women with OASI reported “having difficulties with coitus” compared with any other group of women with perineal trauma (48 vs 22 % for second-degree tears, 32 % for episiotomy, and 37 % for first-degree perineal tears, p < 0.02) and compared with women with intact perineum (23 %, p < 0.02).

Participants were asked to specify their difficulty, either by choosing from a predefined list and/or by giving their own written specification of the problem. None of the responders provided a written specification. The most frequently reported problem was lack of sexual desire (63.4 %), followed by dyspareunia, defined as either pain at the vaginal orifice during penetration (41.5 %) or pain during deep penetration (37.2 %). Forty-seven women (28.7 %) specified that the vaginal orifice felt too tight or sore during coitus, whereas 27 participants (16.5 %) considered the vaginal orifice to be too large. Very few reported having incontinence issues: only 5 women listed urinary incontinence as the reason for their coital difficulty, 12 women reported flatus incontinence, but none reported fecal incontinence. However, 17 women (10.4 %) stated that their problem was fear of incontinence of any type.

Table 4 shows dyspareunia types 1 year postpartum by degree of perineal trauma and mode of delivery. Women with OASI were significantly more likely to experience pain at the vaginal orifice during coitus compared with women without OASI (39 vs 6–19 %, p < 0.05, Table 4). We also found a borderline significance concerning pain at the vaginal orifice when comparing vacuum deliveries with spontaneous and cesarean deliveries, p = 0.07, but when comparing vaginal deliveries with cesareans per se, there was no significant difference between groups (p = 0.41, data not shown).

Table 4 Self-reported types of dyspareunia 1 year postpartum in relation to degree of perineal trauma and mode of delivery
Full size table

With regard to deep penetrational pain, there was no significant difference between different degrees of perineal trauma, although the group of women with vacuum deliveries had a significantly higher rate of such pain than other delivery groups (p < 0.05, Table 4).

Crude and adjusted logistic regression analyses of the 102 women reporting dyspareunia are presented in Table 5. The crude analysis showed that OASI significantly increased the risk of dyspareunia by a 4-fold compared with intact perineum (OR 4.00). Maternal BMI, household income, education level, delivery mode, and breastfeeding were not significantly associated with dyspareunia 1 year after birth, nor were fetal variables such as birthweight, head circumference, gestational age (prematurity), twins, occiput posterior presentation, shoulder dystocia or Apgar scores (data not shown).

Table 5 Crude and adjusted OR for self-reported dyspareunia 1 year after the index delivery (n = 102)
Full size table

In multivariate regression analysis of all factors associated with dyspareunia, worries concerning relations with friends and family were significantly associated with dyspareunia (Table 5, model 1). In model 2, OASI (all were third-degree perineal tears) remained the only significant and strong predictor for dyspareunia (aOR 2.98). In model 3 OASI again remained the only significant predictor for dyspareunia (aOR 3.57). None of the other degrees of perineal trauma were significantly associated with dyspareunia (Table 5, model 3), despite episiotomy being a significant risk in the non-adjusted analyses.

When separately comparing women with episiotomy to women with second-degree laceration in multivariate regression analysis, there was no significant difference between these two groups with regard to dyspareunia, p = 0.67 (data not shown).

Discussion

Our study showed that women with OASI had an increased risk of postponed resumption of coitus (defined as after 8 weeks) by a 4- to 5-fold in addition to being at a significantly increased risk of dyspareunia by a 3- to 4-fold at 1 year after delivery, assessed in two different multivariate analysis models (models 2 and 3; Tables 3, 5).

This is the first study to assess both time of resumption of coitus as well as dyspareunia 1 year postpartum in relation to OASI and mode of delivery adjusted for potential confounding factors in multivariate analyses. We demonstrated that OASI increased the risk of postponed resumption of coitus by a 5-fold compared with intact perineum, and that a significantly larger proportion of women with OASI postponed resumption of intercourse until after 3 months as compared to all other degrees of perineal trauma. A postponed resumption of coitus after OASI is in compliance with findings from previous studies from Sweden and the USA. Rådestad et al. found that adjusted relative risks for not having had sexual intercourse within 3 and 6 months were 2.1 and 2.2 for tears in the sphincter ani and rectum [8]. Similar to the study by Rådestad et al., our question concerning time to coital resumption was retrospective, with the possibility of a recollection bias. Had we sent questionnaires addressing the same issues at several occasions postpartum, the response rates would most likely have decreased throughout a 1-year follow-up period. Hence, we might have lost valuable 1-year postpartum data. We also recognize that time of resumption of coitus after delivery may not only be influenced by obstetric factors or the great variety of individual aspects, but also by cultural variations and attitudes toward female sexuality after childbirth. In Norway, all women with vaginal delivery are encouraged to undergo a postpartum gynecological examination approximately 6–7 weeks postpartum. Many Norwegian women may therefore not resume (or consider resuming) intercourse before having had such a check-up, which is supported by our finding that only half of our responders reported being sexually active 8 weeks after delivery. Consequently, we chose to dichotomize resumption of coitus to 8 weeks in our logistic regression analyses.

We did not find episiotomy to be a risk factor for postponed resumption of coitus (defined as after 8 weeks) in multivariate analysis, nor did we find a second-degree spontaneous perineal laceration to be associated with the postponed resumption of coitus. When categorizing coital resumption into three groups (by 8 weeks, by 3 months, and after 3 months) there were no significant differences between episiotomy and second-degree tears in percentage distribution of time to coital resumption. This is an interesting finding as results from previous studies generally differ with regard to coital resumption and dyspareunia comparing episiotomy with spontaneous perineal lacerations [2, 3, 8–10, 13, 19–21]. The differing results may be explained by studies grouping mixed degrees of perineal trauma into one single group [9–13, 19, 21, 22]. A strength of our study is the separate categorization of all degrees of perineal trauma in addition to separately studying episiotomy and spontaneous second-degree tears. Another advantage of our study in terms of securing accuracy is that the recorded delivery data on perineal trauma in the local obstetrical database was checked against individual medical charts (considered the gold standard) of all the 561 responding women by the first and last author.

A noteworthy finding is the quite large proportion of women with an intact perineum and a first-degree tear (19 % in both groups) reporting postponed resumption of coitus till after 3 months. A plausible explanation for this finding might be sexual habits and complaints prior to pregnancy and delivery, on which we have no information. We cannot exclude an inclusion bias, as women experiencing coital difficulty or women who have more worries about resumption of coitus after childbirth may be more prone to answer a sexual activity questionnaire than women who do not experience such problems. On the other hand, our finding of OASI being the only significant and major risk factor for postponed resumption is in concordance with existing literature.

Difficulty with coitus/dyspareunia the first year after delivery

Obstetric anal sphincter injury increased the risk of dyspareunia 1 year postpartum by a 3- to 4-fold in our multivariate regression analyses. In contrast, other degrees of perineal trauma, episiotomy included, were not independently associated with dyspareunia 1 year postpartum in our adjusted analyses.

Long-term follow-up studies of more than 6 months after childbirth in relation to dyspareunia or sexual functioning are rare, and even fewer address OASI in particular. Our finding of OASI being the biggest risk factor for dyspareunia 1 year after delivery is in slight contrast to an Australian study of 440 primigravidae. The authors found that at 12 months postpartum sexual function had returned to early pregnancy levels, irrespective of mode of delivery or degree of perineal injury [23]. The OASI group, in this study was very small (n = 9), and therefore the conclusions may have limited value. A weakness of our study is that we did not address sexual satisfaction, and we recognize that dyspareunia in itself is not a measure of female sexual dissatisfaction.

Ejegård et al. found that episiotomy, second-degree perineal lacerations, and a history of dyspareunia were independent risk factors for dyspareunia 12–18 months postpartum [19]. In our study neither episiotomy nor second-degree lacerations were risk factors for dyspareunia 12 months postpartum compared with intact perineum, which again is in compliance with the findings of De Souza et al. [23]. What strengthens our findings is that we have analyzed all obstetric factors associated with dyspareunia in a multivariate regression, which Ejegård et al. did not [19].

A limitation of our study is that we lack information on potential coital difficulties prior to pregnancy and delivery. However, we have no reason to believe that our OASI population would differ in terms of dyspareunia or other sexual problems prior to delivery as compared with our control group.

Van Brummen et al. showed that primiparous women were five times less likely to be sexually active 1 year postpartum after a third-/fourth-degree tear compared with women with an intact perineum [5]. However, the majority of our OASI population were in fact sexually active 1 year after delivery, but none had sustained a fourth-degree perineal injury, in accordance with the currently very low rate of the most severe form of OASI in our delivery unit. Whether or not our nonresponding OASI women were sexually active, is unknown to us, and again, we cannot exclude an inclusion bias, even though our responders and nonresponders did not differ with regard to age, degree of perineal trauma, mode of delivery or parity.

We did not use a validated questionnaire in our study, as our aim was not to assess detailed female sexual function, merely to address potential sexual problems, such as dyspareunia, 1 year postpartum. The Norwegian Institute of Public Health has conducted several epidemiological survey studies on health and sexual behavior in Norway using non-validated questionnaires, but used, similar to us, simple questions such as “Are you sexually active or not,” with the response options being “yes” and “no.” If the answer was “yes,” the degree of sexual problems was measured by the question “Have you experienced any of the sexual problems listed below during the past 12 months/or since sexual activity was resumed after birth” [24]. We therefore considered our simple questions about coital difficulties 1 year after delivery to be appropriate for assessing our outcome measure of interest.

Despite potentially redundant confounding factors, such as individual differences in sexual activity and other nonregistered variables of importance, our persistent and clear study finding is that OASI remains a large risk factor for postponed resumption of sexual intercourse after delivery and for dyspareunia 1 year postpartum. Our study is not a randomized controlled trial comparing episiotomy use and risk of OASI, but we found no support neither for episiotomy being a risk for postponed sexual intercourse nor for episiotomy being a risk factor for dyspareunia 1 year postpartum. Importantly, moreover, we found no difference in coital resumption when comparing episiotomy with second-degree lacerations separately. Even though episiotomy should only be used on indication, and not routinely, we advocate that episiotomy should be used when indicated to reduce the risk of OASI. Our previous studies have documented that a reduction in OASI rates is possible without a large increase in episiotomy rates [25, 26]. Our main finding of affected sexual activity after OASI strongly supports the need to reduce the rates of this obstetric injury to a minimum.