Women's sexual health after childbirth

Authors


Correspondence: Ms G. Barrett, Sexual Health Programme, Health Promotion Research Unit, Department of Public Health and Policy, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK.

Abstract

Objective To investigate the impact of childbirth on the sexual health of primiparous women and identify factors associated with dyspareunia.

Design Cross-sectional study using obstetric records, and postal survey six months after delivery.

Setting Department of Obstetrics and Gynaecology, St George's Hospital, London.

Population All primiparous women (n= 796) delivered of a live birth in a six month period.

Methods Quantitative analysis of obstetric and survey data.

Main outcome measures Self reported sexual behaviour and sexual problems (e.g. vaginal dryness, painful penetration, pain during sexual intercourse, pain on orgasm, vaginal tightness, vaginal looseness, bleeding/irritation after sex, and loss of sexual desire); consultation for postnatal sexual problems.

Results Of the 484 respondents (61% response rate), 89% had resumed sexual activity within six months of the birth. Sexual morbidity increased significantly after the birth: in the first three months after delivery 83% of women experienced sexual problems, declining to 64% at six months, although not reaching pre-pregnancy levels of 38%. Dyspareunia in the first three months after delivery was, after adjustment, significantly associated with vaginal deliveries (P= 0.01) and previous experience of dyspareunia (P= 0.03). At six months the association with type of delivery was not significant (P= 0.4); only experience of dyspareunia before pregnancy (P < 0.0001) and current breastfeeding were significant (P= 0.0006). Only 15% of women who had a postnatal sexual problem reported discussing it with a health professional.

Conclusions Sexual health problems were very common after childbirth, suggesting potentially high levels of unmet need.

INTRODUCTION

Women's mental health in the postnatal period has been researched extensively1–5. More recently, studies have demonstrated that women also experience physical problems, and urinary and faecal incontinence after childbirth6–19. However, women's sexual health after birth remains under-researched, in particular the experience of dyspareunia and other sexual problems.

The few reported studies of postnatal sexual health have had limitations, either with the sample or the extent to which sexual health was examined. For example, some studies have volunteer samples20–23 or only have included women who meet certain criteria, such as women with episiotomies24, vaginal deliveries25, women with adequate English, a stable relationship, and living in a certain area25–27, or met the entry criteria for a randomised controlled trial28–30. Two well conducted studies used large unselected samples to examine women's general postnatal health; these studies included questions on postnatal sexual health, but the questions were limited9,31.

Despite the limitations of available studies, general trends emerge: childbirth brings about a change in the sexual relationship; perineal pain and dyspareunia are common experiences for postnatal women22–32; and there is generally a decrease in the frequency of sexual intercourse26,27,33 and the woman's sexual desire20,24,27,28,34,35. Hormonal effects associated with breastfeeding appear to be associated with vaginal dryness and/or loss of libido21,22,31,34,36, although this finding is not consistent across all studies37,38.

There is evidence for a positive association between levels of dyspareunia/perineal pain and perineal damage (in particular, episiotomy)22,23,25,28 and assisted vaginal delivery23,25,31. Such evidence is relevant to the debate about women's choice of elective caesarean section as preferred mode of delivery39–42. Parity is also important because primiparous women have higher rates of epi-siotomy and assisted delivery than multiparous women23,41–45 and report higher levels of dyspareunia and perineal pain23,28.

There is a lack of information about consultation for postnatal sexual problems and information provided by health professionals on this subject. It appears that while the majority of women discuss contraception with a health professional, they rarely discuss sexual intercourse23,31, and even when they feel a need for help with a sexual problem, only a minority seek it31.

In this study we describe the nature of women's sexual health after childbirth and focus on the factors associated with dyspareunia.

METHODS

We studied a series of consecutive primiparous women giving birth to a live infant at a London teaching hospital between 1 July and 31 December 1997. Three women were excluded from the sample: one because she died some months after the birth; one because of the infant's subsequent death; and one because one of her twins was stillborn. The final sample included 796 women with surviving infants. All preterm births, twins (where both were alive), and infants with malformations were included. Primiparous women were chosen to avoid the confounding effects of a previous birth. A six month cohort of women was chosen because of administrative convenience and funding constraints (this study was funded in-house). We expected the six month period to provide approximately 750 primiparous women. In our pilot study the proportions of women reporting dyspareunia varied significantly prior to pregnancy (22%), in the first three months (58%), and at approximately six months after birth (26%). However, the sample size was insufficient to perform multifactorial analyses on dyspareunia. We estimated that the current study size would give at least 100 women with dyspareunia at six months, and so would be large enough do such analyses and thus explore factors related to dyspareunia.

Information about each woman (e.g. age, social circumstances, medical history, obstetric details) and her infant (e.g. birthweight, gestational age) was collected from the computerised birth records. All antenatal and peripartum data in the hospital are entered onto computer by trained midwives, and all midwives receive training on how to use the system and enter the data. There is also a midwife in overall charge of the data system who regularly validates the data entered. Overall, the information in the computerised birth records is detailed and has a high level of completeness.

Postal questionnaires were also sent to women six months after delivery. The questionnaires were sent out in one month batches in the sixth month after delivery (for example, July deliveries received their first posting in December, August deliveries in January etc). Nonresponders in each batch were sent two reminder questionnaires, at two week intervals. Most women in the final sample responded to either the first of second mailing; very little was gained by the third mailing. Overall, data collection lasted from December 1997 to early June 1998.

The questionnaire enquired about general health, bowel and bladder function, sexual health and mental health46, but only the findings relating to sexual health are reported in this paper. Our other findings will be reported elsewhere, as appropriate.

In the questionnaire women were asked if they had resumed sexual intercourse or had attempted to do so. All women who had resumed (or attempted) sexual intercourse were asked a detailed set of questions about problems experienced (prior to pregnancy and postnatally), sexual practices (using the definitions developed for the National Sexual Attitudes and Lifestyles Survey47), frequency of sexual intercourse, satisfaction with sex life, and consultation for postnatal sexual problems. With the exception of the questions on sexual practices, all other areas of postnatal health that we included in the questionnaire have been identified as areas of potential change by previous studies20–36. All women were asked about the information they received on postnatal health before the birth and any information, help or advice they received from health professionals after the birth. Women were asked to recall information (e.g. problems before pregnancy, problems in the first three months after delivery, whether they had gone for their six week check etc), but also included contemporaneous questions (e.g. the postnatal problems they were experiencing now). The questionnaire was piloted successfully with 158 primiparous women in 199648 and is available on request.

Women were classified as experiencing dyspareunia if they answered positively to questions about ‘painful penetration’ and/or ‘pain during sexual intercourse’ and/or ‘pain on orgasm’.

Differences between proportions were tested using χ2 tests. For paired proportions McNemar's test was used (an exact binomial test was used in instances where discordant pairs were < 10), and for three related proportions Cochran's Q test was used.

Logistic regression was used to investigate factors associated with dyspareunia in the postnatal period. Two models were fitted: one for dyspareunia reported in the first three months after delivery and one for dyspareunia at six months after delivery. Variables significantly related (P < 0.05) to dyspareunia in the unifactorial analyses were put into the models. Variables were entered simultaneously in each model. Results of the logistic regression analyses are presented as odds ratios and 95% confidence intervals. Analyses were carried out using SPSS for Windows and ethical approval for this study was given by the Local Research Ethics Committee.

RESULTS

Of the 796 women, 484 (61%) returned a questionnaire. Of the 312 nonresponders, at least six refused because the questionnaire was too personal (we know this from letters and telephone calls from the women), and at least 45 women (6% of total sample) never received the questionnaire as our mailings were returned by the Post Office. There were no differences between responders and nonresponders in any obstetric feature. However, responders and nonresponders were significantly different with respect to age, ethnicity, occupation, marital status, and birthplace; responders were more likely to be older, white, employed, married, and born in the UK (Table 1).

Table 1.  Characteristics of responders and nonresponders. Values are given as % (n).
 Responders (n= 484)Nonresponders (n= 312)P
  1. No significant differences were found on any of the following variables: type of delivery, perineal/genital damage, number of infants, birthweight of infant, preterm birth, induction of labour, artificial rupture of membranes, use of syntocinon, any type of pain relief, length of labour, previous miscarriages, terminations, fertility treatment prior to pregnancy, or smoking.

  2. *Includes divorced/separated.

  3. Employment status not known for eight responders and seven nonresponders.

  4. Place of birth not known for two responders and four nonresponders.

Age (years)  <0.0001; χ2= 30; df = 4
  15–199 (42)14 (44) 
  20–2414 (67)24 (75) 
  25–2931 (150)32 (100) 
  30–3432 (157)22 (69) 
≤= 3514 (68)8 (24) 
Marital status  0.0009; χ2= 11; df = 1
  Married61 (294)49 (152) 
*  Single*39 (190)51 (160) 
Ethnicity  <0.0001;χ2= 34; df = 2
  White71 (344)51 (159) 
  Black18 (86)33 (103) 
Asian11 (54)16 (50) 
Employment at time of booking appointment  < 0.0001; χ2= 43; df = 3
  Employed76 (363)54 (166) 
Unemployed11 (53)25 (76) 
Student5 (22)9 (28) 
  Housewife8 (38)11 (35) 
Birthplace of woman  < 0.0001; χ2= 16; df = 1
  UK72 (347)58 (180) 
Outside UK28 (135)42 (128) 

Resumption of sexual intercourse

Of the responders, 86% (415/480) had resumed sexual intercource since the birth of their child, and 2% (10/480) had attempted to resume, although they had not achieved penetrative sexual intercourse. Sexual intercourse had not been resumed by 55 women (11%) since the birth of their child. For 15 women, this was because they had no partner. The other 40 women had partners, but had not resumed intercourse. Reasons given by women included: tiredness/lack of energy (n= 14 women), physical problems, such as unhealed perineum (n= 9), loss of libido (n= 8), need for contraception (n= 3), feeling unattractive to partner (n= 2), and partner ill or working away (n= 2). Table 2 shows when women resumed sexual intercourse.

Table 2.  Resumption of sexual intercourse.
When sexual intercourse was resumed (or attempted, if not yet resumed)n (%) [cumulative%]
Weeks 1–347 (10) [lo]
Weeks 4–6107 (22) [32]
Weeks 7–8142 (30) [62]
Month 389 (19) [81]
Month 430 (6) [871
Month 58 (2) [891
Month 62 (<1) [89]
Not resumed 
  No partner15 (3) [92]
  Other reason40 (8) [100]
TOTAL480 (100) [100]

The sexual relationship

The nature of sexual practices changed after the birth, with the main change being a decline in oral sex by both the women and their partners (Table 3). Of the women who had resumed or attempted to resume sexual intercourse, 67% (282/420) reported that sexual intercourse was less frequent than before their pregnancy. Only 5% (21/420) said that sexual intercourse was more frequent. Assessment of quality of sex life was more variable with 38% (157/416) describing it as ‘less good’, 47% (194/416) saying it was ‘about the same’, 10% (42/416) saying it had improved, and 6% (23/416) saying they did not know.

Table 3.  Nature of sexual activity before and after pregnancy. Values are given as % (n).
Sexual activity*Sexual activities in year prior to pregnancySexual activities since the birthP
  1. *Question answered by 91% (388/425) of women who had resumed sexual activity.

  2. McNemar's test.

  3. Exact binomial test carried out because there were only nine discordant pairs.

Vaginal intercourse99 (386)98 (379)0.04
Oral sex by woman71 (277)58 (225)< 0.0001
Oral sex by partner72 (278)52 (203)< 0.0001
Anal sex6 (25)3 (10)0.002
Genital contact not involving intercourse71 (276)69 (269)0.2

Problems with sexual intercourse

Compared with the year before pregnancy, problems such as pain, lack of vaginal lubrication, and loss of sexual desire all increased significantly in the first three months after delivery; these problems declined by six months but not to pre-pregnancy levels (Table 4). Dyspareunia (as defined by experience of painful penetration and/or pain during sexual intercourse and/or pain on orgasm) was particularly common: 12% (48/403) had experienced dyspareunia in the year before pregnancy, compared with 62% (225/364) in the first three months after birth and 31% (123/403) at six months. Figure 1 shows the pattern of dyspareunia and other sexual problems in the postnatal period. In the year before pregnancy, 62% of women (251/403) reported not experiencing any sexual problems, as defined in Table 4. In the first three months after delivery, this figure had dropped to 17% (61/363), and by six months after delivery it had risen to 36% (146/403).

Table 4.  Problems experienced with sexual intercourse. Values are given as % (n).
Problem*Ever experienced in year prior to pregnancy (n= 403)Ever experienced in first 3 months after birth (n= see individual denominators)Experiencing this problem now (n= 403)P
  1. *Question answered by 95% (403/425) of women who had resumed sexual activity

  2. Of the 403 women who answered the question, not all answered the section ‘in first 3 months’. Individual denominators are given for each problem in the ‘in first 3 months’ column.

  3. Cochran's Q test; relates to women who gave information at each time point (i.e. denominator in f'in first three months' column).

Lack of vaginal lubrication12 (49)46 (169/368)26 (106)< 0.0001
Painful penetration9 (38)55 (200/364)27 (110)< 0.0001
Pain during sexual intercourse8 (31)45 (162/363)20 (80)< 0.0001
Pain on orgasm<1 (2)3 (11/364)2 (9)0.004
Difficulty reaching orgasm14 (57)33 (121/369)23 (91)< 0.0001
Vaginal tightness14 (58)33 (120/365)20 (80)< 0.0001
Vaginal looseness/lack of muscle tone3 (12)20 (72/365)12 (49)< 0.0001
Bleeding or irritation after sex4 (16)15 (54/363)6 (23)< 0.0001
Loss of sexual desire9 (36)53 (204/385)37 (150)< 0.0001
Figure 1.

Sexual problems. —♦— dyspareunia; —▪— lack of vaginal lubrication; —Δ— difficulty reaching orgasm; inline image vaginal tightness; —*— vaginal looseness; —•— bleeding/irritation after sex; —+— loss of sexual desire.

Factors associated with dyspareunia

In the unifactorial analysis, dyspareunia experienced during the first three months after birth was significantly associated with type of delivery, perineal damage, and having experienced dyspareunia before pregnancy (Table 5). In the multifactorial analysis, only type of delivery and history of pre-pregnancy dyspareunia remained significant (Table 6).

Table 5.  Factors associated with the experience of dyspareunia in the first three months after delivery.
Factor% (n)Unadjusted odds ratio 95% CIP
  1. No significant differences were found on any of the following variables: age, marital status, ethnicity, birth place of woman, induced labour, artificial rupture of membranes, any type of pain relief, syntocinon used during labour, length of labour, birthweight of infant, gestations, sex of infant, and breastfeeding around three months.

Type of delivery  < 0.0001
 Vaginal unassisted62 (108/174)1.00 
 Forceps/ventouse78 (78/100)2.17 (1.23–3.81) 
 Caesarean (+ labour)41 (23/56)0.43 (0.23479) 
 Caesarean (no labour)47 (16/34)0.54 (0.26–1.14) 
Perinedother genital damage  0.006
 Intact perineum49 (61/124)1.0 
 1 st degree tear68 (19/28)2.18 (0.92–5.19) 
 2nd degree tear61 (35/57)1.64 (0.87–3.11) 
 3rd degree tear64 (7/11)1.80 (0.50–6.49) 
 Episiotomy73 (87/119)2.81 (1,64–4.80) 
 Labial tears50 (8/16)1.03 (0.36–2.93) 
 Vaginal tears89 (8/9)8.24 (1.00–67.72) 
Experienced dyspareunia in year prior to pregnancy  0.03
 Yes78 (31/40)2.31 (1.06–5.01) 
 No60 (194/324)1.o 
Table 6.  Logistic regression analysis of factors associated with dyspareunia in the postnatal period. Values are given as adjusted OR (95% CI)*. NA = not applicable to model.
FactorExperiencing dyspareunia in first three months (n= 364)PExperiencing dyspareunia at six months (n= 400)P
  1. *Each odds ratio adjusted for other variables in model.

  2. These were the women for whom full data were available.

Experienced dyspareunia in year prior to pregnancy 0.03 < 0.0001
 Yes2.44 (1.09–5.46) 4.97 (2.57–9.60) 
 No1.0 1.0 
Type of delivery 0.01 0.4
 Vaginal unassisted1.00 1.0 
 Forceps/ventouse2.41 (1.24–4.69) 1.38 (0.74–2.57) 
 Caesarean (+ labour)0.42 (0.18–1.00) 1.91 (0.69–5.28) 
 Caesarean (no labour)0.52 (0.20–1.38) 1.19 (0.37–3.78) 
PerineaVother genital damage 0.6 0.7
 Intact perineum1.0 1.0 
 1st degree tear1.20 (0.42–3.46) 2.26 (0.68–7.53) 
 2nd degree tear0.83 (0.34–2.00) 2.17 (0.80–5.88) 
 3rd degree tear0.52 (0.12–2.36) 1.87 (0.38–9.12) 
 Episiotomy0.97 (0.41–2.31) 2.21 (0.83–5.91) 
 Labial tears0.62 (0.19–2.03) 2.18 (0.57–8.38) 
 Vaginal tears4.3 1 (0.48–38.92) 3.84 (0.81–18.33) 
Breastfeeding at six months   0.0006
 YesNA 2.25 (1.42–3.57) 
 NoNA 1.0 

Dyspareunia at six months was significantly associated with breastfeeding and a history of pre-pregnancy dyspareunia in both the unifactorial and multifactorial analyses (Tables 6 and 7). The association of dyspareunia with type of delivery was not significant at six months in either the unifactorial or multifactorial analyses; the proportion and odds ratio for forceps/ventouse deliveries were still raised, but the confidence interval straddled one. Figure 2 illustrates the pattern of dyspareunia by delivery (caesarean section is presented as one category).

Table 7.  Factors associated with the experience of dyspareunia six months after delivery.
Factor% (n)Unadjusted odds ratio (95% CI)P
  1. No significant differences were found on any of the following variables: age, marital status, ethnicity, birth place of woman, induced labour, artificial rupture of membranes, any type of pain relief, syntocinon used during labour, length of labour, birthweight of infant, gestation or sex of infant.

Type of delivery  0.3
  Vaginal unassisted30 (59/198)1.00 
  Forceps/ventouse37 (39/106)1.38 (0.84–2.27) 
  Caesarean (+ labour)28 (17/60)0.94 (0.50–1 .78) 
  Caesarean (no labour)21 (8/38)0.63 (0.27–1.46) 
Perineal/other genital damage  0.3
  Intact perineum24 (32/135)1.0 
  1st degree tear28 (8/29)1.21 (0.49–3.00) 
  2nd degree tear31 (22/71)1.43 (0.75–2.72) 
  3rd degree tear33 (4/12)1.59 (0.45–5.64) 
  Episiotomy36 (47/129)1.83 (1.07–3.12) 
  Labial tears29 (5/17)1.33 (0.4–4.05) 
  Vaginal tears50 (5/10)3.19 (0.87–11.71) 
Experienced dyspareunia in year prior to pregnancy  <0.0001
  Yes60 (29/48)4.22 (2.26–7.88) 
  No27 (94/354)1.0 
Breastfeeding at six months  0.001
  Yes40 (65/164)2.02 (1.31–3.10) 
  No25 (58/236)1.0 
Figure 2.

Dyspareunia by type of delivery. —♦— vaginal unassisted; —▪— forceps/ventouse; —Δ— caesarean.

Breastfeeding and contraception

Sixty percent of women (289/483) reported fully breastfeeding in the first six weeks, 23% (110/483) reported mixing breast and bottle, and 17% (84/483) reported bottle feeding only. Rates of breastfeeding then dropped. Thirty-seven percent of women (180/482) reported fully breastfeeding at around three months, 21% (103/482) reported mixing breast and bottle, and 41% (199/482) reported bottle feeding only. By six months 42% of women (282/482) reported breastfeeding (full or partial). There was no association between type of delivery and breastfeeding in first six weeks (df = 6, P= 0.2), at three months (df = 6, P= 0.3), or at six months (df = 3, P= 0.9).

Of the women who had resumed (or attempted to resume) sexual intercourse, 82% (350/425) reported using a method of contraception. Four women were pregnant again. Of the 350 women using contraception, 45% (n= 156) were using hormonal contraception (predominantly the pill), 43% (n= 151) were using condoms, 2% (n= 6) were using an intrauterine device, 3% (n= 12) were using the diaphragm, and 7% (n= 25) were using safe period or withdrawal methods. There was no association between method of contraception and dyspareunia at three months (df = 3, P= 0.8), or at six months (df = 3, P= 0.8).

Health services and postnatal sexual health

Twenty-eight percent of women (134/476) said someone talked to them about sex after childbirth before the birth of their child. These conversations were predominantly with antenatal teachers, midwives, family or friends.

After the birth of their child, 69% of women (332/480) said that a health professional talked to them about resuming sex after childbirth. These discussions were predominantly with general practitioners, midwives, and health visitors and predominantly related to contraception. Ninety-six percent of women (314/328) reported that contraception was discussed with them. Only 29% (94/328) reported that the right time to resume intercourse was discussed, and only 18% (60/328) were advised about possible changes or problems they might experience.

Ninety-one percent of women (436/481) reported attending the six week postnatal check. During the check, 62% (266/429) had a vaginal examination, and 45% (189/420) were asked about problems with their perineum/vagina. Nine percent (n= 41) reported that they had wanted to ask something but felt they could not. The main topics these women wanted to ask about were sexual matters (particularly pain and other problems) and problems with urinary and faecal continence.

Of the 337 women who reported a postnatal sexual problem, as defined in Table 4, 15% (n= 49) reported discussing it with a health professional, usually a general practitioner. Help varied from ‘none’(n= 9), through ‘advice and reassurance’ or some form of treatment (e. g. oestrogen pessaries or vaginal lubricant), to referral to secondary services (n= 9). The vast majority did not discuss their problem(s) with a health professional.

DISCUSSION

Our study showed that primiparous women experienced high levels of sexual morbidity after childbirth, with dyspareunia, vaginal dryness and loss of libido being very common. The level of reported sexual problems was very consistent: high levels of problems were reported in the first three months after delivery, which then declined by six months but not to pre-pregnancy levels.

Our findings on frequency and assessment of quality of sexual intercourse were similar to those reported in previous studies21,26,27,33. The pattern of sexual practices altered in the postnatal period, although remained broadly similar to that of the wider population of women in this age group47.

We were surprised at how few factors influenced the occurrence of dyspareunia in the postnatal period. Type of delivery and perineal damage were associated with higher rates of postnatal dyspareunia, as indicated in previous studies22,23,25,28,31, but neither of these factors were significant by six months. Since dyspareunia is cited as one of the reasons why women might opt for an elective caesarean section as preferred mode of delivery39,40, our finding is especially relevant in counselling women in the antenatal period about mode of delivery.

Women who experienced dyspareunia before pregnancy had over a fourfold chance of experiencing dyspareunia six months after the birth, compared with women who had not experienced dyspareunia before having a baby. These women are a small group but may have specific needs; they could be identified antenatally and offered appropriate help and advice.

The association of dyspareunia with breastfeeding has been documented before21,22 and is most likely to be due to the changed hormonal profile of women36, associated loss of libido31,34,36 and vaginal dryness21,22. New mothers could be alerted to this side effect and given appropriate advice regarding use of vaginal lubricants and oestrogen pessaries if necessary, while being reassured about the benefits of breastfeeding.

Health professionals take it for granted that women will resume sexual intercourse following delivery, hence they discuss contraception. However, they do not seem to be concerned about the quality of women's sexual health, since this issue is rarely discussed. The explanation may lie in the fact that at the six week check, where issues of contraception are usually discussed, approximately 60% of women will not have resumed sexual intercourse. Thus the six week check may be too early to discover chronic problems22,24,49. We also found low rates of consultation for problems with sexual intercourse (similar to a previous study31), and of the women who consulted, not all were helped. This suggests that advice and treatment relating to postnatal sexual health could be vastly improved. Our data provide at least some basis on which health professionals might reconsider the information they impart to women before and after childbirth, whether or not women are symptomatic. For example, in the antenatal period women might be counselled on both choice of mode of delivery and what they might expect in terms of sexual health outcome, and after delivery they could be given advice on how to deal with any problems they experience, an idea of when various problems are likely to resolve, and an indication of when to seek further medical help.

Limitations

Our study achieved a 61% response rate. Ideally, this would have been higher, but given the subject of the questionnaire and the inner-city location of the research, it was an acceptable response rate and comparable to other studies7,9,50. Also, it was possible from our birth records to establish whether there was a non-response bias. In fact, we had lower levels of response from women who were younger, single, Asian or Black, born outside the UK or not in full time employment. Explanations for this nonresponse are likely to include higher mobility and/or cultural and language differences. It is also possible that women experiencing postnatal sexual problems were more likely to reply, but we think this is unlikely to be a major bias. Our questionnaire asked about other areas of postnatal health; therefore, applying the same logic, women with higher rates of other problems (e.g. back pain, urinary incontinence, depression) might be more likely to reply, thus diluting the effect on postnatal sexual problems. We were also reassured by the fact that none of the variables on which we had response biases were in any way related to dyspareunia in our analyses.

Another possible limitation of our study is that we asked women to recall information from before their pregnancy and the early months of the postnatal period. Ideally, we would ask women about their experiences prospectively, but this is virtually impossible; for example, women would need to be identified and recruited before they became pregnant, retained in the study, and followed up over the years. Comparison of our current findings with the findings from our pilot study48, carried out in 1996, offers reassurance as the results from the two surveys are remarkably, similar, and therefore suggests that the findings have external validity.

We carried out a large number of statistical tests on our data (e.g. the analyses for Table 5 and Table 7 included 40 tests), and there is the possibility of a Type 1 roni correction did not substantially alter our findings. For example, using the Bonferroni correction to Table 5 and Table 7 (i.e.P < 0.05/40 =P < 0.00125) would mean that three tests were significant rather than five, and the Bonferroni correction is known to be overconservative. In fact, in our study we were surprised at how few factors were related to dyspareunia.

Our study provides new data on postnatal sexual health in women, although the picture remains incomplete as the male perspective is not included. To date, there are no reports on male perspectives of postnatal sexual health. The current practice of encouraging the father's presence at the birth may well influence postnatal sexuality and warrants further research as postnatal sex life depends on the couple, not just the woman.

CONCLUSION

Childbirth in primiparous women is associated with a high prevalence of postnatal sexual morbidity; over 80% of women in our study experienced at least one postnatal sexual problem in the first three months after birth, and two-thirds were still experiencing problems at six months. Despite the high frequency of problems, only a minority of women receive information about sexual health, and rates of consultation for problems with sexual intercourse are low. Our study suggests there may be high levels of unmet need which postnatal care services do not currently address. We also examined factors associated with dyspareunia in the postnatal period and found that in the first three months after birth, dyspareunia was significantly associated with vaginal deliveries. At six months there was some evidence of an association but it was not significant. These findings contribute to the debate about provision of elective caesarean section at the request of the woman, as dyspareunia is often cited as a reason why women might opt for this mode of delivery.

Ancillary