Ms E. Schytt, Center for Clinical Research Dalarna, Nissers väg 3, 791 82 Falun, Sweden.
Objective The aims of the present study were to describe the prevalence of a number of physical symptoms, as described by women themselves, two months and one year after childbirth in a national Swedish sample and to investigate the association between specific symptoms and women's self-rated health.
Design Cohort study.
Setting Swedish antenatal clinics.
Population A total of 2413 women recruited from 593 antenatal clinics in Sweden during three one-week periods evenly spread over one year (1999–2000), representing 54% of women eligible for the study and 76% of those who consented to participate. The representativity of the sample was assessed by comparison with the total Swedish birth cohort of 1999.
Methods Data were collected by means of questionnaires in early pregnancy, two months and one year after the birth and from the Swedish Medical Birth Register.
Main outcome measures Self-reported symptoms and self-rated health.
Results Tiredness, headache, neck, shoulder and low back pain were common problems at two months as well as one year after childbirth. At two months, pain from caesarean section, dyspareunia and haemorrhoids were frequent problems, whereas stress incontinence was often reported at one year. Ninety-one percent of the women said self-rated health was ‘very good’ or ‘good’ at two months after birth, and 86% at one year. Low self-rated health was associated with symptoms that affected general physical functioning and wellbeing, such as tiredness, headache, musculoskeletal problems, mastitis, perineal pain, dysuria, stomachache and nausea. Complaints related to more specific situations, such as dyspareunia, constipation and stress incontinence were not associated with self-rated health.
Conclusion Despite the fact that physical symptoms were common two months and one year after the birth, the vast majority of women rated their health as ‘very good’ or ‘good’.
Pregnancy, delivery and puerperium are associated with great changes in a woman's body and may, even when these processes are normal, affect maternal functional health.1 Many studies have focused on specific physical symptoms during the puerperium, but only a smaller number have investigated women's health during the first year of motherhood. Brown and Lumley2 showed that 94% of Australian women had one or more health problems six to seven months after birth, and Glazener et al.3 reported a range of physical symptoms in Scottish women from the time of delivery up to 18 months later. In a large English study of long term health problems after childbirth, McArthur et al.4 showed that a majority of symptoms that lasted for more than a year after birth were still present one to eight years later. Differences in prevalence rates of symptoms have been reported in primiparas and multiparas.3,5
Common postnatal health problems such as tiredness,2,3,5 headache,3 haemorrhoids,2,3 perineal pain,3,6 constipation3 and breast problems3 may be regarded as minor inconveniences restricted to the postpartum period, which could be self-treated. Symptoms such as backache,5,7,8 urinary incontinence,9–11 faecal incontinence,12–15 flatus incontinence15 and dyspareunia16 may either increase or decrease over time but for some women these problems will persist.4,17,18 However, a majority of women do not consult a health professional even if they feel that they would need help or advice.2–4,16,19
Women usually receive little information about long-lasting physical problems that may occur as a consequence of childbirth and early motherhood, and they may therefore be unnecessarily unprepared when faced with such problems. Clinicians may find it difficult to provide relevant information because of the dearth of epidemiological data on physical health problems during the puerperium.1 The postnatal check up gives an opportunity for health consulting and is attended by 80% of the Swedish women six to eight weeks after the delivery.20 Greater awareness of the risk of long-lasting problems could lead to earlier reporting of symptoms and adequate treatment21 and possibly even to a reduction of sick leave.22 Self-rated health is a commonly used health indicator,23,24 which may predict morbidity23 and the use of health services.25 To our knowledge, self-rated health is not investigated in women after childbirth, and knowledge about women's physical health problems after childbirth in Sweden is limited.
The aims of the present study were to describe the prevalence of a number of physical symptoms, as described by women themselves, two months and one year after childbirth in a national Swedish sample and to investigate the association between specific symptoms and women's self-rated health.
For the purpose of the present study, selected data from a longitudinal cohort study of a national sample of Swedish childbearing women were used. Details of the methods have been reported earlier.26 All Swedish-speaking women who came for their first booking visit to any of the antenatal clinics in Sweden during three one-week periods, evenly spread over one year (1999–2000), were invited to participate in the study. After informed consent was obtained, the first of three questionnaires was posted, including questions about women's socio-demographic background (age, education, civil status, native language, smoking habits) and parity. The second questionnaire, which was mailed eight weeks after the birth, included questions about mode of delivery and a list of possible physical problems. Women were asked to make a retrospective assessment of each of the listed symptoms during the previous four weeks (four to eight weeks after birth) and the response alternatives were: ‘no problems’, ‘minor problems’, ‘severe problems’, and ‘very severe problems’. Women could add additional problems to the list. Self-rated health was measured by the question ‘How would you summarize your state of health at present?’ and the response alternatives were ‘very good’, ‘good’, ‘neither good nor bad’, ‘bad’, ‘very bad’. This single-item question on perceived health has shown good test–retest reliability (agreement of responses of 85–90% and kappa values of 0.64–0.73 in persons of age 18–73).27,28 The third questionnaire was mailed one year after the birth, and included an extended list of physical symptoms. Women made the same assessment of these symptoms as in the second questionnaire, but the period for occurrence of symptoms was the week prior to the questionnaire. The same question about self-rated health was asked as in the postpartum questionnaire. Obstetric data (single/multiple birth, gestational age, perineal trauma), maternal weight and height prior to pregnancy, and infant outcomes (birthweight and head circumference) were retrieved from the Swedish Medical Birth Register.29
For the purpose of the present study, only women who responded to all three questionnaires were included. The background characteristics of the women were compared with those of a cohort of all women who gave birth in Sweden in 1999 (n= 84,729) according to the Swedish Medical Birth Register in order to determine on how representative the study population was to the general population.
All 608 antenatal clinics in Sweden were approached and 593 chose to participate in the study. Of those eligible for the study, 3191 (71%) women consented to participate, and 3061 completed the first questionnaire, 2762 the second and 2563 the third. In total, 2450 women completed all three questionnaires. From this group, 37 women who did not answer any of the questions about physical health were also excluded. The final study group comprised 2413 women (i.e. 76% of those who consented to participate and 54% of all women eligible for the study). The average length of gestation when filling in the first questionnaire was 16 weeks. The second questionnaire was completed at 10 weeks and the third at one year and two weeks after the birth.
The non-response rates to the questions on physical symptoms in the second questionnaire varied from 0.4% to 1.3% except for the questions on pain from perineal tear or episiotomy (2.3%) and pain during sexual intercourse (22.8%). Of the 550 women who did not answer the last question, 496 said they had not had sexual intercourse after the delivery. Data are presented with and without these women included. The non-response rates to the corresponding questions in the third questionnaire ranged from 0.6% to 2.0%, except for the question on pain during intercourse (2.4%).
In the analysis of severity of symptoms (0 = no problems, 1 = minor problems, 2 = severe problems and 3 = very severe problems), the two last response alternatives were collapsed and defined as ‘major problems’. Comparisons between primiparas and multiparas as well as between study group and all women who gave birth in Sweden in 1999, were calculated by χ2 test and Student's t test. Differences in prevalence of symptoms over time were estimated by McNemar's test after categorising the symptoms into ‘no symptoms’ (0 on the scale) and ‘symptoms’ (1–3 on the scale). Self-rated health was dichotomised into ‘good’ (very good + good) and ‘fair/poor’ (neither good nor bad + bad + very bad). The association between self-rated health at two months and one year after birth was described as risk ratio (RR), with 95% confidence intervals (CI). Associations between physical symptoms and self-rated health were tested by means of logistic regression analyses for each subgroup, and are presented as odds ratios (OR) with 95% CI.
The Statistical Package SPSS for Windows (SPSS, Chicago, Illinois, USA) was used for the statistical analyses. The study was approved by the Regional Research and Ethical Committee at the Karolinska Institute, Sweden (Ref. No. 98–358).
Women's average age at recruitment was 29.5 years and 44% of the study population were primigravidas. Seventy-nine percent had a normal vaginal delivery, 7% an instrumental vaginal delivery and 14% a caesarean section. Of women with a vaginal delivery, 56% had a perineal tear of first or second degree, 3% a third or fourth degree tear and 12% an episiotomy. Eighty-nine percent of the women breastfed their babies at two months, and 21% at one year after the birth. Women in the sample did not differ from the national birth cohort of all 84,729 women who gave birth in Sweden in 1999 regarding marital status, body mass index (mean BMI), mode of delivery, infant birthweight and infant head circumference. However, fewer women in the study group were younger than 25 years (14%vs 16%, P < 0.001), older than 35 years (10%vs 12%, P < 0.01), multiparas (56%vs 58%, P < 0.05), smokers (9%vs 12%, P < 0.0001) and born in another country (8%vs 17%, P < 0.0001), compared with the total birth cohort.
Four to eight weeks after the birth, tiredness was the most common problem, reported by almost two-third of the women (Table 1). In women who had a caesarean section, more than one-third reported minor or major problems with post-operative pain. Neck and shoulder pain, low back pain and dyspareunia came third in ranking by being listed by 28–29% of the women. Other frequently reported problems were headache, haemorrhoids and constipation. Most women who had any of the listed symptoms said their problems were minor. However, the most common symptoms caused women the largest problems.Figure 1 shows the number of symptoms in women two months and one year after the birth. Few women had none of the listed symptoms, only 9% of the primiparas and 11% of the multiparas. Additional health problems not listed in the questionnaire were reported by 3.4%, equally distributed between primiparas and multiparas, and included endometritis, symphysiolysis, fever, joint ache and Candida infection in the breast.
Table 1. Self-reported physical health problems 4–8 weeks after childbirth. Values are presented as n (%).
Some of the physical symptoms were more common in primiparas, such as sore nipples, dyspareunia, perineal pain and dysuria. Primiparas assessed their symptoms of constipation as of major degree to a larger extent than multiparas. Dyspareunia was the second most common problem reported by the primiparas, affecting 39% and causing 9% major problems. In women who said that they had actually resumed sexual intercourse, there was a slightly higher prevalence: in primiparas minor and major problems were reported by 40% and 11%, respectively, and in multiparas the corresponding figures were 25% and 3%.
Similarly to the postpartum assessment, tiredness was the most common symptom one year after the birth and this accounted for most of the cases classified as a major problem (Table 2). One in four women had sleeping problems. Symptoms that were still common one year after birth, and reported by about one-third of the women, were headache, neck and shoulder pain and low back pain. Other common problems were colds, which were only assessed at one year, haemorrhoids, stomachache and dyspareunia. Similarly to the four to eight weeks' assessment, the most frequently reported symptoms also caused women the highest degree of problems. Thirteen percent of the primiparas and 12% of the multiparas did not report any symptoms at one year (Fig. 1).Table 2 also shows that most symptoms did not differ by parity. However, multiparas had more problems with colds and stress incontinence, and primiparas had more problems with perineal pain and dyspareunia.
Table 2. Self-reported physical health problems one year after childbirth. Values are presented as n (%).
The fact that the one-year assessment included symptoms that had occurred during a shorter time period (the week prior to the questionnaire) than the two months' assessment (the last four weeks prior to the questionnaire) made it difficult to analyse whether specific symptoms increased or decreased over time. Lower rates at one year could be explained by the shorter assessment period, however, higher rates probably indicated an increase of symptoms. Such symptoms, which increased in prevalence from two months to one year after birth, in primiparas as well as in multiparas, were headache (P < 0.0001), neck and shoulder pain (P < 0.0001), low back pain (P < 0.0001), stomachache (P < 0.01), nausea (P < 0.0001) and sleep disorders (P < 0.0001). Other symptoms than those listed in the questionnaire were reported by 2.8% of the women at one year after birth, and included joint ache, symphysiolysis, thyroid diseases and skin problems.
The vast majority of women said that their self-rated health was ‘very good’ or ‘good’ at two months after the birth (primiparas: 92%; multiparas: 91%) with a slight reduction one year after the birth (primiparas: 86%; multiparas: 85%) (Table 3). The primiparas were more satisfied with their health than the multiparas at both time points. For all women, reporting fair/poor health at two months after birth increased the risk of making the same assessment at one year (RR 4.1, 95% CI 3.4–4.9).
Table 3. Self-rated health two months and one year after delivery. Values are presented as n (%). Differences between primiparas and multiparas are analysed by χ2 test.
All women (n= 2407)
Primiparas (n= 1062)
Multiparas (n= 1345)
All women (n= 2389)
Primiparas (n= 1054)
Multiparas (n= 1335)
Neither good nor bad
Primiparas vs multiparas: P value
Table 4 shows the association between specific symptoms and women's self-rated health at four to eight weeks and one year after the birth. General symptoms associated with fair/poor self-rated health at two months as well as at one year after birth were headache, sleeping problems and tiredness. Low back pain seemed to have an equal effect on self-rated health at both time points, while neck and shoulder pain was associated with self-rated health only at one year after the birth. Breast problems (mastitis) and perineal pain were associated with self-rated health at two months after delivery only, and gastrointestinal symptoms (nausea and stomachache) and dysuria at one year only.
Table 4. Statistically significant associations between physical symptoms and assessment of self-rated health as fair/poor, two months and one year after childbirth, analysed by logistic regression. Values are presented as OR (95% CI).
Two months (n= 2184)
One year (n= 2285)
Neck and shoulder pain
Low back pain
The association between physical symptoms and self-rated health differed by parity. At two months after delivery, six factors seemed to have a negative effect on self-rated health in primiparas: sleeping problems (OR 1.8, 95% CI 1.1–3.1), tiredness (OR 6.4, 95% CI 2.5–16.4), low back pain (OR 2.0, 95% CI 1.2–3.3), sore nipples (OR 2.0, 95% CI 1.1–3.5), mastitis (OR 4.2, 95% CI 1.3–3.9) and perineal pain (OR 2.3, 95% CI 1.3–3.9). In multiparas, three factors were statistically significant: headache (OR 2.9, 95% CI 1.9–4.4), sleeping problems (OR 2.7, 95% CI 1.7–4.1) and neck and shoulder pain (OR 2.1, 95% CI 1.4–3.1). At one year, the corresponding variables in the primiparas were: tiredness (OR 4.1, 95% CI 2.5–6.9), neck and shoulder pain (OR 1.7, 95% CI 1.2–2.6), nausea (OR 3.1, 95% CI 1.9–5.1) and stomachache (OR 3.4, 95% CI 2.2–5.4); and in multiparas: headache (OR 1.9, 95% CI 1.3–2.6), sleeping problems (OR 1.6, 95% CI 1.1–2.3), tiredness (OR 3.1, 95% CI 1.9–5.0), colds (OR 1.5, 95% CI 1.1–2.1), neck and shoulder pain (OR 2.0, 95% CI 1.4–2.9), low back pain (OR 1.6, 95% CI 1.1–2.3) and stomachache (OR 2.2, 95% CI 1.5–3.3). Age and mode of delivery had no confounding effect and were removed from all models.
Physical health problems were common and few women were free from symptoms two months and one year after the birth. Some of the symptoms seemed to be temporary for most women, such as perineal pain, dyspareunia and breast problems. However, others seemed to remain or even increase over time, such as tiredness, headache, sleeping problems, back pain and neck and shoulder pain.
The sample of this study was fairly well representative to all women who gave birth in Sweden in 1999, with the exception of non-Swedish-speaking women. Younger and older women, multiparas and smokers were slightly under-represented, which may have resulted in a small under-estimation of women's health problems. A limitation of our study was the different time intervals during which symptoms were assessed, at two months and one year after the birth, which made conclusions about changes over time hazardous. The one-year questionnaire was developed between the two time points and included an extended and more specified list of symptoms, and a follow up of some of the symptoms during the puerperium could not be made.
Tiredness was by far the most common symptom and was causing women most problems at both time points. Sixty percent of the primiparas as well as the multiparas suffered from tiredness during the puerperium, with only a slight reduction in prevalence at one year. The corresponding figure in women of the same age in the Swedish population was 15%.30 Lack of sleep is probably one, but not the sole, explanation to the high proportion of women reporting tiredness. The adjustment to motherhood, an altered relationship with the partner and siblings, and limited time for own self may all be contributing factors.31,32 Lee and Zaffke33 found that women's sleep patterns changed after childbirth, an alteration that was more rapid in multiparas than in primiparas. Three months after the birth, the mothers were more able to quickly get into deeper, more restorative stages of sleep after awakenings than during pregnancy. This finding may explain the low prevalence of self-reported sleep disorders four to eight weeks after the birth, with no differences between primiparas and multiparas. However, the fairly low prevalence of sleeping problems could also be an under-estimation due to a non-specific question. Women may have reported difficulties in falling asleep rather than total sleep time, sleep efficiency and quality of sleep.
The prevalence of neck and shoulder pain increased during our study period from 30% to 36%, but did not reach the level in Swedish women of corresponding age (45%) as reported by Statistics Sweden.30 Research has shown that a rest from physical activity and mental relaxation is needed in order to prevent symptoms of neck and shoulder pain.34 It is possible that Swedish women, who have the opportunity of 14 months of parental leave, experience lower levels of stress when they have their focus mainly on children and household chores, and can relax from the demands of their position in the workforce. It is also possible that women without symptoms have better support from their partner and others than women with symptoms, as social support has been associated with low levels of stress after childbirth.35
Symptoms of low back pain were also less frequent in the study group during the puerperium (28%) than in women in the total population (34%),30 but the prevalence had increased at one year (34%). Contradictory to our findings, Östgaard et al.8 found a decrease in the proportion of women with back pain after delivery, and this decrease levelled off at approximately six months after delivery. The increasing proportion of women with symptoms of low back pain in the current study may be explained by the demands of childcare rather than an effect of pregnancy and childbirth as such, but this needs further investigation.
Of the urogenital problems, dyspareunia was the most common symptom four to eight weeks after the birth. These problems were more frequent and intense in primiparas than in multiparas (40%vs 20%). Even if the symptoms were less common one year after delivery, 13% of the primiparas and 9% of the multiparas still had problems. Barrett et al.16 found that within three months after the birth, 62% of the women experienced dyspareunia and 31% at six months. The pre-pregnancy level of 12% was fairly close to our finding at one year (11%), which indicates that women may have recovered from this specific consequence of childbirth at one year. In our study, 45% of the primiparas and 41% of the multiparas said that they had not yet had sexual intercourse at two months postpartum, which would explain the great number of non-responders. However, of these women, 12% of the primiparas and 4% of the multiparas said that they had suffered from dyspareunia, suggesting that a completed sexual intercourse had been postponed due to pain. Other common urogenital symptoms at four to eight weeks after delivery were post-operative pain, which troubled more than one-third of those who had undergone a caesarean section, and stress incontinence, which has been reported in detail previously.26
The vast majority of the women rated their general health as ‘very good’ or ‘good’ at two months after birth, however, fewer at one year, despite the fact that physical health problems were common. Fair/poor self-rated health at two months increased the risk of more inferior health 10 months later. Symptoms associated with pain and general physical functioning, such as tiredness, headache, low back pain and neck and shoulder pain, seemed to have a stronger impact on self-rated health than physical problems related to specific situations, such as allergic symptoms, constipation, incontinence of urine and faeces and flatus.
A comparison of our findings with a health survey conducted by Statistics Sweden,30 including the question ‘How would you assess your general state of health?’ with the same response alternatives as in the current study, showed that women in our sample rated their health as better at two months (P < 0.0001) and one year (P < 0.05) after the birth than women of the same age in the Swedish population (Fig. 2). The more positive assessment of women in our study may have several explanations. In Western Europe, choosing to get pregnant and keep the pregnancy can be seen as an indication of satisfaction with life situation, including one's health. Otchet et al.36 suggested that women's more positive perception of their health status soon after delivery, compared with a community sample, could be explained by the view that childbirth itself is as an expression of good health. Relief and euphoric reactions to the birth of a healthy baby may also influence women's overall feelings in the immediate period after the delivery.37 Women may expect a certain amount of physical symptoms as a consequence of pregnancy and childbirth and regard these problems as natural and of temporary nature. The symptoms may also be considered as minor inconveniences in relation to pregnancy itself or labour and delivery. The decline in self-rated health from two months to one year after the birth was mirrored by the increase of most symptoms that were associated with self-rated health. The symptoms may actually have increased, or problems that women may have regarded as natural during the puerperium may have been less well tolerated at one year after the birth, and therefore been more easily reported.
Symptoms that previously have been shown to have a significant impact on women's lives in terms of daily life activities and emotional wellbeing, such as dyspareunia,1,16 stress incontinence38 and anal incontinence,39 were not associated with self-rated health in this study. Previous research suggests that self-rated health is a multidimensional and complex concept that is only partly predicted by the presence or absence of various symptoms and conditions, even if these symptoms are associated with great suffering.23,40,41 Further research including not only the effect of physical symptoms, but also of socio-demographic and psychosocial factors, health behaviour and mental health, is needed in order to better understand self-rated health in the context of childbirth.
SUMMARY AND CONCLUSIONS
Very few Swedish women reported fair/poor self-rated health, despite the fact that physical health problems were common. Low scores for self-rated health were mainly associated with symptoms that affected general physical functioning and wellbeing. Physical symptoms related specifically to pregnancy and delivery seemed to be temporary for a majority of women. However, symptoms that may be associated with childcare such as tiredness, sleeping problems and low back pain remained or increased during the first year of motherhood. For some women, there is a need for greater attention of health problems after childbirth. On the other hand, our results draw attention to childbirth as a health promoting event, as women rated their health as better than women of corresponding age in the Swedish population, and also that some symptoms were less frequent during the postpartum period. The contradictory finding of high level of self-rated health and large number of symptoms warrants further research.