Women's health after childbirth: a longitudinal study in France and Italy


Correspondence: Dr M.-J. Saurel-Cubizolles, INSERM Unit 149, 16 avenue Paul Vaillant Couturier, 94807 Villejuif Cedex, France.


Objective To describe the prevalence of various symptoms five months and twelve months after childbirth in two European countries, according to employment, financial difficulties and relationship with partner.

Design Longitudinal multicentre survey with a similar design in France and Italy. Women were interviewed three times: at birth and twice after childbirth. At 12 months after childbirth, the survey was postal in both countries.

Population Women who had been delivered of their first or second baby in three maternity units in France and five units in Italy. Women who had multiple pregnancies, a stillbirth or known neonatal death were excluded.

Results The response rates were 83% in France and 88% in Italy. Fifteen symptoms were considered. The results were similar in the two countries and showed that the prevalence of most symptoms was higher at 12 months than at five months after childbirth. When their baby was one year old, more than half of the women reported backache, anxiety and extreme tiredness. Around one-third of women reported headache, lack of sexual desire, sleep disorders and depression. Piles, constipation and painful intercourse were also common. One year postpartum women with financial problems or a difficult relationship reported poor psychological health more frequently.

Conclusion Physical and emotional problems are common after birth, and they tend to increase over time. Backache, headache and piles can seriously interfere with day-to-day life. Sexual problems also may be a source of unhappiness for the woman and her partner. Extreme tiredness, anxiety and depression may make a woman feel guilty for not corresponding to the image of a healthy, happy and well-coping mother. There is a link between financial problems or a difficult relationship with her partner and her own wellbeing. Health professionals should be aware of the high prevalence of health problems among new mothers, and of the social context of women who are more likely to suffer from them. They should counsel the women in their care, in order to help them to find the best solution, be it medical or social in nature.


In industrialised countries medical attention is usually high during pregnancy but drops quickly after childbirth. In France one medical examination is mandatory six weeks postpartum, while in Italy, it is generally scheduled 40 days after the birth. In both countries the duration of stay in hospital is becoming shorter. It has recently been emphasised that ‘maternal morbidity is an under-researched and neglected field’1. The few studies, all recent, on mothers’ health after birth have found consistent and worrying results: the incidence of various health problems is high; many of these problems persist well beyond what is usually considered as the postpartum period; and only a small proportion of affected women ask for professional help2,3.

The aim of this report is to present results on the health of new mothers from a longitudinal study carried out in France and Italy, in which women who had given birth to their first and second baby were interviewed immediately after the birth, and at five and 12 months postpartum. The prevalence of symptoms is described for the whole sample and according to women's employment status, reported financial hardship and the quality of the relationship between the woman and her partner.


The data were collected from a study on mothers'work and health after childbirth4,5. All women who had been delivered their first or second child were eligible, except mothers of twins, those whose baby was stillborn or had died after the birth, and those unable to speak the national language. In France, three hospitals were involved (two in the Parisian region and one in the Champagne-Ardennes region); in Italy, five hospitals were involved, all in the north-east (three urban and two rural). From September 1993 to July 1994, eligible mothers were contacted in the maternity ward and invited to participate: in France, 761 out of 770 women agreed to participate and answered the first questionnaire (a face-to-face interview) two to three days after delivery; in Italy, 820 out of 858 accepted. When the baby was five months old, mothers received a postal questionnaire in France and were personally interviewed at home in Italy; this was the only difference between the two countries. On the child's first birthday, in both countries mothers received the third questionnaire through the post with an accompanying letter and a stamped addressed envelope. Overall, 632 women in France and 723 in Italy answered the three questionnaires, making a response rate of 83% and 88%, respectively.

Nonrespondents differed significantly from respondents regarding a number of variables. In both countries mothers who were young, poorly educated, not employed or employed with a low qualification job were more likely to be lost to follow up. In France lone mothers, mothers whose partner was unemployed, and those reporting financial problems were also less likely to answer all three questionnaires6. However, in both countries maternal health indicators before pregnancy, whether the pregnancy was wanted, type of delivery and the baby's health at birth did not discriminate between respondents and nonrespondents.

The analysis includes women who answered the three questionnaires, who were not pregnant and had not recently given birth to another child 12 months after the index delivery. There were 589 women in France and 697 in Italy.

Data on self-reported morbidity were obtained by asking women at the time of the interview if they had each one from a list of physical or psychological health problems. Women were asked to specify whether the problem bothered them ‘a little’ or ‘quite/a lot’. In the following analysis, these two positive answers have been combined. The same list of health problems was proposed in France and Italy, at five and 12 months postpartum. Moreover, at the first interview, women were asked about their health status before the pregnancy. Fewer symptoms were considered at this time than after childbirth.

The McNemar test was used to compare prevalence of symptoms at five and 12 months postpartum. As a reference, the prevalence of symptoms is described before pregnancy, if available. The prevalence of symptoms 12 months after childbirth is also described according to employment status (having a job vs not having a job), reported financial hardship (since the birth of the child, did you have serious financial problems? yes/no) and relationship with the partner. The quality of the relationship was described by the following question: ‘At the moment, the relationship with your husband or partner is: very good, good, fairly good, not too good, difficult, and very difficult’. The last four answers were collapsed and coded ‘not good’. To also take lone mothers into account, this variable was combined with the information of whether or not the woman was cohabiting with a male partner. The resulting scores were: 1 = for very good cohabiting relationship; 2 = for good cohabiting relationship; 3 = for not good cohabiting relationship; 4 = lone mother. Statistical analysis was performed with SAS in France and with SPSS in Italy.


Table 1 shows the prevalence of symptoms reported by French and Italian mothers five and 12 months postpartum. In both countries the prevalence of symptoms are high; moreover, except for piles in both countries, and backache and painful intercourse in Italy, there was a significant increase in the reporting of all symptoms between five and 12 months. When their baby was one year old, more than half of the mothers reported backache, anxiety and extreme tiredness, while around one-third reported headache, lack of sexual desire, sleep disorders and depression. Piles, constipation and painful intercourse were also common. Overall, except for headache, French mothers reported more symptoms than Italian mothers. Nevertheless, the increase of symptoms between the two periods shows a very similar tendency in both countries.

Table 1.  Health status reported by women by months after childbirth. Values are given as %, unless otherwise indicated. NS = non significant.
 Time after childbirth
 Italy (n= 697)France (n= 589)
Symptom5 months12 monthsP5 months12 monthsP
  1. The statistical comparisons have been performed using the McNemar test

Backache49.450.9NS47.464.7< 0.001
Headache22.545.1< 0.00120.738.4< 0.001
Piles16.520.7< 0.0216.426.2< 0.001
Varicose veins8.28.5NS17.919.9NS
Constipation12.616.9< 0.0114.525.7< 0.001
Urinary incontinence1.75.0< 0.0017.614.0< 0.001
Faecal incontinence1.03.0<< 0.02
Urinary infections1.43.6<< 0.01
Vaginal infections2.39.0< 0.0014.611.7< 0.001
Painful intercourse11.913.9NS15.620.1< 0.02
Lack of sexual desire17.530.6< O.OOl24.939.6< 0.001
Sleep disorders14.131.6< 0.00126.734.2< 0.001
Anxiety36.255.8< 0.00135.256.0< 0.001
Tearfulness/depression20.431.6< 0.00119.035.4< 0.001
Extreme tiredness46.160.7< 0.00148.467.5< 0.001

For some symptoms, prevalence before pregnancy were available. Headache was reported by 34% of French respondents and 48% of Italian respondents, and backache by 46% in both samples. In the French sample, 48% of women reported being depressed or anxious, compared with 45% and 20% of Italian women reported being anxious or having symptoms of depression, respectively. Moreover, in the Italian sample 12% of women described piles, 10% lack of sexual desire and 5% painful intercourse.

As our sample included only women who had their first or second baby, we looked at possible differences between these two groups. Differences were scarce. Five months postpartum, French and Italian primiparae were significantly more likely to report painful intercourse than women who had had their second baby. French women who had a second baby more often reported varicose veins, and Italians more often reported urinary infections. Twelve months postpartum, primiparae were significantly more likely to report painful intercourse in France and a lack of sexual desire in Italy, while Italian women with two children suffered more often from extreme tiredness.

The prevalence of symptoms did not vary in relation to employment status (Table 2). In Italy, no significant difference was observed. In France, only few differences were found. Employed French women reported constipation, lack of sexual desire and tiredness more frequently, while they reported urinary infections and depressive feelings less frequently. The rate of women's employment was similar in the two samples: 67% in Italy and 69% in France.

Table 2.  Employment and health status 12 months after childbirth. Values are given as %, unless otherwise indicated. NS = non significant.
 Employment status
SymptomWith a job (n= 469)Without job (n= 228)PWith a job (n= 405)Without job (n= 184)P
  1. The statistical comparisons have been performed using the χ2 test

Varicose veins8.19.2NS19.520.6NS
Constipation16.018.9NS28.519.6< 0.05
Urinary incontinence5.14.8NS13.412.6NS
Faecal incontinence2.63.9NS5.55.5NS
Urinary infections4.12.6NS3.58.2< 0.02
Vaginal infections8.111.0NS11.412.5NS
Painful intercourse14.313.2NS19.820.6NS
Lack of sexual desire32.325.9NS42.533.2< $0.04
Sleep disorders32.230.3NS32.737.5NS
Tearfulness/depression30.932.9NS31.643.7< 0.004
Extreme tiredness62.956.1NS70.560.9< 0.02

In both samples, the prevalence of symptoms were more pronounced in combination with financial hardship (Table 3). In France, women experiencing serious financial problems had higher prevalence of backache, headache, faecal incontinence, sleep disorders, anxiety, depressive feelings and tiredness. In Italy, significant differences were observed for headache, lack of sexual desire, anxiety and depressive feelings. Some prevalences are very close in the two countries (e.g. the percentages of depressive feelings). In the Italian sample, fewer women reported financial hardship than in France (7%vs 23%).

Table 3.  Serious financial problems and health status 12 months after childbirth. Values are given as %, unless otherwise indicated. NS = non significant
 Serious financial problems
SymptomYes (n= 49)No (n= 647)PYes (n= 133)No (n= 440)P
  1. The statistical comparisons have been performed using the χ2 test.

Backache59.250.4NS78.860.1< 0.001
Headache59.244.0< 0.0553.434.1< 0.001
Varicose veins12.28.2NS20.319.6NS
Urinary incontinence6.14.9NS17.411.6NS
Faecal incontinence6.12.8NS10.03.9< 0.006
Urinary infections6.13.4NS7.54.1NS
Vaginal infections14.38.7NS15.810.0NS
Painful intercourse18.413.6NS25.017.5NS
Lack of sexual desire42.929.5<0.0542.138.6NS
Sleep disorders36.731.2NS45.130.3< 0.002
Anxiety71.454.6<0.0268.452.0< 0.001
Tearfulness/depression53.129.8< 0.00155.628.0< 0.001
Extreme tiredness71.460.0NS78.264.2< 0.003

The variables concerned with the relationship with the partner were strongly linked to women's wellbeing, especially psychological, in both samples. Moreover, the differences in symptoms were similar in France and Italy, with respect to the quality of the relationship (Table 4). Women describing their relationship as ‘not good’ were much more likely to report symptoms related to sexual life and psychological wellbeing, with a progressive increase in the three groups of women evaluating their relationship as ‘very good’, ‘good’and ‘not good’. In Italy, the percentage of headache was also significantly related to the relationship, with high values among women with ‘good’ or ‘not good’ relationships. In France, this tendency was observed for vaginal infections. In the Italian sample, 22% of women living with a partner described their relationship as ‘not good’ 12 months after childbirth, while this prevalence was 17% in the French sample. The proportion of noncohabiting women was low in both samples (3% in Italy and 4% in France). Nevertheless, the prevalence of symptoms tended to be lower for noncohabiting women than for women who described their relationship with their partner as ‘not good’.

Table 4.  Relationship with partner and health status 12 months after childbirth. Values are given as %, unless otherwise indicated. NS = non significant.
 Women's assessment of relationship with partner
SymptomVery good (n= 294)Good (n= 236)Not good (n= 149)Lone mothers (n= 18)PVery good (n= 307)Good (n= 155)Not good (n= 96)Lone mothers (n= 25)P
  1. The statistical comparisons have been performed using the χ2 test.

Headache37.451.352.327.8< 0.00233.844.542.740.0NS
Varicose veins7.56.812.811.1NS16.920.627.124.0NS
Urinary incontinence5.
Faecal incontinence2.73.82.70NS3.65.29.512.0NS
Urinary infections3.44.72.70NS2.
Vaginal infections7.58.912.111.1NS8.116.114.616.0< 0.05
Painful intercourse11.214.818.85.6NS14.025.231.920.0< 0.001
Lack of sexual desire17.337.347.716.7< 0.00129.249.764.616.0< 0.001
Sleep disorders26.527.548.327.8< 0.00127.036.846.948.0< 0.001
Anxiety43.556.877.266.7< 0.00147.159.475.4868.0< 0.001
Tearfulness/depression22.430.149.750.0< 0.00123.133.670.860.0< 0.001
Extreme tiredness54.460.673.855.6< 0.00261.671.681.264.0< 0.002


These results show that French and Italian mothers reported several symptoms in the year following the birth of a first or second child, and that the prevalence of most of these symptoms increased significantly between five and 12 months postpartum.

This study has a number of strengths and limitations. Beginning with limitations, the samples were not representative at the national level. However, in both countries, different hospitals catering for different populations of women, urban and rural, were included. Moreover, as far as France is concerned, the demographic and social characteristics of the women interviewed in this study correspond closely to those of the French population of new mothers7.

We studied only women who had had their first or second child. This represents 73% of deliveries in France7 and 80% in Italy8. Our results show only a few differences between the health of these two groups of mothers. Researchers comparing the frequency of symptoms between primiparae and all multiparae considered together found the same result. In Australia, Brown and Lumley2 sent 1300 women a postal questionnaire six to seven months postpartum: out of 14 symptoms, the only difference consisted in primiparae reporting more often a painful perineum and sexual problems. In the United Kingdom, a postal questionnaire study of 438 mothers reported health problems experienced between two and 18 months postpartum; again, there were few differences related to parity1. Primiparae suffered from a painful perineum and from piles more often, while multiparae reported more urinary infections. Overall it seems that, except for primiparae consistently reporting more sexual difficulties and soreness in the perineum9, there are few differences in the health of new mothers according to parity. Therefore, the fact that we studied only women who had had a first or second baby should not limit the generalisation of our results to other parities. Another limitation was that the study was observational and not experimental. Causality cannot be inferred from these associations. Moreover, some of our indicators (e.g. health symptoms, the quality of the relationship with the partner and financial difficulties) were subjective self-reports. We analysed the links between employment, financial difficulties and quality of the relationship with the partner and symptoms using indicators collected at the same time (12 months postpartum). This may entail a problem of ‘contamination’between these sets of data. But, as social conditions may change over a year, it seemed logical to examine the relationship between social conditions present at a certain time and health experienced at the same time.

The strengths of this study are that it is longitudinal and a high response rate was obtained. The time in which we collected information about health problems was very precise. Most women answered the questionnaires within a very tight span of time and respondents were asked to report the symptoms they experienced at the time of the interview. This gives a more exact assessment of the health of new mothers than results obtained in retrospective enquiries of, for example, Glazener et al.1 and Mac Arthur et al.10. Another strength was that the same questionnaire was used in two different populations of women. Notwithstanding differences in cultural beliefs, social organisation and obstetric practices in France and Italy, the proportion of symptoms was similar and the same steady increase between five and 12 months was observed. We conclude that these results are not peculiar to our samples, but may be general to all new mothers in industrialised countries. The fact that French women tended to report more symptoms than Italian women is consistent with other data concerning adult female population in these countries11, and may be explained by a generally worse health status or by a lower threshold in perceiving and reporting ill health.

It is difficult to compare our results with those of other studies, as different questions were asked in different groups of women at different times. Nevertheless, mothers interviewed in Australia six to seven months postpartum reported a similar amount of backache (43.5%), piles (24.6%), urinary incontinence (10.7%) and extreme tiredness (69.4%)2. MacArthur et al.12 interviewed women 10 months after delivery and found that 6% suffered from faecal incontinence, a proportion very near to that reported by French mothers 12 months postpartum. The comparison with results by Glazener et al.1 is less easy, as they present data on symptoms experienced between two and 18 months1. The frequencies they reported are lower than ours and those by Brown and Lumley2, but, interestingly, in all surveys the most common health problems are the same: extreme tiredness, backache, depression, headaches, piles and sexual problems.

An unexpected result is the increase in reported symptoms between five and 12 months postpartum. This runs against the popular belief that as the baby grows older and the event of childbirth recedes, mothers will recover their health, strength and wellbeing. However, no other longitudinal data in a population are available on the matter. Glazener et al.1, studying women at different points in time, noted that “almost every health problem mentioned was more common after hospital discharge”. Moreover, they found that, while the proportion of women with health problems slowly declined with time, symptoms such as depression, tiredness, backache, piles and headaches did not decrease significantly between eight weeks and two to eighteen months postpartum. In the United States Gjerdingen et al.13 studied a sample of first time mothers, all with a paid job during pregnancy, interviewing them one, three, six, nine and twelve months postpartum. They found that some physical problems (e.g. breast and vaginal discomfort) showed a higher prevalence one month postpartum. Several of these problems (e.g. haemorrhoids, tiredness and constipation) persisted beyond one month, and were joined by other ‘late’symptoms, including respiratory symptoms and sexual concerns. Other results from a small sample of American primiparous women show that they were more likely to report tiredness and loss of energy at 14 and 19 months after the birth than they were at the end of the sixth week postpartum14.

In our study it is impossible to explain this increase in health problems. The health of the women really could have worsened, as babies actually tend to become more demanding over time and social pressure may have increased as the mother is expected to have recovered from childbirth. Another explanation could be that, while health had not actually worsened, over time women found some symptoms less ‘normal’and more bothersome. For instance, they may consider ‘normal’to be exhausted or to suffer from urinary or faecal incontinence five months postpartum. But if these problems persist seven months later, they may consider them as ‘symptoms’and therefore report them. Another explanation is methodological: the very fact of answering the questionnaire may render women more perceptive to health matters. But the fact that some symptoms, such as varicose veins in both countries, and backache and painful intercourse in Italy, did not increase at all is inconsistent with this explanation.

Without such data in the general female population in these two countries, it is difficult to be sure that the prevalence are much higher after childbirth than at any other point in the lives of young adult women. It is a fact that, for general symptoms, prevalences are high among women: for instance, in Nordic countries in 199315, complaints of headache concerned 56% of women under 30 years of age and 51% of women between 30 and 49, while lower back pain affected 40% and 37%, worry 47% and 39% and tiredness 67% and 56%, respectively. These rates are not very far from those derived from our data. For some symptoms, we have a description of the previous situation: 12 months after childbirth, prevalences were higher than before the pregnancy, except for headaches where the prevalence was similar. The difference was clear for anxious or depressive troubles, and specially high for piles, painful intercourse and lack of sexual desire.

As regards employment, women had not resumed work at the same time in both countries: in France, most women returned to work at the end of three months, while in Italy they returned to work later, and the times at which they resumed work were evenly distributed throughout the year4. Overall, in both the French and Italian samples, being employed had no adverse effect on the mothers’physical or mental health one year after birth. This result may appear counterintuitive, as employed mothers of young children spend more hours working, with the accumulation of paid employment, housework and child-care work, and conversely dedicate less time to leisure, personal care and sleep16. But employment also means an increase in social contacts and financial autonomy, a respite from domesticity and sometimes intellectual stimulation and an improvement in self-esteem, all factors known to promote health. Italian national data show that among cohabiting mothers in the age group 20–34 years, those who are employed spend more time working than full-time housewives (42%vs 22% work ≥ 70 hours per week), but are also more likely to be ‘very satisfied’with family relationships (53%vs 46%) and less likely to be ‘unsatisfied’with their work (20%vs 33%)8. Previous results from an Italian study of primiparae who were employed during pregnancy show that 84% of them were back to work 15 months after the birth. Being employed was not associated with the health problems considered (i.e. backache, extreme tiredness or lack of sleep)17. In a British longitudinal study of first-time mothers, the relation between employment, physical and mental health was inconsistent. The only difference was that employed mothers were more tired 11 and 18 months postpartum, but not at the earlier and later contacts18. In the United States, Gjerdingen et al.13 asked first-time mothers about 76 symptoms: the only difference related to employment was that employed women reported more often symptoms of respiratory infections than those who stayed at home. This could be explained by greater exposure to micro-organisms in the workplace and in child-care centres. As far as mental health is concerned, results from a number of studies suggest that what is important for new mothers is not only employment, but rather the congruence between their values and wishes and their actual occupational status19. This is confirmed by the results of an analysis of the same sample of French and Italian women, using the General Health Questionnaire as an indicator of psychological distress 12 months postpartum5. Furthermore, in the French sample, being unemployed and looking for a job was actually associated with more psychological distress20.

Women experiencing economic difficulties reported more symptoms. More particularly, in both countries, financial problems were significantly associated with headache, anxiety, and feelings of depression. In a Spanish sample restricted to women having had a job during pregnancy, women who reported serious financial problems 12 months after childbirth were also more likely to be depressed21. A similar trend was found in a British study with a sample of mostly disadvantaged mothers. One year after childbirth, financial worries were strongly related to cumulated indices of mental and physical health22. Unemployment, whether by the woman or her partner, was also strongly associated with poor health. The relation between unemployment and poor health was consistent, regardless of class, as measured by housing tenure and of previous mothers’health problems. Other studies of postpartum depression have found a positive association between unemployment of both parents or of the mother and depression in the mother23,24.

In both the French and the Italian sample, the quality of the relationship with the partner was strongly associated with different symptoms related to psychological distress. Mothers with a less than ‘very good’relationship reported more often lack of sexual desire, sleep disorders, anxiety, depression and fatigue. The association between a difficult relationship and mothers’psychological distress is well-known in the literature on postpartum depression. Single mothers did not report more health problems; on the contrary, they tended to report fewer symptoms than cohabiting mothers whose relationship was ‘not good’. This result is confirmed by the multivariate analysis carried out on the same samples, using the General Health Questionnaire as an indicator of depression5. In a British study Oakley et al.22 found that one year after birth, problems with partners were strongly related to poor psychological and poor physical health. In the French sample, where data were collected, 4% of mothers reported violence by the partner in the year after childbirth; the proportion was 3% among women who were steadily cohabiting from pregnancy through the first postpartum year, 7% among the steadily noncohabiting and 24% among those who changed their status from pregnancy through 12 months postpartum (P < 0.01). Battered mothers were significantly more likely to report psychological distress and to use psychotropic drugs and were more likely to have been hospitalised in the year after the birth (15%vs 8%)25.


Physical and emotional problems are common after birth, and they tend to increase over time. Backache, headache and piles can seriously interfere with day-to-day life, sexual problems may be a source of unhappiness for the woman and her partner, and extreme tiredness, anxiety and depression, besides lowering the quality of life, may make women feeling guilty for not corresponding to the image of a healthy, happy and well-coping mother. Furthermore, urinary and faecal incontinence, while rarer, may be socially devastating.

In our study, a majority of mothers had paid employment 12 months after birth, and employment was not clearly associated with either physical or mental health. A small group of mothers were not cohabiting with a male partner, but they did not report symptoms more frequently than the women living with a partner. Financial worries and a difficult relationship with the partner were strongly associated with a higher prevalence of symptoms, especially of a psychological nature. This emphasises the importance of the social context and the quality of the relationship in the wellbeing of new mothers. Women were asked in open question on what could make the life of new mothers easier. The most common answer concerned the sharing of the responsibility and the concerns of motherhood. Regardless of their employment status, women asked for day care that was accessible and of good quality. They also were interested in joining parents’groups to discuss issues such as infant development, breastfeeding, mothers’health and difficulties with their partner26.

Most health problems in new mothers appear to go unrecognised1–3,12. Some physical problems, such as urinary and faecal incontinence, can be improved medically27. Other problems seem to require a wider, more social response. In both cases, more attention should be paid to the difficulties and needs of new mothers, well beyond the short six weeks period traditionally allotted for recovery from childbirth.


The French survey was funded by a contract CNAMTS-INSERM 1992. The Italian survey was funded by the Region Friuli-Venezia Giulia and by the IRCCS Burlo Garofolo, Trieste. The authors obtained a grant for collaborative research between France and Italy in the context of an INSERM-CNR agreement (1997–1998). The authors would like to thank the women who participated to this study, the personnel of the maternity units involved, the interviewers and the technical research staff.