National Maternity Survey 2014

This study used data from the cross-sectional National Maternity Survey conducted in England in 2014.34 A random sample of 10 000 women who gave birth during a 2-week period, excluding those aged less than 16 years and those whose baby had died, were selected by the Office for National Statistics from birth registrations. They were sent a questionnaire 12 weeks after the birth asking about clinical events and care during pregnancy, labour and birth, and in the postnatal period. Questionnaires could be returned by Freepost, completed online or completed by telephone with the aid of an interpreter if necessary. Up to three reminders were sent to non-respondents using a tailored reminder system.35

Exposure and outcome measures

Women were asked questions about postnatal health, including ‘Did you experience any of the following 10 days, one month, and three months after the birth of your baby?’ as used in previous National Maternity Surveys, with answer options including ‘Fatigue/severe tiredness’ among various other postnatal symptoms. The time points of 10 days, 1 month and 3 months had been selected pragmatically to reflect the range of experience up to the time of the survey. Women were asked whether they had mental health problems and asked to complete the Edinburgh Postnatal Depression Scale (EPDS). Questions were also asked about sociodemographic variables, including age, parity, Index of Multiple Deprivation (an area-based measure representing the level of socioeconomic deprivation of the neighbourhoods in which the respondents lived, comprising elements related to income, employment, education, health, crime, barriers to housing and services, and living environment), ethnicity and age on completing full-time education; prepregnancy and antenatal health and well-being, variables related to labour and birth, including duration of labour and mode of delivery; multiple birth, a checklist relating to postnatal maternal health and symptoms; infant health and feeding at 3 months; perceptions of their baby at 3 months indicated by the number of positive and negative adjectives circled from a list of 16; and indicators of maternal–infant relationship reflected in women’s sense of when their baby belonged to them, on a 6-point scale ranging from ‘during pregnancy’ to ‘not quite yet’, and whether she considered her baby more or less difficult than average. Postnatal partner support was estimated by summing the scores for five activities: changing baby’s nappy, supporting feeding, helping when the baby cries, bathing the baby and playing with the baby; each scored 1–4 and summed (high score=more support). Midwifery support was estimated by women responding that they had/had not received enough help and advice about the baby’s crying, sleeping and feeding, and also whether the woman would have liked to see the midwife more or less often in the postnatal period. The full list of variables included in the analysis is given in the online supplementary appendix.

Analyses

Descriptive statistics were used to describe the prevalence of fatigue at 10 days, 1 month and 3 months, with proportions and means as appropriate. Associations with PPF were tested using the χ² test. A cut-off of 13 or more was used for the EPDS. To assess the sociodemographic and clinical characteristics of women who experience PPF, separate logistic regression models were estimated for each time point. Analyses were thus carried out separately for PPF at 10 days, 1 month and 3 months as it was anticipated that different variables may be important at different time points. Four groups were entered: (1) sociodemographic variables; (2) antenatal and intrapartum variables; (3) indicators of the mother–infant relationship; and (4) partner and midwife support. Due to the large number of comparisons made in determining the key predictors of PPF, only variables which were statistically significant at p<0.001 were entered into binary logistic regression. All logistic regression analyses were adjusted for maternal age, parity, Index of Multiple Deprivation, ethnicity and age left full-time education as these were potential confounding factors. A full case analysis was carried out as missing data were generally less than 5%.

Patient and public involvement

Patients were not directly involved in the conceptualisation or design of this study. Women were selected at random for the survey by the Office for National Statistics from birth registrations. The survey questions, including those relating to postnatal health, were developed in consultation with a research advisory group with representatives from user groups, maternity services liaison committees and members of national charities associated with maternity care. The reports from the National Maternity Surveys are available on the National Perinatal Epidemiology Unit (NPEU) website.

Prevalence of PPF

According to the symptom checklist, 38.8%, 27.1% and 11.4% of women experienced fatigue/severe tiredness at 10 days, 1 month and 3 months, respectively. Of those women who reported PPF at 10 days, 46% also reported it at 1 month; however, of those who reported PPF at 1 month, only 30% also reported it at 3 months. These figures varied significantly by maternal age, Index of Multiple Deprivation (an area-based measure of deprivation), whether the woman had left full-time education before age 16 and parity, such that PPF was more common in women who were older, primiparous, more highly educated and residents of less deprived areas. There were marginal differences by ethnicity (table 1).

Associations between risk factors and PPF

Sociodemographic and predictive variables that were significantly associated with PPF (p<0.001) in univariate analyses were entered into binary logistic regressions separately for PPF at 10 days, 1 month and 3 months; the results are shown in table 2. Different variables were important at different times, with the exception of maternal age which was significant throughout. Women aged 20–24 years were significantly less at risk of PPF at each time point compared with those aged 30–34, with ORs of 0.60, 0.45 and 0.40, respectively, at 10 days, 1 month and 3 months. At 1 month women aged 25–29 years were also at significantly reduced risk (OR=0.58), and at 3 months women aged 40 or over were at significantly increased risk (OR=2.00) compared with those aged 30–34 years. Multiparous women were at significantly reduced risk of PPF at 1 month and, to a lesser extent, at 10 days compared with primiparous women. Leaving full-time education aged 16 years or less was significantly protective at 10 days but not subsequently. Reported postnatal health problems, particularly depression and anxiety, were significantly associated with PPF at each time point. In addition, at 10 days and 1 month, post-traumatic stress disorder symptoms were significantly raised, and at 3 months sleep problems not associated with the baby were higher in women who had PPF. Women who were still breast feeding at 3 months, either exclusively or partially, were also at increased risk of PPF. Raised EPDS score, although statistically significant in the univariate analysis, dropped out in the multivariate logistic regression.

The mother–infant relationship

Table 3 shows the indicators of maternal–infant attachment in women with and without PPF at 10 days, 1 month and 3 months, adjusted for sociodemographic variables. At each time point, women who had PPF used significantly more negative adjectives to describe their baby and perceived their baby as more difficult than average. Women who had PPF, especially at 3 months, experienced significant delay in feeling that their baby belonged to them in addition to having more negative feelings about their baby, although it was relatively uncommon for women to describe their baby as belonging to them ‘not quite yet’ (n=29).

Women were also asked about physical well-being in the first few days and at 3 months after giving birth, with answer options ‘very well’, ‘quite well’, ‘tired and uncomfortable’, ‘exhausted’ and ‘very ill’. In the first few days and at 3 months, 13% and 5%, respectively, reported feeling ‘Exhausted’, but of these women only 68% and 55%, respectively, also reported feeling ‘Fatigue/severe tiredness’. Those who reported feeling ‘exhausted’ or ‘very ill’ were also significantly more likely to report feeling that their baby belonged to them ‘only recently’ or ‘not quite yet’ (15.6% of ‘exhausted’, 14.3% of ‘very ill’ women reported ‘only recently/not quite yet’ compared with 4.8% overall).

Partner and midwife support

We postulated that women who had less support from their partner may be at greater risk of PPF. However, this was not borne out in the results (table 4). On the contrary, after adjustment for sociodemographic variables, women whose partner was more involved in practical postnatal care (changing nappies, supporting feeding, helping when the baby cried, bathing and playing with the baby) were more likely to be experiencing PPF at 3 months and, to a lesser degree, at 1 month. Similarly, the number of days of paternity leave was positively associated with PPF at 1 month, although this was of only marginal significance.

Women with PPF at 3 months were significantly more likely to be seen for longer by their midwife in the postnatal period; however, these women were also significantly less likely to report that they saw the midwife as much as they wanted. Women with PPF were also more likely to report not receiving enough help and advice about the baby’s feeding (at 10 days), crying (at 1 month) and sleeping (at 3 months). Women with PPF also tended to use more peer, online and web support, with OR (95% CI) of 1.45 (1.26 to 1.67), 1.25 (1.08 to 1.45) and 1.35 (1.17 to 1.55), respectively, for women with PPF at 10 days, and slightly lower ORs at 1 and 3 months (data not shown).

Strengths and limitations

This study was based on a large, population-based random sample of recent mothers, with considerable diversity among respondents. The importance of some specific risk, care and individual factors was identified. However, the response rate was 47% and, in common with many other surveys,42–44 there was significant under-representation of hard-to-reach groups. PPF was as reported by women using a single item as part of a symptom checklist rather than one of the many scales. The cross-sectional survey was conducted at 3 months post partum, and thus causality cannot be inferred from associations found and women may not have accurately remembered some of the details reported. In particular, while fatigue at the time of the survey is likely to be accurately reported, fatigue at 10 days and 1 month may be less well recollected and subject to recall bias. However, studies comparing women’s reports of events around childbirth with medical records or other recorded data have demonstrated good recall.45 46

Implications for healthcare professionals

These findings highlight that, for some women, PPF can be severe and long-lasting and may require intervention. Several interventions to reduce PPF have been evaluated in good-quality randomised controlled studies.14 22 31 47 48 These were a mixture of self-management, telephone support, exercise and education, and all were reported to be beneficial except for the purely self-directed intervention which used the Tiredness Management Guide to help women to self-manage their fatigue.48 Supportive interventions and availability of additional advice and support targeted to those women with the greatest need are to be encouraged. The nature of the additional advice and support should be tailored to the needs of individual women. The association with more negative perceptions of their infant and of infants who may have been ill or born preterm may be a key point for health professionals working in postnatal care to consider. There is also a need for antenatal preparation for women and their partners in approaching the transition to parenthood so that they have realistic expectations, can enlist practical help and emotional support when required postnatally, and enhance their own coping skills.

Implications for research

The cross-sectional nature of this study has precluded teasing out of the interaction between PPF and postnatal depression. Although the two are clearly linked, the nature of the relationship is unclear. Further exploration of the impact of PPF on the developing mother–child and partner relationships, and the influence of method of infant feeding, in prospective longitudinal studies, is also likely to better illuminate the way in which individual differences may contribute. There is a wide range of estimates of point and period prevalence of PPF in the literature which requires further elucidation.