This qualitative serial interview study investigates the perspectives of women and their nominated significant others from late pregnancy until 6 months after birth. Serial qualitative interviews allow trust to develop between the researcher and participant, allow early questions generated from data analysis to be explored in depth later and can help validate study findings.³⁴ Including significant others captures how relationships change over time and enables infant feeding to be understood in a wider socio-cultural context. Our approach is informed by environmental and ecological theory of behavioural change, which understands health-related decisions as constantly adapting to changes in the micro, meso and macro context in which the decisions are made.³⁵ The study was conducted in two contrasting Scottish Health Boards around 100 miles apart, where maternity units were implementing the BFI. Two qualitative researchers were based at each site. The research team brought together considerable infant feeding research experience from different backgrounds: nutrition, the voluntary sector, social policy, midwifery and general practice. Conducting qualitative research with a multidisciplinary team can help challenge researcher assumptions and biases, which is important when one of the aims is to develop theoretically informed interventions to test in trials.

Breastfeeding initiation refers to the baby receiving any breast milk, even if only once. Exclusive breast feeding is defined as the infant receiving only breast milk since birth with no other liquids or solids with the exception of drops or syrups consisting of vitamins, mineral supplements or medicines.³⁶ Introduction of solids is defined as the first ever solid food offered to and taken by the baby, even if it is only a small amount. Pregnant women recruited to the study were considered as the index cases, and relationships are described in relation to them. We define significant other(s) as the person(s) identified by the woman who has the strongest influence on feeding decisions, regardless of the direction of influence (either for or against the decision).

Maternity unit databases were used to identify 459 (site 1) and 533 (site 2) women due to give birth between September and October 2009. As mothers living in disadvantaged areas are less likely to breast feed and to participate in research,⁶ we invited all women living in the three more deprived postcode quintiles of the Scottish Index of Multiple Deprivation (SIMD)³⁷ (n=420) and a smaller sample of women living in the two more advantaged SIMD quintile areas (n=121). In more advantaged areas, we recruited families where the woman or her partner had a low age of leaving full-time education, a non-professional occupation, or were immigrants to the UK which may be a disadvantage, particularly around the time of childbirth. The research commissioning brief was to aim to recruit over 75% of participants from the three more disadvantaged SIMD quintiles and select women with diverse characteristics who intended to breast feed or who had breastfed a previous baby. Invitation packs included an introductory letter on Maternity Unit-headed paper signed by a lead health professional, an information leaflet and a short opt-in characteristics questionnaire (S1) with a free post envelope to inform purposive sampling. Of 541 invitation letters sent out 4–8 weeks prior to a woman's estimated date of delivery, 72 (13%) women volunteered to participate and provided socio-demographic data. Using a sampling frame, we selected 18 women from each site for the characteristics listed in table 1. The index women were asked to identify significant others (partners, family, friends and health professionals) throughout the study and the researcher negotiated informed consent to interview a diverse range of information rich significant others at different points.

Our aim was to interview women and their significant others every 4 weeks, at a time and place to suit them. We negotiated frequency of contact, being sensitive to the emotional and physical impact the arrival of a new baby can have on a family. Face-to-face interviews took place at home during pregnancy, within 4 weeks of birth and at 6 months, with shorter, mostly telephone, interviews (0–5) in between. Two participants preferred face-to-face interviews throughout as English was not their first language. Prior to contact after birth, we consulted midwives who accessed NHS records to ensure a safe delivery had occurred. A website discussion forum was available throughout the study. This complemented interview data and enabled contributions from volunteer parents who had not been selected to participate. However, only 25 people registered; one was a woman participating in the study (4 posts) and two of the 72 volunteers who were not selected for the study posted twice each. The research team posted five questions to stimulate discussion (S2).

Interviews were semistructured, using topic guides that were modified over the course of the study to probe emerging themes in more depth and to search for disconfirming data (S3). At the end of each interview, researchers collected structured information about significant others influential since the last interview (age, relationship, distance from the family and feeding experience). In particular, any inconsistencies or changes in the person(s) nominated as significant at different time points could be explored. Similarly, researchers collected structured data at each time point about breastfeeding duration, exclusivity and introduction of non-milk liquids and solids, based on the Office for National Statistics five yearly UK survey questions.⁶

Prior to the final interview, the research team constructed seven vignettes describing a range of health or community services to help with infant feeding informed by the emergent data analysis and the research evidence (S3). The vignettes were multi-component and designed as research tools to assist the development of interventions for future research. They were given to participants to read and comment on at the final 6-month interview. Details are provided in the full report.³⁸

Data collection and analysis progressed iteratively, with the four authors involved in listening to interview recordings, reading verbatim transcripts, identifying and interpreting themes and agreeing modifications to topic guides according to the emerging analysis. All interview transcripts were entered as data units onto FrameWork software.³⁹ FrameWork is a rigorous systematic data management tool, which allows original data and researcher interpretations to be transparently documented and maintains the important direct link between coded themes, interpretations and the original interview data.⁴⁰ The four researchers independently constructed a thematic index by reading a sample of six information rich and diverse transcripts of antenatal and first postnatal interviews and then reached consensus through discussion. A further six interviews were selected in a similar manner to add to the index to cover the introduction of solids. A final thematic index for the antenatal and early postnatal interviews was agreed approximately half way through data collection when these interviews were complete and finalised for the introduction of solids towards the end of data collection. The index was used to organise, label and summarise data, which facilitated the construction of different charts, with cases (rows) and themes (columns). Charts compared summarised theme data for couples with differing attributes, for example, primiparous compared with multiparous women, early cessation of breast feeding compared with late, early introduction of solids compared with late and differences in the level of partner or significant other involvement with infant feeding. Analysis proceeded by researchers listening to interviews, reading transcripts, keeping reflective diaries, identifying interpretive themes, discussing them, generating research questions, creating different FrameWork charts to explore patterns and to search for disconfirming data. Towards the end of our analysis, we returned to the web discussion forum data to search for any disconfirming data and no such data were found.

In pregnancy, many women and their significant others anticipate breastfeeding difficulties and failure, describing scenarios where they would behave differently from their own ideals and those of the health service. Parents hedge their feeding outcomes saying they will try breast feeding and decide from day to day rather than aim for the 6-month goal stated in health service information. New parents do not want to set themselves up to fail, and the expectation-reality gap in women's breastfeeding experiences has been widely documented.¹⁰ Frustration with this gap is particularly strongly expressed by partners, who comment that antenatal information for breast feeding did not prepare them well for reality. Antenatal breastfeeding preparation ‘makes it sound so easy’ and fails to acknowledge the more negative word on the street. Most parents prefer to hear about a wide range of positive and negative feeding experiences from recent parents, including those who ‘break the rules’, whereas such parents seem less likely to be selected by health professionals as role models in antenatal education sessions. Some parents report feeling ‘pressured’ to breast feed and in extreme cases view breastfeeding promotion as ‘propaganda’ and suggest that the ‘breast is best message’ has been overdone, confirming earlier research.⁴²

Breastfeeding education classes or workshops which intensively teach positioning and attachment prior to birth, assisted by videos, dolls and knitted breasts are viewed negatively by many as ‘school-like’ and ‘patronising’. Overly scientific information about anatomy, physiology and the constituents of breast milk is seldom appreciated, as described by others.¹⁰ A strongly stated preference is to learn about positioning a baby at the breast immediately after birth, with skilled patient and timely help. Messages are perceived as ‘breastfeeding centric’, presenting breast or bottle as a dichotomy ‘you can't do both’. The provision of separate antenatal breastfeeding preparation classes rather than general feeding classes assumes an ideal of fixed early rational decision making and can infer an ‘all or nothing’ rather than a ‘try it and see’⁴² breastfeeding culture. Many women prefer not to be separated according to feeding intention and welcome open discussions about both formula and breast milk to help them to reconcile health service information with the word on the street. They want to discuss what breast feeding feels like, how it will fit into family life, the practicalities of expressing, what health service help will realistically be available, return to work and feeding in public. Families report that health professionals vary in the extent to which they take a realist rather than an idealist approach to feeding education, with the former rarely encountered but highly valued. Voluntary sector groups and helplines were infrequently mentioned by participants in our study.

What would make a difference? Skilled facilitation of interactive discussions with individuals, families or groups regardless of feeding intention, which cover the practical and emotional realities of breast and formula feeding and involve parents who have had feeding difficulties and not always lived up to ideals.

Although some women are happy with the breast feeding help available on postnatal wards, for most there is a large gap between antenatal ideals or expectations and the reality, confirming international qualitative research synthesis.¹⁰ Women value proactive help that reassures and builds confidence, continuity of care and staff who help them to latch their baby on themselves. Families describe some healthcare professionals as lacking good communication skills and breastfeeding expertise and, above all, staff were seen to be ‘rushed off their feet’, leading women to feel ‘really bad’ about asking for help, and feeling a ‘burden rather than a priority’. Occasionally, staff encourage longer hospital stays to establish breast feeding with the implied assumption that hospital care will be better than community care, although evidence does not support this.²³ Mothers are often distressed by the hospital environment and perceive that they will feel more relaxed and better able to breast feed at home. Early discharge from hospital was a primary goal for some women above breast feeding and it could be the trigger for introducing formula milk to expedite discharge. Women particularly value a member of staff sitting through a feed to provide reassurance that she is ‘doing it right’.

What would make a difference? Participants unanimously prioritise the period immediately after rather than before the birth for resources to help women breast feed. Women want someone sitting through feeds to help with breast feeding who increases their confidence.

New parents are aware that exclusive breast feeding and delaying the introduction of solids until 6 months are promoted by the health service because of health benefits and many are keen to ‘give it a go’, confirming earlier research.⁴² Participants' beliefs and opinions about these recommendations vary over the infant feeding journey both within and between families. At one extreme, some parents believe in the recommendations and persevere with exclusive breast feeding—the ease of this depending largely on personal qualities like self-confidence, determination and the availability of support. For some, the strongest motivator is to avoid a specific health outcome like food allergy or eczema, particularly in families with a history of the condition. For others, the value of ‘nature’ is paramount, with a desire to avoid formula milk and commercial baby foods, which are seen as artificial, processed and linked to future adverse health outcomes. At the other extreme, some families express doubt about the causal relationship between exclusive breast feeding with introduction of solids at 6 months and longer term health outcomes, particularly obesity. There is an awareness of mixed messages from health professionals about the exact timing of introducing solids, contributing to this doubt. Later childhood sedentary lifestyles and consumption of ‘junk food’ are perceived as stronger determinants of future health, a view expressed particularly by men. The health consequences of not exclusively breast feeding for 6 months are not readily visible or tangible, and parents who were formula milk fed perceive themselves to be as healthy as others who were breastfed. As in earlier longitudinal qualitative research, narratives describe generations of families given formula milk or solids from 3 months, with no evidence of harm and these accounts are employed to counter any implications of being a ‘bad mother’ or putting a child at risk.⁴³ For all families, current emotional and physical well-being is paramount. But those who are more sceptical about or do not prioritise health recommendations introduce formula milk or solids more readily when feeding difficulties arise, expecting immediate benefits, such as reduced breast pain, more sleep or a contented baby who is gaining weight and resolution of anxiety. Longer term future health benefits are not considered, and the priority is to ‘make sure he's getting enough to eat’ now, ‘regardless of where it comes from’.

What would make a difference? Health professional–family communication about infant feeding would benefit from a move away from checklists and instead ask open questions about experiences, values, priorities and goals with discussion about how feeding will fit into family life. This could reduce the mismatch between ideals, expectations and reality and would allow infant feeding care to be tailored to the family.

The hopes and fears of the immediate family and its social network are raised by the arrival of a new baby. The overriding goal for new parents is emotional and physical well-being, so that the important social ritual of introducing the new baby to the social network is a joyful and fulfilling experience for all, with minimal anxiety and tension. Scenarios of changing feeding behaviour to prioritise family well-being over health values are described, for example, not wanting to ‘shun to the side’ a 6-year-old daughter who is ‘desperate to help’ by giving a bottle. Breast feeding is widely promoted as a way for mothers to bond with their babies, leading to anxiety that partners, grandparents or older children will not bond with the baby unless they are actively involved in giving feeds. For some mothers and significant others, actively feeding the baby by bottle or spoon has a higher value than all other baby care tasks and is ‘therapeutic’ and ‘relaxing’. Women are keen to express breast milk as soon as possible to enable others to bottle feed breast milk and if expressing fails some introduce formula milk or solids early to fulfil this ideal. Accounts of prioritising bonding and family emotional well-being are employed to counter any implications that families are not doing the best for the health of their baby by not exclusively breast feeding. A few parents prioritise the ideal of exclusive breast feeding above all other facets of individual or family well-being. A model of ‘intensive mothering’ has been described where a strong mother–infant bond is fulfilled by exclusive breast feeding on demand and is of paramount importance.⁴⁴ For first-time parents, a mismatch can occur between expectations and the reality of how family, friends and health professionals help them practically and emotionally with feeding, parenting and household tasks. For many, there are barriers to asking for help, with few feeling confident enough to ask and women prefer professionals and others to offer help.

What would make a difference? Opportunities for significant others to be involved in discussions about roles, bonding and coping with breast feeding after birth might help families to adapt and maintain family well-being. Proactive offers of help are preferred.

A priority for many new parents is ‘getting the baby into a routine’, as attaining a regular feeding and sleeping pattern helps to adjust to changes in roles and use of time. First-time parents in particular can find the lack of time for non-baby-related activities and ‘me time’ a major challenge for which they are not prepared. Even those who have anticipated devoting themselves to feeding may struggle with the time taken to breast feed, feeling anxious to ‘get back in control’ of their lives. At the other extreme, a few women who fit the ‘intense mothering’ model willingly sacrifice other activities for their ideal and thrive on fulfilling this role. Priorities for the use of time vary but often focus on restoring pre-pregnancy activities and values, in particular sleep, image (hair, nails, clothes), exercise, friendships, couple or family time including attention to other children and housework. Reconciling these ideals and values with exclusive breast feeding can be challenging. Parents who do reconcile ideals with reality can feel empowered and those who struggle can feel guilty or inadequate as parents in relation to their own ideals or those of others. Return to work can symbolise ‘the end for the boob’ for some, who introduce formula milk or solids prior to 6 months to facilitate childcare arrangements. The moral work of breast feeding and perceptions of being a good or bad mother have been explored in depth elsewhere.^{45, 46}

What would make a difference? Women benefit from helpers who understand the tensions between ideals and reality experienced while feeding a baby, who are woman and family centred rather than focused primarily on breast feeding and can build confidence and increase self-efficacy.

In family accounts of interactions with health professionals, there is a lack of clarity about the breastfeeding ideals derived from current evidence-based guidance: exclusive breast feeding for 6 months is best; breast feeding for longer than 6 months is ideal and any breast milk is better than none.^{1, 19} In contrast, accounts of advice about the ideal time to introduce solids were more clearly recalled, but inconsistencies in health professional advice were recounted. Some parents report not always being honest about their feeding behaviour with health professionals if they were not following advice. This reflects the complexity and ambiguity of the feeding messages within current guidance, leading to communication challenges for health professionals.⁴⁷ Dichotomies of ‘good’ or ‘bad’, ‘right’ or ‘wrong’, ‘breast or bottle’ prevail in accounts of interactions with health professionals, with support for giving both formula and breast milk or self-evaluations on a continuum like doing ‘well enough’ less common. The emphasis on the technical skills of correct positioning and attachment: ‘if it hurts you are doing it wrong’ can leave some women feeling a failure, whereas other women distrust this dictum and attribute success to being ‘lucky’, allowing their self-esteem to be more protected. Rules are recounted where the rationale to parents is unclear: expressing must not begin before 8 or 12 weeks, no nipple creams or shields, bottles and teats confuse the baby and babies must feed within a certain time after birth. Our interpretation is that well-meaning health professionals who wish to promote the ideal of exclusive breast feeding until 6 months convey some advice in a black and white manner, where the supporting research evidence is less clear-cut.⁴⁸ Concern about doing infant feeding ‘properly’ adds to the pressure on unconfident new parents, with some describing it as ‘just one big guilt trip’ and feeling ‘heartbroken’ when a decision to stop breast feeding is made before the woman's ideal time. It is widely acknowledged that health professionals ‘have always got to be seen to be saying the right thing’. However, women appreciate communication with professionals that embraces the reality of how difficult it is to attain feeding ideals and is attentive to the emotional consequences of compromising between the different health and family values. Such accounts invariably feature reassurance, confidence building, normalising experiences and time spent developing a trusting relationship.

What would make a difference? Whereas a woman-centred communication style can ameliorate parental distress, a breastfeeding-centred or formulaic guideline and rules-based style can increase it. The latter can influence feeding decisions and actions as it can undermine women's confidence, experience and the wider family situation.

When physical, emotional or social difficulties arise in the first 6 months after birth, parents believe that changing the feeding behaviour is one of the few immediate actions within their control that might restore their primary goal of family well-being. Our interpretation is that these pivotal points where feeding behaviour changes represent a conflict between ideals and reality. The trigger for a pivotal point may be somatic, emotional, social, cultural or environmental and examples are well documented in the research literature describing the reasons why women change their feeding behaviour.^{6, 9, 10} For some in our study, the pivotal points were brief, intense, accompanied by significant emotional distress and feeding behaviour changed promptly. These intense pivotal points mostly occurred with transition from breast feeding to formula in the early weeks when the mother and baby are learning a new skill at the same time as recovering from the physical and emotional consequences of childbirth. They also occurred at crisis times like illness. For others, pivotal points and feeding transitions were a more gradual cognitive process, weighing up the pros and the cons and accompanied by less emotional dissonance. More gradual pivotal points are recounted particularly by parents with older children, those with strong family and social network support and around the introduction of solids. Families provide accounts of timely intervention around pivotal points which enabled them to continue breast feeding or delay the introduction of solids. There are missed opportunities for help, for example, midwife home visits when the baby is asleep or busy ward staff who have other priorities like drug rounds. Distress can increase when care is provided by an unknown person, staff are ‘too busy’, conflicting advice is encountered or where help is not available. Our analysis suggests that mothers are particularly likely to stop breast feeding or to introduce solids early at a pivotal point if they lack health professional support at an important time, lack a readily available network of family and friends to provide emotional support and practical help or have significant others who formula fed or had negative breastfeeding experiences.

From a new mother's perspective, availability of trusted family, friends or professionals who are able to provide emotional and practical help are perceived as key contributors to advantage or disadvantage with regard to infant feeding. Yet not all help provided is perceived as beneficial to breast feeding, introduction of solids or maternal well-being and it can create tension and a conflict in ideals. Pivotal points are common at around 4 months, if babies change their crying, feeding or sleep behaviours. Even parents who have ‘persevered’ with their ideals may reach a point when they introduce formula or solids earlier than planned in search of the immediate gain of a more settled contented baby, relief from their anxiety and more sleep. Many parents collect tips and suggestions from a variety of information sources, the media, family and friends, health professionals and try many strategies before introducing formula or solids. Influence depends on the value and trust attached to the source and what fits best with their circumstances.

What would make a difference? Our interpretation is that infant feeding care would improve by providing more proactive rather than reactive care, anticipating mismatches between ideals and reality that underlie pivotal points and providing skilled family-centred help to resolve difficulties.