In the Netherlands, the majority of women (80%) start breastfeeding, which drops to 57% after one month [1, 2]. These figures represent a common pattern in breastfeeding rates in most western countries, with the highest dropout occurring in the first weeks after birth [3, 4]. Although the importance of breastfeeding in high-income countries is less evident than it is in low-to-middle income countries, as stated by several authors [5], including the Netherlands, public health campaigns that target women have been developed with the aim of extending the initiation and duration of breastfeeding. These campaigns follow international guidelines from the World Health Organization (WHO) [6, 7], which recommend exclusive breastfeeding until the age of six months [8]. The extension of the 2025 targets includes the aim to increase the rate of exclusive breastfeeding in the first six months up to at least 50% [7]. Although the duration of breastfeeding in the Netherlands has increased [2], the national campaigns have been more successful in increasing initiation rates than in extending duration [2, 9].

To explore the reasons that underlie women’s decisions to stop or continue breastfeeding in the first month after delivery, a qualitative research project was designed in the northern part of the Netherlands [10]. This research project had been conducted from an anthropological perspective by examining breastfeeding practice in the broader context of socio-cultural and socioeconomic features, including the unique tradition of how reproductive (home) care is organized in the Netherlands.

Empirical qualitative data were collected by interviewing primiparous women. An initial study was aimed at gaining insight into the women’s intentions to start breastfeeding as well as their actual breastfeeding practice [10]. Themes identified from the women’s narratives in the initial study were: ‘making arrangements for childbirth,’ ‘reflecting on the intention,’ ‘becoming a mother,’ ‘combining breastfeeding with work’ and ‘learning about breastfeeding’ [10]. Among those themes, the theme ‘learning about breastfeeding’ required further examination to contribute to a better understanding of women’s decisions to stop or continue breastfeeding. Therefore, we collected additional data and performed a secondary analysis of the previous and new data. This secondary analysis is addressed in this article and aims to identify and categorize the content as well as the sources of breastfeeding knowledge.

There are a few specific features in the medical as well as the sociocultural and socioeconomic context of breastfeeding in the Netherlands. One crucial characteristic is the high rate of home deliveries as part of the unique Dutch system of integrated maternity healthcare, including clinical obstetric as well as midwifery home care [11]. In this system, postnatal maternity nursing care at home is available from day one to day eight, and, although the rate of home deliveries has been decreased from 38% in 1991 to 18% in 2013, the majority of women (95%) make use of this home care for an average of seven days [12, 13]. In this postpartum period, maternity nurses support breastfeeding initiation and assist women in practicing feeding breast milk. Prior to that, in the prepartum period, women who intend to breastfeed receive information from their antenatal care providers during any of the medical consultations or in antenatal breastfeeding classes.

Another characteristic of the breastfeeding context in the Netherlands is that additional information and support can be obtained from lactation consultants. However, this expertise is not always included in the integrated (funded) care system, which could therefore generate differences in access to this service among women with different socioeconomic backgrounds [10].

Apart from the information from professionals, women can obtain information from non-professionals around them, such as relatives or friends. However, considerable transformations in knowledge acquisition have occurred in society because of decreasing breastfeeding rates. The probability of women observing other women practicing breastfeeding has become minimal nowadays, and observational learning in new mothers, which relied on seeing women around them breastfeeding, has become a thing of the past [9, 14].

Additional development in breastfeeding knowledge acquisition is the use of the internet, which generates an immense increase in access to information. Some authors suggest that, from the perspective of the generation of young adults aged between 20 and 35 years, if information concerning breastfeeding is not distributed on the internet, it would be as if this information did not even exist [15]. Women’s knowledge acquisition is not limited to a single type of knowledge or to one specific source, and new mothers acquire their breastfeeding knowledge from multiple sources [16].

The literature presents a variety of concepts and classifications referring to types of knowledge sources. Jordan [17] described biomedical and authoritative knowledge as professional sources of knowledge. Informal and lay knowledge are presented as popular sources of knowledge, for example, by McFadden et al. [18]. The categories of professional and popular sources have also been described by Kleinman [19], where ‘professional sources’ are further classified as specialist professional sources, and ‘popular sources’ generate non-specialist nonprofessional knowledge. In our analysis, we use this classification of ‘professional sources’ and ‘popular sources’ to categorize the sources of knowledge.

A better understanding of the content and sources of breastfeeding knowledge in Dutch women may contribute to theory building as well as policy development concerning infant feeding practices in high-income countries and the various ways in which breastfeeding can be understood [20]. Identifying and categorizing the content and sources of breastfeeding knowledge could guide professionals such as midwives and maternity nurses as well as others in the women’s surroundings to engage in disseminating knowledge and, therefore, support women in achieving their breastfeeding goals.

The research project was designed as a cyclic process [21]. This allows moving back and forth between the different stages of a research process and conducting a secondary analysis of data previously collected [21].

Data collection was done in two rounds, as described in the introduction, by interviewing two samples of primiparous women who received prenatal care from midwives in the Northern part of the Netherlands. All women were interviewed twice, prepartum and postpartum. The prepartum interviews (1) were conducted in the women’s third trimester, and the postpartum interviews (2) between four and six weeks postpartum.

The series of prepartum and postpartum interviews for the first round of data collection was conducted in 2008 among a sample of eight women in the province of Friesland. The second round, aiming to collect additional data to get insight into one specific theme identified in the first round, i.e., ‘learning about breastfeeding,’ was accomplished in 2011 among a sample of five women in the province of Groningen to complement the previous sample. Friesland and Groningen are provinces in the north of the Netherlands. During the time between the first and the second round of data collection, there had been no adjustments to any of the breastfeeding guidelines, recommendations, or campaigns, neither at international (e.g., by WHO) nor at the national level. Therefore the time interval has no impact on the findings. Data from both rounds were analyzed jointly to address the research question for this paper.

All women were recruited through their midwife practices. During their checkup visits in gestation week 20, their midwives inquired whether they had the intention to breastfeed and, if so, invited them to participate in the research. Apart from primiparity and the intention to breastfeed, inclusion criteria were that the women were in good health and able to speak Dutch or English. Additional criteria for the postpartum interviews were that the mothers as well as their infants were in good health and that there were no contraindications for breastfeeding.

The prepartum interviews were conducted at the midwife clinics and took 30-45 minutes. The postpartum interviews were conducted at the participants’ homes and took 45-60 minutes. One infant’s father was present at the prepartum interview, and one was present at the postpartum interview. During the postpartum interviews, all infants were present in the same or a different room.

The interview guides covered topics derived from the theory of planned behavior [22, 23], including breastfeeding knowledge [24]. The prepartum interview questions thus concerned breastfeeding intentions, attitudes, social environment, and norms, perceived behavioral control, and breastfeeding knowledge [9]. The postpartum interview questions dealt with these same topics in retrospect and included childbirth experience and actual feeding practice. The interview guides were semi-structured, and questions were open-ended with probes in variable sequences. The interview guide for the second round of data collection included additional probes to get a deeper understanding of the content and sources of breastfeeding knowledge.

The records of all interviews were anonymized and transcribed ad verbatim for analysis. These transcripts were first analyzed by open coding, applying the principles of grounded theory [25-27]. One key theme identified in the initial study was ‘learning about breastfeeding' [9], which was selected for further analysis, along with additional data collected in the second round. Analysis of the additional data was performed according to the same procedure [25-27]. Subsequently, the concepts’ professional sources and popular sources were used as sensitizing concepts for the deductive categorization of those five knowledge themes, after which cross categorization was conducted. Sensitizing concepts are background ideas that inform the exploration of the research problem and offer ways of seeing, organizing, and understanding experiences [28-31]. This process included analytic questioning [21], which was accomplished by a thorough analysis of all codes that were covered by the five inductive knowledge content themes and by exploring whether the source of this knowledge was professional or popular.

A total of 26 interviews (13 prepartum and 13 postpartum) were conducted, at which point the analysis indicated that data saturation had been achieved [26], as no new codes emerged.

Before starting the interviews, written informed consent was given by all participants. The women could withdraw from the study at any time without explanation. Ethical approval was obtained from the Institutional Review Board.

Particular consideration had been given to data collection ethics, such as the researcher’s positionality. As a cultural anthropologist, the first author had conducted ethnographic fieldwork in different foreign cultures but had never done so in her own culture. She had been a breastfeeding mother herself once, and, as Hoppe [32] had also pointed out, the women talked openly to her because they felt she had been through the same experience [32]. The disadvantages of being familiar with the research topic are that knowledge and experience can also block the researcher from recognizing unexpected issues [21]. The author was reflecting continuously on her own experiences and positionality [33]. She explained to the women that she was interested in their personal experiences and opinions from their own perspectives and in their own words. All responses would be considered equally right or relevant. The author indicated that since she was a researcher and not a physician, midwife, nurse, or social worker, she would not give her opinion or provide information. At the same time, she assured thems that the information in their narratives would be treated confidentially and would not be shared with other professionals. In doing so, minimizing risks of harm or discomfort was accomplished [21].

Data analysis ethics included achieving consensus among the authors concerning interpretation and coding of data, and overall reflexivity generated scientific rigor. Reporting was accomplished according to the Consolidated Criteria for Reporting Qualitative Research (COREQ).

The aim of this study was to identify and categorize the content and sources of breastfeeding knowledge in a population of primiparous mothers in the northern part of the Netherlands. Five knowledge content themes were identified: the pros and cons of breastfeeding, how breastfeeding works, individual breastfeeding practice, expressing milk, and formula feeding. Knowledge from professional sources was present and helpful for some of the knowledge themes. The benefit from popular sources, however, was perceived as limited.

When further examining the women’s popular sources, we found that the infants’ grandmothers were no important source of knowledge, even though they are essential significant others in the women’s environment [34, 35]. Most grandmothers had not practiced breastfeeding themselves and, therefore, could not share their experiences. This has also been described by other researchers [14]. Furthermore, Grassley & Eschiti [34] found that the sharing of experiences between generations is limited.

The second group of significant others involved the infants’ fathers. As suggested by the result of the earlier study of this research project [9], fathers were important in supporting the women's intentions, but not in contributing to knowledge acquisition. Hauck & Irurita [36] also found that although the infant’s father is usually important in supporting the woman’s attitude toward breastfeeding, he is generally not perceived as a knowledge source.

A third group representing popular sources consists of ‘others,’ referred to as ‘they.’ These others were often perceived as making normative and critical comments, therefore, not contributing to knowledge dissemination. Previous researchers have reported prejudiced comments by ‘others’ as well [36-39].

The limited dissemination of knowledge from popular sources has substantial consequences because it is from those sources where community members start their health activities [19]. As a result, women should begin their search for breastfeeding information from non-professional sources, which they may find in their intuition. Also other researchers have corroborated the perception of intuitive knowledge as a satisfactory knowledge source [40, 41].

Using empirical data from the emic point of view in a population of primiparous mothers in the northern part of the Netherlands, five knowledge content themes were identified: the pros and cons of breastfeeding, how breastfeeding works, individual breastfeeding practice, expressing milk, and formula feeding. Knowledge from professional sources was present and helpful in some of the knowledge themes. However, the usefulness of popular sources was limited. Intuition was inductively identified as a specialist, non-professional knowledge source. Women must rely on additional sources other than the knowledge shared by their colleagues, relatives, and friends due to limited exposure to practising breastfeeding. In doing so, they perceive the pressure to follow breastfeeding recommendations and perform according to the standards of natural motherhood at the same time. This study’s contribution to the breastfeeding discourse is twofold: first, it contributes to theory building concerning professional and popular knowledge sources by identifying intuition as an inductive concept of specialist non-professional knowledge. Second, messages from breastfeeding campaigns should address not only the health benefits of breastfeeding but also all knowledge content themes and target peers, relatives, and friends, according to recommendations for breastfeeding practise. Recognizing intuition as an adequate expert source of knowledge supported by professionals will facilitate women in making informed decisions, thus contributing to a supportive environment for breastfeeding women in the Netherlands.