Life course learning experiences and infant feeding practices in rural Rwanda

Abstract Most studies about infant and young child feeding (IYCF) practices are often perceived as an individual choice depending on mothers' or caregivers' knowledge or attitudes and are focused on mothers' failure rather than successes in adequately feeding their children. However, the role of life course experiences in IYCF is less investigated. Applying a Salutogenic Model of Health, this study on 14 mothers looks at women's life course learning experiences shaping appropriate IYCF practices during the first year of child's life in a rural district of Rwanda. Transcripts from in‐depth interviews were analysed using thematic analysis. Results indicate that positive social interaction with parents or grandmothers during childhood such as sharing meals, parental role models for dietary choices and cooking skills gained by participating in household food preparation played a role in shaping appropriate IYCF practices. Negative experiences during childhood also had a positive influence on IYCF practices for some participants by converting life course constraints into learning opportunities. Motherhood increased mothers' sense of responsibility over their children's health and nutrition. Moreover, mothers' participation in community cooking classes and role modelling approach were strong avenues that enabled their learning through positive interactions and encouragement. Nutrition promotion interventions should consider tailoring nutrition advice to the complexity of mothers' life course experiences by creating opportunities for positive learning experiences of appropriate IYCF practices.


| INTRODUCTION
Optimal infant and young child feeding (IYCF) practices have great potential for reducing child malnutrition and thereby contributing to a reduction of child mortality rate (Black et al., 2008). Many factors interact to determine how a mother feeds her child. IYCF practices are often perceived as an individual choice depending on mothers' or caregivers' knowledge, attitude or belief (van Woerkum & Bouwman, 2014), and most IYCF studies focus on exploring why mothers or caregivers fail rather than on why some mothers or caregivers manage to have healthy IYCF practices (Ramani et al., 2019). IYCF practices (unhealthy and healthy) are learned, supported and expressed within the dynamics of everyday context in which mothers live (Ahishakiye et al., 2019). Past life experiences are known to direct how people make food choices and decisions in the future (Bisogni, Jastran, Seligson, & Thompson, 2012;Devine, 2005). However, the interaction between individuals (in this case mothers) and their environment and the way in which experiences throughout the life course play a role in mother's decision-making to manage optimal IYCF practices are less investigated.
The Salutogenic Model of Health (SMH) developed by Antonovsky (Lindström & Eriksson, 2010) can support addressing the above knowledge gap. The SMH acknowledges the active role of people in creating health and that health develops from the interaction between people and their everyday life context (Mittelmark, Bull, & Bouwman, 2017 Antonovsky viewed health not as a state but as a dynamic move along a continuum, referred to as the ease/dis-ease continuum (Antonovsky, 1987). Along the life course, human beings are confronted by stressors (challenges) and learn to identify and apply resources that enable them to cope in either a health-promoting or health-damaging manner (Lindström & Eriksson, 2010). If people deal successfully with the stressors, they can maintain their health or move towards the ease end of the continuum. Unsuccessful coping with stressors can lead people to a movement towards the disease end of the continuum (Antonovsky, 1987;Lindström & Eriksson, 2010). The SMH has two central concepts: generalized resistance resources (GRRs) and sense of coherence (SOC) (Idan, Eriksson, & Al-Yagon, 2017). The GRRs are resources within the individuals (e.g. attitudes, self-efficacy and knowledge) or in their environment (e.g. social support) that can be used to counter or to cope with the stressors of everyday life. The GRRs provide one with sets of meaningful, coherent life experiences that are characterized by consistency, a balance between overload and underload and participation in shaping the outcome (Vinje, Langeland, & Bull, 2017).
The ability to recognize and use these resources is the meaning of the second central concept: the SOC (Eriksson, 2017). Individuals with a strong SOC are better able to identify resources and use them in a health-promoting way. Those with a strong SOC have the feeling that life is comprehensible, manageable and meaningful. Life experiences are defined as experiences when people learn how to deal with life in general and acquire problem-solving skills that help to shape one's SOC (Vinje, Langeland, & Bull, 2017). Life experiences characterized by consistency, an underload-overload balance of stimuli and participation in shaping outcomes, contribute to the development of a strong SOC and allow one to reach out in any situation and use the resources to deal with stressors (challenges) (Vinje, Langeland, & Bull, 2017).
In line with the SMH, the central question in this study is, 'why do some women manage to feed their children well despite their poor living conditions?' Because of the role of life experiences within the SMH, it is important to consider the life course perspective when studying food-related behaviours, such as IYCF practices. From a life course perspective, the sequence of socially defined events and roles over time plays an important role (Kok, 2007). The life course research approach takes into account the history and path dependency (the dependency of later developments on previous ones) of the development of practices over time and links the practices to the dynamics of the social and cultural environmental contexts in which the individuals live (Aschemann-Witzel, 2013). Life course research applied to nutrition-related health practices helps to understand how past experiences interact with current food practices or choices (Devine, 2005;Devine, Connors, Bisogni, & Sobal, 1998 • Life course experiences including both positive and negative past food upbringing, social interactions with parents or grandmothers during childhood and parental role models for dietary choices play a role in shaping subsequent appropriate IYCF practices.
• The sense of responsibility over child's health and participation in community cooking classes during motherhood are powerful experiences in shaping IYCF practices.

| Participants
Our study is part of the larger longitudinal research that involved 36 mother-child pairs to explore actual breastfeeding and complementary feeding practices and influencing factors from birth until 1 year. Table 1 provides a brief description of the larger longitudinal study.
Mothers' age in the larger study ranged from 18 to 40 years, with a mean age of 34 years. Most mothers (89%) were married. The main occupation for all was farming. Ninety-four per cent had the ability to read and write their mother tongue. Five per cent had not attended any formal education, 47% did not complete primary school and 42% completed primary school.
In this study, participants consisted of a specific group of 14 mothers 'doing well' out of the 36 mother-child pairs. The selection was based on two criteria: (1) following recommended IYCF practices and (2) good growth of the child (growth according to the World Health Organization growth standards) at 1 year of age. The recommended IYCF practices included initiation of breastfeeding within 1 h of birth, exclusive breastfeeding for the first 6 months, continued breastfeeding at 1 year and timely introduction of complementary foods at 6 months. Anthropometric measures that were collected as part of the larger study were used to check that the children were growing well. Children's recumbent lengths were taken at birth and then monthly up to 12 months post-partum following standardized procedures. Height-for-age difference defined as the difference in observed height and median height of a child of the same age and sex from the World Health Organization 2006 growth standard (Leroy, Ruel, Habicht, & Frongillo, 2015) was calculated. Using individual linear regression with height-for-age difference as dependent and age as independent variable, a child was qualified as positive grower when having a beta-coefficient being zero or significantly (p < 0.05) positive reflecting growth according or above the standard. A total of 14 mothers complied with the above two criteria, and their respective community health workers confirmed the selected 14 months as 'doing well'.

| Data collection
Narrative inquiry consisting of systematically listening to people's life stories and providing access to peoples' life experiences (Keats, 2009) was used to collect information about participants' life course experiences. Specifically, 'Food-Life-Story' narrative inquiry that recognizes the active role of people in constructing their own life and their eating practices was used to further map out specific life experiences and how people reflected on these experiences in relation to their food practices (Mittelmark, Bull, & Bouwman, 2017). One visual method, the timeline, was used as a form of graphic elicitation to encourage the construction of rich temporal stories on how one's pasts shape presents (Sheridan, Chamberlain, & Dupuis, 2011). Data were collected through two consecutive sequences of interviews: preliminary and in-depth interviews. During the preliminary interviews, participants were given details on how to draw their timeline representing their key life events and past experiences. Participants were asked to think back and to create their personal chronological timeline in which they were free to include all relevant aspects of their life trajectories, key moments and past life experiences with food and child feeding practices. The same opportunity was used to check participants' demographic and socio-economic characteristics. During the followup in-depth interviews centred on their timelines, participants were asked to talk about their life experiences. By use of probing questions, T A B L E 1 Socio-demographic characteristics of women in the larger longitudinal study

| Data analysis
The interview recordings were transcribed verbatim and translated from Kinyarwanda into English. Each participant was assigned an identity to guarantee anonymity. Atlas.ti software was used for coding and analysing the data. Analysis across all transcripts was done thematically following the protocol described by Braun and Clarke (2006). The first step of analysis consisted of reading and rereading of the transcripts to familiarize with the data. Second, researchers generated initial codes within the transcripts in the left side margin of the transcripts. The initial coding was done by the first author. The second author also coded independently half of the transcripts. Codes were compared and discussed by the first and the second authors to ensure that the data segments were categorized correctly. The two researchers' categorization was almost identical.
Any inconsistencies were discussed with the fourth author until agreement was reached. The coding was guided by the research question on life experiences that play a role in shaping appropriate IYCF practices. Potential themes were formed from initial codes and reviewed for their relevance in relation to the coded extracts and entire dataset. Examples of these themes are key learning moments and life experiences through the life course highlighted by participants as significantly influencing appropriate feeding practices. The findings were further discussed and challenged in a series of team meetings involving all authors, from which further analytical refinements emerged.

| Ethical considerations
The study was approved by the institutional review board of the College of Medicine and Health Sciences in Rwanda (approval notice: No. 058/CMHS IRB/2016). Written informed consent was obtained from all participants.

| RESULTS
This section starts with a brief description of the participants' characteristics, followed by key moments or events on the timelines as indicated by participants. Additionally, as represented in Figure 1, the section elaborates on past food upbringing experiences, the influence of motherhood as a major life course transition and the learning from others through positive interactions.

| Key moments or events on the timelines as indicated by participants
All participants indicated key moments (events) and experiences during their life course that influenced their food practices. Most timelines started from the participants' early childhood until motherhood. The narrated moments or events varied from one participant to another. Important moments or events reported as influencing food practices included • death of mother, father or both parents; • living with grandmother or stepmother; • responsibility as family cook at young age; • taking care of older sisters' or neighbours' babies at young age; • sharing meals with siblings or parents; • parental examples of preparing foods, both positive (e.g. variety of foods, including vegetables) and negative (monotonous meals); and • life course transitions such as getting married, pregnancy and childbirth.
Both positive and negative life experiences and events played a role in shaping the participants' IYCF practices. For instance, childhood events such as the death of the father or the mother and being raised by stepmother gave rise to stressful situations because they used to eat monotonous diets and sometimes with food scarcity during childhood. However, later in their life course, some mothers managed to convert those life course constraints into learning opportunities that motivated them to work very hard to be able to feed their children differently.

| Past food upbringing experiences
The influence of dietary practices learnt early in the life course was one major recurring subject throughout most of the interviews.
Most participants reported remembering their upbringing with respect to food and emulating some of their parents' food habits as well as what they learned from their upbringing in terms of foods.

| Major life course transition: Motherhood
Participants discussed how they became more aware of the impor- Three major insights emerge from our study, related to (1) past food upbringing experiences, (2) learning from others and (3) transition to motherhood (see Figure 1). Below, we discuss these findings within the framework of the SMH, more specifically in relation to the three dimensions of the SOC: comprehensibility, manageability and meaningfulness.
Comprehensibility vegetables trajectories that found that early experiences provide lasting food roots that provide reference points for later comparison (Devine, Connors, Bisogni, & Sobal, 1998). For some participants, unpleasant food experiences during childhood triggered them to do things differently during motherhood for the benefits of their children; that is, they managed to convert life course constraints into learning opportunities. The SMH postulates that people are exposed to events that may be considered as stressors and affect health as they can reduce health temporarily but can also in the long term strengthen people in a way that makes it possible to deal with stress in another situations (Eriksson, 2017). In a study to unravel the mechanisms underlying healthful eating, Swan et al. (2018), concluded that it is never too late to promote healthful eating, because even in the face of adverse experiences, challenges could be overcome later in life. shown that social support had influence on individuals' past and current ways of managing dietary practices (Bisogni, Jastran, Shen, & Devine, 2005). This finding implies that dietary behaviours, for instance, IYCF decisions and actions, are learnt, supported and embedded in the social environment (Ahishakiye et al., 2019).
Meaningfulness is the motivational dimension of SOC. Life experiences that evoked meaningfulness-a belief that there is a good reason to maintain or care about their IYCF practices-included participants' sense of responsibility over their children's health and awareness of the benefits of optimal IYCF practices during motherhood.
Becoming a mother (motherhood) triggered an increased sense of responsibility over children's health and nutrition and increased their awareness of the benefits of optimal IYCF practices. This made mothers to be more internally motivated. With this internal motivation, healthy food practices, including IYCF practices, become more meaningful to mothers because they are important for the good health of their children. Previous studies found that the health of the child is the most important motive for mothers to manage healthy food practices (Edvardsson et al., 2011;Hingle et al., 2012). Or, as Aschemann-Witzel (2013)  the learning process about appropriate practices of their children, especially positive food upbringing experiences. Also, learning from others (role models) contributes to this. This is in line with a study by Bull, Mittelmark, and Kanyeka (2013), which found that participation in community groups of various kinds build individual women's skills and play a key role in making any activity meaningful, comprehensible and manageable. This in its turn supports mothers to develop their sense of agency and thus gain greater control of their IYCF practices. Social support, from both family and community members, provides an individual with coherent life experiences (Antonovsky, 1979). Finally, sense of responsibility, triggered by motherhood, is a powerful experience. As Szwajcer, Hiddink, Koelen, and van Woerkum (2007) argue, those periods are occasions when women become more aware of the health aspects of nutrition and are more motivated to eat healthy. It is the meaningfulness of the life event or experience (IYCF practices in this case) that determine the person's comprehension as well as their willingness to invest resources to succeed (Thomson & Dykes, 2011). Once individuals are motivated, goal-directed activities are initiated and maintained (Cook & Artino, 2016), and through those activities, they also learn to identify and use certain resources (GRRs).

| Strengths and limitations
One of the strengths of this study was the use of a visual research tool of timelining. The use of timelining elicited self-reflection and story-telling among study participants (Sheridan, Chamberlain, & Dupuis, 2011). In such conditions, the interviews were led by what participants marked as important aspects of their past life experiences without interference from the researchers.
This study has some limitations as well. First, the study was limited to a specific group of mothers, from one rural geographical location in Muhanga District, with a minimal sample size (n = 14).
Therefore, the generalizability of the findings to a different or wider population is low. However, generalizability of the findings was not of primary importance as this study rather aimed to obtain richer and indepth accounts in the case studied, which would not have been achievable with a large sample. Second, only mothers who followed the recommended IYCF practices and whose children were growing well were included in the study. Future research could explore similar investigations by considering all mothers of the larger study and classify them into categories based on the criteria of IYCF and child growth (meeting IYCF criteria, meeting growth criteria, meeting IYCF and growth criteria and meeting no criteria) to analyse the different experiences and try to see where patterns emerge. Third, besides nutrition-related factors like feeding practices, other factors such as genetic factors are related to child growth in height (Addo et al., 2013). Future research should consider mother's height while determining children who are growing well. Also, child's characteristics such as birth order are known to influence breastfeeding and complementary feeding practices (Dhami, Ogbo, Osuagwu, & Agho, 2019;Shiferaw, Mossa, & Gashaw, 2017). However, there are conflicting findings with regard to the consistency of the associations from one setting to another. In our study, the sample size was too small to explore its influence on mothers' feeding practices. Fourth, the questions were mainly focused on nutrition rather than on life in general.
Consequently, how women cope in general and how this relates to their IYCF practices were not explored. Future research should explore these questions to gain insights into more general healthpromoting factors along the life course that generates healthy IYCF practices.

| Implications
The findings of this study have a number of practical implications for nutrition promotion professionals and programmes: • Tailoring food and nutrition advice to the complexity of mothers' life course experiences by creating opportunities for mothers to reflect on their lived experiences and the role of past life experiences on how they deal with their daily lives to manage well IYCF practices (mothers-driven learning) instead of the one-size-fits-all educational approach is fundamentally important.
• Health professionals should also focus on the social environment in which mothers live, because the interaction of the mothers with their social environment is important for appropriate IYCF practices. Facilitating food-and nutrition-directed learning through community actions, including community cooking classes where mothers can practice appropriate practices, is an example.