Menu
Menu
Menu
Menu
Background: Improving child nutritional status is an important step towards achieving the Sustainable Development Goals 2 and 3 in developing countries. Most child nutrition interventions in these countries remain variably effective because the strategies often target the child’s mother/caregiver and give limited attention to other household members. Quantitative studies have identified individual level factors, such as mother and child attributes, influencing child nutritional outcomes. Methods: We used a qualitative approach to explore the influence of household members on child feeding, in particular, the roles of grandmothers and fathers, in two Nairobi informal settlements. Using in-depth interviews, we collected data from mothers of under-five children, grandmothers, and fathers from the same households. Results: Our findings illustrate that poverty is a root cause of poor nutrition. We found that mothers are not the sole decision makers within the household regarding the feeding of their children, as grandmothers appear to play key roles. Even in urban informal settlements, three-generation households exist and must be taken into account. Fathers, however, are described as providers of food and are rarely involved in decision making around child feeding. Lastly, we illustrate that promotion of exclusive breastfeeding for 6 months, as recommended by the World Health Organization, is hard to achieve in this community. Conclusions: These findings call for a more holistic and inclusive approach for tackling suboptimal feeding in these communities by addressing poverty, targeting both mothers and grandmothers in child nutrition strategies, and promoting environments that support improved feeding practices such as home-based support for breastfeeding and other baby-friendly initiatives.
The study was conducted in two Nairobi informal settlements: Korogocho and Viwandani. The sites are located approximately 10 km from the city centre and about 7 km from each other (Emina et al., 2011). Each occupies about 0.5 km2 and is densely populated: 30,736 residents in Korogocho and 52,852 residents in Viwandani (African Population and Health Research Center, 2017). The sites are characterized by high unemployment, poverty, and poor access to facilities such as water, sanitation, and health care services when compared with Nairobi as a whole (APHRC, 2002). However, there are few differences between the two sites. For instance, Viwandani is bordered by an industrial area and attracts a youthful and highly mobile population, whereas in Korogocho, the population is more stable, about a quarter of the residents aged 12 years and above were born in this informal settlement (Emina et al., 2011). A variety of ethnic groups live in the sites, specifically Kikuyu, Luhya, Luo, Kamba, and Somali, with different cultural norms and practices with regard to child development including feeding. The study targeted mothers of children aged 12–59 months and family influencers such as grandmothers and fathers. The participants were selected purposively from a 2006–2013 Maternal and Child Health project nested within the Nairobi Urban Health and Demographic Surveillance System (NUHDSS). The NUHDSS collects birth, death, and migration data every 4 months, and the Maternal and Child Health study collected child anthropometric measurements. From these data, households with stunted and not stunted under‐five children were identified. Community health workers in the study area helped identify the selected households based on their NUHDSS unique identifier. In total, 30 in‐depth interviews were conducted in the two sites: 12 with mothers of stunted children, eight with mothers of nonstunted children, six with grandmothers, and four with fathers, as shown in Table 1. Respondents were interviewed if they met three inclusion criteria: (a) mothers, grandmothers, and fathers of at least one child aged 12–59 months present at the time of interview, (b) living in one of the study sites since birth of the last child, and (c) willing to participate in the study. In most cases, mother, father, and grandmother from the same household were interviewed. Data collection took place in September 2016 using three different interview guides (one each for mothers, grandmothers, and fathers). The study collected information on respondents’ knowledge on child stunting, practices and beliefs on child feeding, and child health care. The interviews were all conducted in Kiswahili by two experienced research assistants after intensive training by the study principal investigator into the purpose of the study and the instruments to be used. The interview guides were pretested on respondents with similar background to the study participants. Field supervision was done by the principal investigator. Number of in‐depth interviews by type of respondent and study site The interviews were audio‐recorded then transcribed before translation into English. Using a deductive coding approach, the analysis was done using Nvivo 10 and guided by the conceptual framework on child health described in Figure 1 to explore the household level factors underlying child health and feeding practices. The coding approach was informed by findings from previous research conducted on child growth in the same settings, in particular, household level factors that were related to child linear growth (Faye et al., 2019). We explored respondents’ knowledge on child stunting (perceptions and perceived causes), practices around child feeding in the household, decision making on child feeding (who decides on what the child should eat and when), and child health and care (what to do when a child is identified or perceived to be stunted and who takes the decision). Consistency checks were applied during the analysis and coding to ensure good understanding and interpretation of the data. The consistency checks consisted of triangulation (utilizing two analysts to review the data and findings) and member checking where each respondent was given a chance to confirm his or her views and statements before leaving the interview venue. Thematic analysis was used to make sense of the data (Grbich, 2012; Vaismoradi, Turunen, & Bondas, 2013). The study was approved by the African Medical and Research Foundation Ethics and Scientific Review Committee in Kenya and the Human Research Ethics Committee (Medical) at the University of Witwatersrand in South Africa. All interviews were conducted in private, and written informed consent was sought from all participants.
N/A
