Menu
Menu
Menu
Menu
Background: Stunted physical growth during early childhood is a marker of chronic undernutrition, and the adverse life circumstances that underlie it. These have the potential to disrupt normal brain development and the acquisition of foundational cognitive, language, social and motor skills. Stunting is prevalent in most low-and middle-income countries. Because the prevention of stunting requires large-scale structural and attitudinal changes, several psycho-educational interventions have been developed to mitigate the adverse association between early stunting and skill development. However, the resource-intensive nature of custom-designed interventions limit their sustainability and scalability in resource-limited settings. This study explored the possibility that available resources that promote positive development (existing preschool education programs, and no- or low-cost home-based learning activities and resources) may protect against any negative association between stunting and the acquisition of foundational skills required for academic learning and adaptation at school. Method: Data for 36-to 59-month-old children (n = 3,522; M = 46.7 months; 51.2% male; 74.1% rural) were drawn from the most recent Multiple Indicator Cluster Survey conducted in Côte d’Ivoire (MICS5, 2016). Stunting was assessed using the WHO Child Growth Standards. Preschool attendance and home learning activities and resources were assessed by maternal report. School readiness was assessed using the 8-item form of the Early Child Development Index (ECDI). Results: A high percentage of children met the criteria for stunting (28.5%; 19.7% moderate; 8.8% severe). There were marked urban–rural differences in the prevalence of stunting, rates of preschool attendance, home learning activities and resources, children’s school readiness scores, and the relationships between stunting, the protective factors and school readiness scores. These urban–rural differences in ECDI scores could be fully explained by differences between these settings in stunting and the protective factors. However, only two protective factors (access to books and home-based activities that promote learning) made independent contributions to variance in ECDI scores. There was tentative evidence that stunted children whose homes provided highly diverse learning activities and multiple types of learning resources were more likely than those who did not to have a high level of school readiness. Conclusion: Capitalizing on the existing practices of families that show positive deviance in caregiving may provide a basis for culturally appropriate, low-cost interventions to improve school readiness among children in low- and middle-income countries, including children with stunted growth.
This research draws on data collected during the latest Multiple Indicator Cluster Survey (MICS5) conducted in Côte d’Ivoire. This survey was designed to monitor progress toward the targets of the Sustainable Development Goals and the National Development Plan for 2016–2020. The survey selected a nationally representative sample of households using cluster sampling in a two-stage probability sampling design. A subsample of 512 of the 23,484 enumeration areas from the 2014 General Population and Housing Census was selected. Then, a sample of 25 households was recruited in each area. Data were collected between April and July 2016. In total, 11,879 of the 12,768 selected households responded to the main survey (96.6% participation rate). During data collection, face-to-face interviews were conducted with mothers or primary carers of children under 5 years of age (98.2% participation rate). Trained interviewers recorded the mothers’/caregivers’ responses and mothers’ and children’s anthropometric measurements on tablets. MICS5 collected data on 3,730 children aged 36–59 months of age. The planned statistical analyses assume that datapoints are independent. This assumption is violated if a sample includes multiple children born to the same mother and living in the same household. Solutions that involve statistical analyses that cluster children within families face two challenges. First, within-family differences in the parenting practices and the investment of resources directed to individual children (based on their gender, birth order and other factors) are very common. Second, these within-family differences often show little uniformity in multi-cultural populations. Therefore, the current study ensured the independence of data points by randomly selecting one child per mother for inclusion in the analyses. The final sample consisted of 3,522 children (94.4% of the original sample), most of whom (74.1%) lived in rural settings (Table 1). Demographic characteristics of the samples of urban and rural 36- to 59-month-old children in Cote d’Ivoire. All interviews with mothers/primary caregivers and instructions to children during anthropometric measurements were provided in French or in the participants’ local language. In about 90% of cases, children’s age could be accurately determined from their immunization records. When these was not available children’s date of birth was reported by the mother/caretaker. Children’s sex and ethnicity were assessed by single items. Data concerning the mothers’ education, the number and age of other children, the flooring material of the dwelling, and the household’s access to electricity and improved sanitation facilities were gathered through observations by interviewers and/or parent reports. Mothers were asked to indicate the highest level of education they had attended (did not attend primary school = 0; primary school = 1; secondary school or post-secondary education = 2). This wording prevented conclusions about the highest levels of education mothers had completed. The economic circumstances of households were assessed using a wealth index calculated on the basis of a principal components analysis of data about the ownership of consumer goods, dwelling characteristics, and other characteristics. Separate factor scores were calculated for households in urban and rural areas. Each household in these two samples was then assigned a wealth score based on the assets owned by that household and on the final factor scores. Finally, wealth scores in each setting were ranked and grouped into five categories of equal size (quintiles: poorest = 1; richest = 5). Further information on the construction of the wealth index is provided by Rutstein (79). Escherichia coli (e. coli) contamination of household drinking water at its point-of-use was assessed using samples that were collected by asking for “a glass of water that members of the household would drink”. A sample of 100 ml of water was filtered through a 0.45 micron filter paper that could retain any E. coli. The filter was then placed on a growth medium plate. A second sample of 1 ml of water was pipetted directly onto a different growth medium plate. To encourage the growth of E. coli, both plates from a given household were stored in custom-designed incubation belts carried by the fieldwork staff, or in electric incubators. After a 24-h period, colonies of E. coli (blue) and non-E. coli coliforms (purple/red) that were detectable by eye were counted. The risk of fecal contamination in 100 ml of drinking water was classified as low (0 colonies), moderate (1–10 colonies), high (11–100 colonies) or very high (more than 100 colonies) (80). Children’s place of residence was classified as urban or rural on the basis of population size using the criteria from the 2014 General Census of Population and Housing. Urban and rural settings were sampled in each of the eleven administrative districts in Côte d’Ivoire (except the district for the city of Abidjan, which contains no rural areas). Children’s standing height without shoes or socks was measured in accordance with World Health Organization recommendations (81) by trained field assessment teams using a portable stadiometer with an accuracy of 1 mm. Z-scores for height-for-age were interpreted using the WHO Child Growth Standards (81). Stunting was classified as moderate (height-for-age z-score under −2, but not under −3) or severe (height-for-age z-score <-3). Z-scores below −5 (n = 17) and above +5 (n = 1) were removed from the dataset because these are implausible values. The prevalence of stunting was classified according to the thresholds set by the WHO-UNICEF Technical Expert Advisory Group on Nutrition Surveillance: very low (under 2.5%); low (2.5–<10%); medium (10–<20%); high (20–<30%); very high (over 30%) (82). The mother/caretaker was asked whether the child attended a public or private pre-school. Attendance at either type of pre-school was coded as 1, and attendance at neither was coded as 0. Out-of-home childcare without an educational orientation was excluded. The Family Care Indicators inventory (83) asked whether the child's mother, father, or another adult (person over 15 years of age) had engaged in each of six activities with the target child within the previous 3 days: “read books or looked at picture books with your child”; “told stories to your child”; “sang songs with your child”; “took your child outside the home place”; “played games with your child”; and “spent time with your child in naming things or counting things or drawing”. Responses were integrated to assess whether any adult had engaged the child in each of the activities. A composite score was created by summing positive responses across the six activities (range 0–6). The inventory has good reliability, concurrent validity and equivalence across low- and middle-income countries, including Côte d'Ivoire (84). In the current study, the scale had adequate internal consistency in both rural and urban settings (Cronbach alpha of 0.72 and 0.71, respectively). Scores of 4 or more are considered to indicate “adequate” home stimulation (47, 85). Families' investment in home learning resources was assessed using three questions. The first asked how many children's books (use requires caregivers to be literate) or picture books (use does not require caregivers to be literate) the target child had. The second and third questions were: “I am interested in learning about the things that (name) plays with at home. Does he/she play with home-made toys (such as dolls, cars or other toys made at home)? Toys from a shop or manufactured toys?” Because data on access to books was not expected to have a normal distribution, a dichotomous variable was created (one or more books = 1; no books = 0). Dichotomous data from the two questions about toys were integrated to create a single variable (one or more home-made or manufactured toys =1; no toys = 0). The 3-domain form of the Early Childhood Development Index (ECDI) (86) was used as a proxy measure of school readiness. This caregiver-reported checklist consists of dichotomously scored items that assess skills relevant for 3- and 4-year-old children. The 8 items assess literacy and numeracy (3 items; e.g., “Can (name) identify or name at least ten letters of the alphabet?”), social-emotional skills (3 items; e.g., “Does (name) get along well with other children?”) and learning (2 items; “Does (name) follow simple instructions to do something correctly?”). After reverse-scoring the two negatively worded items, the number of positive responses was summed (range 0 to 8). Previous research has also used this 3-domain composite score (87, 88). However, the wide range of child ages, small number of items and dichotomous scoring of items limit the measure's sensitivity (87, 89). For example, children who can name 9 letters and recognize the symbols for 9 numbers receive the same score (0) on these items as children who cannot name any letters or recognize any numbers.
N/A
