Quantifying the association of natal household wealth with women’s early marriage in Nepal

Background Women’s early marriage (<18 years) is a critical global health issue affecting 650 million women worldwide. It is associated with a range of adverse maternal physical and mental health outcomes, including early childbearing, child undernutrition and morbidity. Poverty is widely asserted to be the key risk factor driving early marriage. However, most studies do not measure wealth in the natal household, but instead, use marital household wealth as a proxy for natal wealth. Further research is required to understand the key drivers of early marriage. Methods We investigated whether natal household poverty was associated with marrying early, independently of women’s lower educational attainment and broader markers of household disadvantage. Data on natal household wealth (material asset score) for 2,432 women aged 18–39 years was used from the cluster-randomized Low Birth Weight South Asia Trial in lowland rural Nepal. Different early marriage definitions (<15, <16, <17 and <18 years) were used because most of our population marries below the conventional 18-year cut-off. Logistic mixed-effects models were fitted to estimate the probabilities, derived from adjusted Odds Ratios, of (a) marrying at different early ages for the full sample and for the uneducated women, and (b) being uneducated in the first place. Results Women married at median age 15 years (interquartile range 3), and only 18% married ≥18 years. Two-thirds of the women were entirely uneducated. We found that, rather than poverty, women’s lower education was the primary factor associated with early marriage, regardless of how ‘early’ is defined. Neither poverty nor other markers of household disadvantage were associated with early marriage at any age in the uneducated women. However, poverty was associated with women being uneducated. Conclusion When assets are measured in the natal household in this population, there is no support for the conventional hypothesis that household poverty is associated with daughters’ early marriage, but it is associated with not going to school. We propose that improving access to free education would both reduce early marriage and have broader benefits for maternal and child health and gender equality.

Association between supportive supervision and performance of community health workers in India: a longitudinal multi-level analysis

Introduction Community health workers (CHWs) deliver services at-scale to reduce maternal and child undernutrition, but often face inadequate support from the health system to perform their job well. Supportive supervision is a promising intervention that strengthens the health system and can enable CHWs to offer quality services. Objectives We examined if greater intensity of supportive supervision as defined by monitoring visits to Anganwadi Centre, CHW-supervisor meetings, and training provided by supervisors to CHWs in the context of Integrated Child Services Development (ICDS), a national nutrition program in India, is associated with higher performance of CHWs. Per program guidelines, we develop the performance of CHWs measure by using an additive score of nutrition services delivered by CHWs. We also tested to see if supportive supervision is indirectly associated with CHW performance through CHW knowledge. Methods We used longitudinal survey data of CHWs from an impact evaluation of an at-scale technology intervention in Madhya Pradesh and Bihar. Since the inception of ICDS, CHWs have received supportive supervision from their supervisors to provide services in the communities they serve. Mixed-effects logistic regression models were used to test if higher intensity supportive supervision was associated with improved CHW performance. The model included district fixed effects and random intercepts for the sectors to which supervisors belong. Results Among 809 CHWs, the baseline proportion of better performers was 45%. Compared to CHWs who received lower intensity of supportive supervision, CHWs who received greater intensity of supportive supervision had 70% higher odds (AOR 1.70, 95% CI 1.16, 2.49) of better performance after controlling for their baseline performance, CHW characteristics such as age, education, experience, caste, timely payment of salaries, Anganwadi Centre facility index, motivation, and population served in their catchment area. A test of mediation indicated that supportive supervision is associated indirectly with CHW performance through improvement in CHW knowledge. Conclusion Higher intensity of supportive supervision is associated with improved CHW performance directly and through knowledge of CHWs. Leveraging institutional mechanisms such as supportive supervision could be important in improving service delivery to reach beneficiaries and potentially better infant and young child feeding practices and nutritional outcomes. Trial registration : Trial registration number: https://doi.org/10.1186/ISRCTN83902145

Cross-feeding between intestinal pathobionts promotes their overgrowth during undernutrition

Child undernutrition is a global health issue associated with a high burden of infectious disease. Undernourished children display an overabundance of intestinal pathogens and pathobionts, and these bacteria induce enteric dysfunction in undernourished mice; however, the cause of their overgrowth remains poorly defined. Here, we show that disease-inducing human isolates of Enterobacteriaceae and Bacteroidales spp. are capable of multi-species symbiotic cross-feeding, resulting in synergistic growth of a mixed community in vitro. Growth synergy occurs uniquely under malnourished conditions limited in protein and iron: in this context, Bacteroidales spp. liberate diet- and mucin-derived sugars and Enterobacteriaceae spp. enhance the bioavailability of iron. Analysis of human microbiota datasets reveals that Bacteroidaceae and Enterobacteriaceae are strongly correlated in undernourished children, but not in adequately nourished children, consistent with a diet-dependent growth synergy in the human gut. Together these data suggest that dietary cross-feeding fuels the overgrowth of pathobionts in undernutrition.

Urban-rural differences in the associated factors of severe under-5 child undernutrition based on the composite index of severe anthropometric failure (CISAF) in Bangladesh

Introduction Severe undernutrition among under-5 children is usually assessed using single or conventional indicators (i.e., severe stunting, severe wasting, and/or severe underweight). But these conventional indicators partly overlap, thus not providing a comprehensive estimate of the proportion of malnourished children in the population. Incorporating all these conventional nutritional indicators, the Composite Index of Severe Anthropometric Failure (CSIAF) provides six different undernutrition measurements and estimates the overall burden of severe undernutrition with a more comprehensive view. This study applied the CISAF indicators to investigate the prevalence of severe under-5 child undernutrition in Bangladesh and its associated socioeconomic factors in the rural-urban context. Methods This study extracted the children dataset from the 2017–18 Bangladesh Demographic Health Survey (BDHS), and the data of 7661 children aged under-5 were used for further analyses. CISAF was used to define severe undernutrition by aggregating conventional nutritional indicators. Bivariate analysis was applied to examine the proportional differences of variables between non-severe undernutrition and severe undernutrition group. The potential associated socioeconomic factors for severe undernutrition were identified using the adjusted model of logistic regression analysis. Results The overall prevalence of severe undernutrition measured by CISAF among the children under-5 was 11.0% in Bangladesh (rural 11.5% vs urban 9.6%). The significant associated socioeconomic factors of severe undernutrition in rural areas were children born with small birth weight (AOR: 2.84), children from poorest households (AOR: 2.44), and children aged < 36 months, and children of uneducated mothers (AOR: 2.15). Similarly, in urban areas, factors like- children with small birth weight (AOR: 3.99), children of uneducated parents (AOR: 2.34), poorest households (APR: 2.40), underweight mothers (AOR: 1.58), mothers without postnatal care (AOR: 2.13), and children’s birth order ≥4 (AOR: 1.75), showed positive and significant association with severe under-5 undernutrition. Conclusion Severe undernutrition among the under-5 children dominates in Bangladesh, especially in rural areas and the poorest urban families. More research should be conducted using such composite indices (like- CISAF) to depict the comprehensive scenario of severe undernutrition among the under-5 children and to address multi-sectoral intervening programs for eradicating severe child undernutrition.

Age-appropriate feeding practices and their association with undernutrition among children aged 6–23 months in aspirational districts of India: a multinomial analysis

‘Health and nutrition’ is one among the five areas covered by the Aspirational District Programme in India, which aims to achieve the Sustainable Development Goals (SDGs). The reduction of undernutrition in under-five children has remained a major focus of the SDGs, especially at the ages of 6–23 months as this affects child development. This study used National Family Health Survey 2015–16 data to examine appropriate feeding practices and their associations with undernutrition among children aged 6–23 months in the 124 aspirational districts of India. Multinomial logistic regression analysis was used to analyse the association between feeding practices and undernutrition, adjusting for covariates. A total of 13,851 children aged 6–23 months were included in the analysis. Child nutritional outcomes, and children receiving the recommended minimum dietary diversity (MDD), minimum meal frequency (MMF) and minimum acceptable diet (MAD), were poorer in the aspirational compared with non-aspirational districts. However, the proportions of children who continued to breastfed, i.e. currently breastfeeding and the proportion of children who were receiving appropriate breastfeeding, i.e. receiving complementary feeding, in addition to breast milk, were higher in the aspirational districts. Appropriate breastfeeding and MDD were found to be associated negatively with undernutrition. While the continuation of breastfeeding increased the odds of children being undernourished, appropriate breastfeeding lowered the odds. The significant predictors of undernourishment among the study children were the child being male, of higher birth order, older and of smaller than average birth size; mother’s lower educational level, mother’s lower BMI of mothers and being a teenage mother; and poor household drinking water, sanitation facilities and lower economic status. This study suggests that educating mothers, especially illiterate and poor mothers, about appropriate breastfeeding and dietary diversity could help prevent and reduce child undernutrition in the aspirational districts of India.

Early childhood undernutrition, preadolescent physical growth, and cognitive achievement in India: A population-based cohort study

Background There is a lack of nationally representative estimates for the consequences of early child undernutrition on preadolescent outcomes in India. Understanding this relationship is helpful to develop interventions that not only prevent child undernutrition but also mitigate its consequences. Methods and findings In this cohort study, we analyzed prospectively gathered data from 2 waves of the India Human Development Survey (IHDS) to investigate the association of undernutrition during early childhood (0 to 5 years) in 2004 to 2005 with physical and cognitive outcomes during preadolescent (8 to 11 years) years in 2011 to 2012. These surveys interviewed 41,554 households across all 33 states and union territories in India in 2004 to 2005 and reinterviewed 83% of the households in 2011 to 2012. Primary exposure was assessed using the Composite Index of Anthropometric Failure (CIAF) based on 2004 to 2005 survey. Primary outcomes were short stature (height-for-age z-score [HAZ] <−2), thinness (body mass index [BMI] <18.5 kg/m2), reading, and arithmetic skills during preadolescence based on the 2011 to 2012 survey. Survey-weighted generalized linear models were used, and effect modification based on child sex and sociodemographic variables were evaluated using 3-way interaction terms. Of the 7,868 children included in this analysis, 4,334 (57.3%) were undernourished. Being undernourished was associated with increased odds of short stature (odds ratio [OR] 1.73, 95% confidence interval [CI] 1.45 to 2.06) and thinness (OR 1.52, 95% CI 1.33 to 1.73) during the preadolescent period, while it was associated with decreased odds of achieving a higher reading (cumulative odds ratio [cumOR]: 0.76, 0.66 to 0.87) and arithmetic (cumOR: 0.72, 0.63 to 0.82) outcomes. The disparity in outcomes based on CIAF increased with age, especially for girls. Increased level of female education within the household reduced the disadvantages of undernutrition among female children. Study limitations include observational and missing data, which limit our ability to draw strong causal inferences. Conclusions In this study, we found that early child undernutrition was associated with several adverse preadolescent physical and cognitive outcomes, especially among female children. Improved female education mitigates this association. Female education promotion should assume a central role in Indian public health policy making.

A machine learning classifier approach for identifying the determinants of under-five child undernutrition in Ethiopian administrative zones

Background Undernutrition is the main cause of child death in developing countries. This paper aimed to explore the efficacy of machine learning (ML) approaches in predicting under-five undernutrition in Ethiopian administrative zones and to identify the most important predictors. Method The study employed ML techniques using retrospective cross-sectional survey data from Ethiopia, a national-representative data collected in the year (2000, 2005, 2011, and 2016). We explored six commonly used ML algorithms; Logistic regression, Least Absolute Shrinkage and Selection Operator (L-1 regularization logistic regression), L-2 regularization (Ridge), Elastic net, neural network, and random forest (RF). Sensitivity, specificity, accuracy, and area under the curve were used to evaluate the performance of those models. Results Based on different performance evaluations, the RF algorithm was selected as the best ML model. In the order of importance; urban–rural settlement, literacy rate of parents, and place of residence were the major determinants of disparities of nutritional status for under-five children among Ethiopian administrative zones. Conclusion Our results showed that the considered machine learning classification algorithms can effectively predict the under-five undernutrition status in Ethiopian administrative zones. Persistent under-five undernutrition status was found in the northern part of Ethiopia. The identification of such high-risk zones could provide useful information to decision-makers trying to reduce child undernutrition.