Causes and consequences of child growth faltering in low-resource settings

Growth faltering, encompassing stunting (low length for age) and wasting (low weight for length), significantly impacts child health and survival in low-resource settings. Globally, a substantial portion of children under five years old experience stunting and wasting, primarily in low- and middle-income countries (LMICs). These conditions contribute to significant mortality and are associated with impaired cognitive development and reduced economic outcomes in adulthood. Despite the recognized global health significance, preventive interventions in LMICs have achieved limited success. Numerous nutritional interventions targeting various life stages have shown benefits, including nutritional supplementation during pregnancy and breastfeeding promotion. However, postnatal interventions and those targeting children after the initiation of complementary feeding have had minimal impact at the population level, highlighting a need for a more comprehensive understanding of the causes and timing of growth faltering. This study addresses this knowledge gap by analyzing a large dataset to identify key causal factors and the age windows during which interventions may be most effective.

Existing literature highlights the detrimental effects of growth faltering on child health, with stunting and wasting contributing to substantial mortality and morbidity in LMICs. Numerous studies have explored the effectiveness of various nutritional interventions, including prenatal and postnatal supplementation, breastfeeding promotion, and improved complementary feeding practices. While some positive effects have been observed, the overall population-level impact has been limited. Furthermore, research on water, sanitation, and hygiene (WASH) interventions has shown inconsistent effects on child growth. The modest success of interventions suggests an incomplete understanding of optimal strategies and timing, prompting a renewed focus on combining rich data sources with advanced statistical methods to elucidate the causes of growth faltering and their temporal relationships. Studies have revealed high rates of incident stunting and wasting within the first three months of life, emphasizing the critical period in early childhood.

This research utilized a population intervention effects (PIE) analysis of 33 longitudinal cohorts encompassing 83,671 children and 662,763 anthropometric measurements. Data were collected from 15 LMICs across South Asia, sub-Saharan Africa, Latin America, and Eastern Europe between 1987 and 2014. The cohorts were part of the Bill & Melinda Gates Foundation's Knowledge Integration (Ki) initiative, focusing on growth and development during the first 1,000 days of life. Five inclusion criteria were applied to select cohorts, ensuring a robust evaluation of growth faltering in representative populations. The primary outcomes included length-for-age z-score (LAZ), weight-for-length z-score (WLZ), weight-for-age z-score (WAZ), stunting, wasting, underweight, and length and weight velocities. Associations between early growth faltering and subsequent growth faltering or mortality were also assessed. PIEs were estimated using targeted maximum-likelihood estimation (TMLE), a doubly robust, semi-parametric method, adjusting for potential confounders using ensemble machine learning. Cohort-specific parameters were estimated and pooled using random-effects models. Thirty exposures, including maternal anthropometry, child birth characteristics, postnatal factors, parental characteristics, and household/environmental conditions, were examined.

The analysis revealed that improving maternal anthropometry (height and BMI) and child birth size substantially increased LAZ and WLZ by 24 months of age. Markers of better household socioeconomic status (number of rooms, parental education, clean cooking fuel, wealth index) were also significant predictors. The population-level effect of season on WLZ was notable, with higher WLZ in drier periods. Exclusive or predominant breastfeeding before 6 months was associated with higher WLZ but not LAZ at 6 months and wasn't a major predictor at 24 months. Girls consistently had higher LAZ and WLZ than boys. The strongest predictors of stunting and wasting closely matched those for LAZ and WLZ at 24 months, suggesting that both continuous and binary measures provide similar inferences. Potential improvements through population interventions were relatively modest, for example an estimated 8.2% reduction in wasting incidence with improved maternal BMI. Maternal height strongly influenced at-birth LAZ, with similar trajectories up to 24 months regardless of maternal height. However, WLZ trajectories diverged substantially after 3-4 months depending on maternal height or BMI. Age-varying effects were observed for many exposures; socioeconomic status was associated with growth faltering after 6 months of age. First-born babies had lower WLZ at birth but showed postnatal catch-up. Persistent wasting from birth to 6 months was strongly associated with incident stunting in older children. Early growth faltering was significantly associated with later severe growth faltering. Finally, growth faltering, particularly severe wasting and stunting, significantly increased the hazard of death before 24 months.

This large-scale analysis provides crucial insights into the causes and consequences of growth faltering in LMICs. The findings emphasize the importance of prenatal exposures, particularly maternal anthropometry, in determining child growth outcomes. The relatively limited impact of postnatal interventions such as breastfeeding promotion and diarrhea reduction at the population level highlights the need for a greater focus on pre-conception and pregnancy interventions to mitigate intergenerational growth deficits. The strong association between early growth faltering and subsequent severe outcomes emphasizes the critical nature of early interventions. The age-varying effects of exposures highlight the need for tailored interventions based on specific age windows, socioeconomic factors, and child-specific vulnerabilities. The study's findings provide valuable evidence for informing policy and programmatic interventions to reduce the global burden of child growth faltering.

This study, through a comprehensive analysis of a large dataset, reinforces the critical role of maternal and prenatal factors in child growth. Early interventions focused on improving maternal health and socioeconomic conditions, particularly in the preconception and pregnancy periods, are essential for preventing growth faltering. Further research should explore cost-effective and scalable interventions targeting these critical windows and evaluate the effectiveness of multifaceted approaches addressing multiple risk factors simultaneously.

While the study's large scale and rigorous methodology are strengths, certain limitations need consideration. The mortality analysis was constrained by the availability of death data, potentially excluding neonatal deaths and lacking cause-specific mortality information. Additionally, the data primarily represent populations under surveillance, which may not fully generalize to the broader population in LMICs. The generalizability could also be affected by the inclusion criteria, resulting in the overrepresentation of particular populations.