progress to address malnutrition has typically been better when policies and programs address such determinants simultaneously, accompanied by economic growth [2].
Fig. 3. Conceptual framework on the determinants of child growth. Developed by the authors building on previously published frameworks [27, 28]. IUGR, Intra-uterine growth retardation; NCDs, Non-communicable diseases.
Stunting
Building on prior frameworks [27, 28], Figure 3 shows known determinants of child stunting organized along a simplified pathway. Being born small, unimproved sanitation, and diarrhea are important risk factors for stunting globally and particularly in South Asia [29]. Low maternal height and household wealth are also important predictors in South Asia [30]. The relationships between stunting and its determinants as well as their relative importance can vary substantially by context, even within a single country, as illustrated previously in India [27]. Within-country diversity in the drivers of stunting has also been documented in Vietnam [28]. Effective action to accelerate progress to address stunting may require interventions tailored to address these determinants, the composition of which may vary by geography or other factors. Path analyses of linear growth faltering using longitudinal data can also help identify where and how to intervene across the range of basic to immediate causes of stunting [31]. While data may be limiting in some contexts, in others, existing data have not been used to its full potential to document and understand the many determinants of child malnutrition and their variability by geographic or other factors like those from India and Vietnam. Longitudinal datasets are particularly powerful for understanding determinants but are rare and usually do not have sufficient geographically representation to explore variability within countries.
Anemia
As noted above, anemia continues to be the only indicator tracked consistently at a national level and used as a proxy for micronutrient status, and it is often interpreted as synonymous with iron deficiency. The determinants of anemia, however, include a complex pattern of biological, infectious, environmental, and genetic factors [32]. Anemia may be related to inadequate intakes of iron, folic acid, vitamin A, and/or vitamin B12. Conditions that cause increased loss or reduced absorption or utilization of iron such as soil-transmitted helminths, malaria, schistosomiasis, and other parasitic infections may also result in anemia, even if intakes appear adequate. Finally, genetic disorders that affect iron or hemoglobin metabolism also result in anemia.
Particularly in low- and middle-income countries, several risk factors may simultaneously cause anemia in an individual. For instance, in sub-Saharan Africa, malaria and iron deficiency are both highly prevalent. In these cases, common approaches to prevent and treat anemia, such as iron supplementation, may not be effective, and recent evidence suggests that iron supplementation may even exacerbate infection. In a study of children living in malaria-endemic Tanzania, prophylactic iron supplementation increased the risk of hospitalization, death, malaria, and other infections among those who were not iron deficient [33]. Consumption of iron-containing multiple micronutrient powders also resulted in increased diarrhea in a study in Pakistan [34] and increased hospitalizations in Ghana [35]. Thus, understanding the etiology of anemia and a balanced assessment of the potential risks and benefits of iron supplementation should be a part of program planning [36].
Fig. 4. Schematic illustration of the nutritional and nonnutritional causes of anemia in children and women. Developed by the authors based on results of a path analysis of anemia determinants in Uttar Pradesh, India [38].
In India, the prevalence of anemia is very high with little progress made over the past decade [37]; a better understanding of the etiology of anemia in context could help inform more appropriate approaches to its reduction. A recent study in Uttar Pradesh, India, examined the etiology of anemia in women and children using a state-representative survey [38]. This comprehensive study measured a range of genetic, environmental, infectious, and nutritional risk factors for anemia (Fig. 4). Results indicate that iron was an important contributor to anemia; 78% of children and 74% of women with anemia had iron deficiency anemia. Other direct and indirect determinants included inflammation, folic acid, vitamin B12, and vitamin A (serum retinol), as well as more distal factors, such as dietary diversity, household hunger, and water, sanitation, and hygiene, which work through effects on status indicators. In this population without malaria (<1%) and minimum intestinal parasites (<9%), iron and inflammation were strong determinants of hemoglobin concentration and anemia. The relationship between iron and inflammation, however, is not straightforward, and the extent to which iron and other nutritional interventions can be fully effective to address anemia is not clear. Nonnutritional causes of anemia may inhibit a response to iron interventions if not concurrently addressed. The type of analysis conducted with the data from Uttar Pradesh should be used to guide effective programs in populations where the contribution of nutritional and nonnutritional determinants to anemia prevalence may vary, such as settings with malaria or high proportions of genetic abnormalities.
Fig. 5. The ecological model applied to determinants of obesity in children living in First Nations communities in Canada, reproduced with permission from [41].
Overweight
There is still much to learn about the determinants and potential actions to prevent and control childhood overweight. As with stunting and anemia, the etiology of excess weight in childhood is complex and includes interactions among genetic makeup, intrauterine factors, the home and built environment, and behavioral factors [39, 40]. The interrelationship among these factors and how they may influence child weight is nicely illustrated for First Nations communities in Canada [41]. In this context, historical factors and persistent inequities dramatically influence many social and environmental determinants of overweight. An ecological model that frames