Precision Nutrition in Maternal and Child Health

Artificial intelligence (AI) and machine learning (ML) methods have expanded into the public sphere and hold much promise for addressing issues in nutrition, from nutrition assessment to designing personalized dietary interventions. Improved computational power and advances in technology have enabled precision nutrition-based approaches that can complement current methods for assessing 1) nutritional status to identify those at risk as well as and 2) use individual-level anthropometric, clinical, biochemical, microbiome, social, and environmental characteristics to predict responses to an intervention or a program to address some of the structural challenges in improving maternal and child health and nutrition, which is critical for the well-being of all and for future generations. As malnutrition continues to be a major driver of poor maternal and child health, it is critical that we develop all the tools possible to address malnutrition and its preventable consequences such as compromised immune and cognitive function, growth, and metabolic health. We review how artificial intelligence-based methods have been applied or tested in populations of women, children, and infants in the literature, including machine learning models to account for the complexities of and across different data types (e.g., dietary intakes, images, microbiome, etc.). Further, we suggest additional opportunities for incorporating precision nutrition to monitor effectiveness of existing programs, measure efficacy of nutrition-based interventions, and ultimately to identify those likely to respond to these interventions. We also discuss myths related to AI and ML in the context of nutrition. Future directions for AI and nutrition should include training the next generation of nutrition scientists in foundational computer science techniques to foster cross-discipline communication and collaboration, key in moving the field forward.