Nutrition for the Sick Preterm: Can We Make It More Precise?

Abstract

In the early era of neonatal intensive care (about 5-6 decades ago), most nutritional approaches were based largely on the physician's intuition, previous experience, and patient's signs and symptoms. This resulted in a large heterogeneity of diagnostic, preventative, and therapeutic measures. More recently, evidence-based approaches, such as data reviews and clinical trials, form the foundation for nutritional guidelines used in most Neonatal Intensive Care Unit (NICUs). These are derived from population statistics aimed toward the average and, thereby, meet the needs of many of these infants, but because of the extreme heterogeneity of the preterm population, they marginalize others. Helpful scoring programs are now available to identify malnutrition in populations of preterm infants using defined indicators. However, similar to growth curves, they do not provide proactive guidance. Newly developed precision-based approaches using algorithms and predictive analytics based on artificial intelligence (AI) and machine learning (ML) will provide for a priori-based preventative approaches. It is likely that these will employ technologies that cluster infants into different risk categories that can then be investigated mechanistically with multiomic integrations that provide mechanistic interactions and provide clues to biomarkers that can be used for the discovery of biomarkers that can be utilized for the development of preventative strategies.