| الوصف: |
Infants born before 32 weeks postmenstrual age (PMA) are at high risk of growth failure. Current guidelines recommend that the growth of preterm infants should match that of the equivalent fetus in utero, both in terms of weight gain and body composition, but that target is commonly missed. There is emerging evidence that nutrition and growth during the neonatal period is associated with neurodevelopmental outcome. However, factors influencing the growth of very preterm infants are incompletely understood, limiting the capability of clinicians to adjust nutritional care to the individual needs of each infant. This research project aims to assess multiple clinical, nutritional, genomic and metabolomic influences on the growth of very preterm infants, working towards a toolkit to guide personalised nutritional care. The work is centred on the formation of a comprehensive quality-assured relational database: the Southampton Preterm Nutritional Database. This contains prospectively gathered information for over 600 infants born before 32 weeks PMA and cared for in Southampton's neonatal unit. It combines nutritional intake data for over 33,000 care days with comprehensive demographic, clinical and biochemical information. Regression and machine learning techniques can be applied to these data to provide insights into the impact of these factors on growth. This research programme also includes total body water analysis of nine infants, using deuterium oxide dilution, allowing a marker of body composition to be tracked longitudinally in this subgroup. Whole exome sequencing has been performed for 13 infants, with these data being included in the database as a pilot for future analysis of genetic factors influencing growth. Metabolomic analysis of weekly urine samples from 14 infants also provides a model for the investigation of the effect of genomic, clinical and nutritional factors on the metabolic maturation of the very preterm infant. In addition to detailed analysis of these infants from a single centre, growth data for around 30,000 infants born across England were analysed to assess changes in growth patterns over time. Random forest machine learning has been used to identify the key factors influencing growth. Growth charts were published based on the local cohort of infants for whom a detailed accompanying description of nutritional and clinical factors was available, setting out an expected growth pattern in response to a defined nutritional approach. Growth, nutritional and biochemical results have also allowed presentation of an exploration of the influence of protein intake on plasma urea. A changing national pattern of weight gain, with generally more rapid growth in the most preterm infants, has been identified. Taken together, the findings presented in this thesis provide a guide to key modifiable factors influencing growth, a range of charts to monitor growth and pilot data on total body water, genomic analysis and metabolomic profiling which will be employed to provide further insights in the future. The work is underpinned by the formation of a comprehensive research database. |
