DESCRIPTION (provided by applicant): Extremely preterm infants (EPIs) (< 28 weeks' gestation) have increased susceptibility for multiple significant short- and long-term medical complications that significantly delay the development of essential milestones, such as oral feeding, required for discharge. Further, infants who develop bronchopulmonary dysplasia (BPD) are at even greater risk for oral feeding difficulties compared to gestationally aged matched controls. Despite the frequency of oral feeding difficulties in this at-risk population, there currently exists no objective assessment tool to predict when an infant may safely feed by mouth and only limited intervention strategies to improve feeding behavior. The overall goal of this research proposal is to pair the expertise of Co-PIs Drs. Steven Barlow and Jill Maron to develop an innovative, novel, personalized and integrative approach that directly addresses the important need for oral feeding treatment strategies for the EPI population. In this multicenter trial, 180 EPIs will be enrolled from three neonatal intensive care units (CHI Health St. Elizabeth, Lincoln, NE; Tufts Medical Center, Boston, MA; and Santa Clara Valley Medical Center, Santa Jose, CA). Infants will be randomized to receive either PULSED orocutaneous somatosensory stimulation (PULSED NTrainer(r)) previously developed in the Barlow Laboratory, or SHAM (blind pacifier) starting at 30 weeks' post- conceptional age (PCA) up to the achievement of full oral feeds or discharge home. Serial salivary samples (n=2/week) will be obtained concomitantly to examine the gene expression profiles of six biomarkers (AMPK, NPY2R, NPHP4, WNT3, PLXNA1, PLXNA3) previously shown by the Maron Laboratory to be associated with oral feeding success in the premature newborn. In Aim #1, we will test the hypothesis that infants who received PULSED NTrainer(r) intervention will have a shortened duration of time to achieve full oral feeds and that their salivary gene expression profiles will correlate with feeding success. Infants will be stratified based upon their gestational age, sex, and BPD status. In Aim #2, computational analyses of both clinical and gene expression data from Aim #1 will be performed to identify responders and non-responders to the PULSED NTrainer(r) intervention and to identify critical times in neurodevelopment when infants may most benefit from the intervention. Finally in Aim #3, infants will undergo Bayley III Developmental Screen at 18-24 months' PCA to assess both the long-term effect of the PULSED NTrainer(r) on feeding behavior and neurodevelopment and the accuracy of neonatal salivary gene expression profiles to predict impairments later in childhood. Infants will also be assessed with the NICHD NRN 18-month Feeding-Growth-Nutrition Questionnaire to further our understanding of the PULSED NTrainer(r) on long-term growth and feeding behavior.