683 - Bioelectrical impedance analysis at birth among preterm infants from 23 to 35 weeks' gestation: associations of measured resistance with gestational age

Publication Number: 683.3763

Katherine Bell, Brigham and Women’s Hospital, Boston, MA, United States; Erin Plummer, Children's Hospitals and Clinics of Minnesota, Minneapolis, MN, United States; Juan D. Gonzalez Villamizar, Children's Hospitals and Clinics of Minnesota, University of Minnesota, Minneapolis, MN, United States; Nicole L. Welke, University of Minnesota Masonic Children's Hospital, Saint Paul, MN, United States; Hunter L. Pepin, Brigham and Women's Hospital, Boston, MA, United States; Wei Wang, Brigham and Women's Hospital, Boston, MA, United States; Mandy Belfort, Harvard Medical School, Boston, MA, United States; Sara E. Ramel, University of Minnesota Masonic Children's Hospital, North Oaks, MN, United States

Background: Body composition is an emerging indicator of nutritional status for preterm infants in the neonatal intensive care (NICU) because greater fat-free mass is linked to improved brain development and excess adiposity may confer cardiometabolic risk. Current methods for measuring body composition are limited to stable infants free from respiratory support. Bioelectrical impedance analysis (BIA) is a portable, low-cost method that can be applied to sick, small infants even in the first days of life. BIA measures the body’s resistance to a weak electric current; resistance indicates the amount of total body water, an indirect measure of fat-free mass. Yet, the extent to which BIA-derived resistance values vary across the gestational age (GA) spectrum is unknown.

Objective: Among preterm infants born across a wide range of GA from 23 to 35 weeks, with BIA assessed within 72 hours of birth, to determine the 1) reliability of BIA measurements, and 2) relationship of BIA-derived resistance with GA.

Design/Methods: We enrolled singleton AGA preterm infants < 36 weeks' GA in an ongoing prospective observational study at four level 3/4 NICUs. Within 72 hours of birth, research staff performed BIA in triplicate to measure whole-body resistance at 50kHz. We assessed intra-rater reliability of the triplicate resistance values using intraclass correlation coefficient (ICC). We used the median resistance value for each infant in a linear regression model to determine associations of resistance with GA.

Results: We studied 116 infants, 51.7% male, median GA 33.4 weeks (range 23.4, 35.9) (Table 1). Intra-rater reliability of BIA-derived resistance was high; ICC=0.98. Resistance decreased with increasing GA (-7.2 Ohms/week, 95%CI:-0.3,-14.2) (Fig 1). There was no significant difference in resistance values of male compared with female infants.

Conclusion(s): To date, methodologic limitations have prevented body composition measurement for sick or very preterm infants, especially early in the NICU hospitalization. Our initial experience performing BIA for small, sick preterm infants in the first few days of life demonstrates feasibility of this method in this population. The whole-body resistance results are biologically plausible. Declining resistance with greater gestational age is consistent with the known decrease in total body water as gestational age increases, and support the need for GA specific BIA reference data. Our findings suggest BIA is a promising potentially scalable method for clinical application in the NICU, which could facilitate longitudinal body composition measurements among nutritionally vulnerable infants.

Table 1


Figure 1