119 DXA, ADP & SFT in preterm infants 6 Skinfolds were measured (to the nearest mm) at biceps, triceps, subscapular and supra-ilial positions with a Harpenden® skinfold caliper by two investigators. One measurement was taken bilaterally for every position. The bilateral measurements were averaged to come to one skinfold thickness for every position. According to previous studies, the intraobserver coefficient of variation is below 3%, however the inter-observer coefficient is up to 10%(15). The anthropometric formula which was used to estimate fat mass at term equivalent age was that of Schmelzle and Fusch (12). This formula was originally modeled to predict fat mass, measured with DXA, in infants 34 weeks gestational age and older, using ∑SFT (mm) and length (cm): Fat mass (g) = 68.2 x ∑SFT ^ (0.0162 x length) – 172.8. Skinfolds were measured at the same site as our study. This formula was selected because it was the only predictive equation based on a population that included preterm infants and gave a high explanation of the variance in fat mass (12, 14). The Pea Pod® (PEA POD Infant Body Composition System, Cosmed, Ltd, Concord, CA, USA) was used to assess whole body fat mass and fat-free mass through ADP. The measurements were performed by two investigators. Infants were measured naked and hair was flattened using hair oil. Infants were allowed to move during the measurement. In case of excessive crying the measurement was stopped. Measurements were done briefly before feeding time, i.e. approximately 3 hours after the last feeding. In line with the manufacturer's guideline, daily quality control checks were done which included chamber calibration. Every two weeks the scale was calibrated. A detailed description of the Pea Pod® measurement is described elsewhere(19). As previously reported, the coefficient of variation for repeated volume measurements lies between 0.02 and 0.09% (19). Fat mass and fat free mass were calculated using gender-specific equations developed by Fomon and colleagues (20). TheHologicQDR4500A, using InfantWhole Body Software version 13.5.3:3 (Hologic Inc, Bedford, MA, USA) was used to assess whole body fat mass and fat-free mass through dual-energy x-ray absorptiometry. During the procedure, the infants were required not to move. The infants were swaddled in a blanket of the same size and type supplied by the investigators, without any clothing or diapers. Infants were swaddled in supine position with the soles of the feet together and knees bent (frog-leg-position) and the arms stretched beside the body. Infants were positioned in the center of the scanning bed with their head near the head end of the bed. The measurement was done after feeding. Typically infants remained awake, but lights were dimmed and a video was played from a mobile device outside of the scanning field. The preparation and positioning of the infants was performed by two experienced investigators. Calibration was done daily using an anthropomorphic spine phantom and a geometric block phantom. In addition a radiographic uniformity test was done once a week and software was regularly
RkJQdWJsaXNoZXIy MTk4NDMw