159 Body composition measurement methods in preterm infants 7 in 1- and 2-year-olds.60 Nevertheless, we only found one study which assessed the predictive value of ultrasound measurement for body composition in preterm infants and they found a poor predictive value.29 Others did report on ultrasound measurements as a means to estimate body composition in preterm infants, but they did not assess the predictive value or validity.58,59 Hence, in our opinion, there seems to be insufficient evidence for the use of ultrasound as a reference method for body composition. However, since ultrasound measurements showed high reliability, it may be of interest to investigate whether other body sites are a better representation of body composition. Accuracy of MRI Though several authors suggested the use of MRI to measure body composition in preterm infants44,61, only one study was found which assessed the validity of MRI for the assessment of body composition in a small study population.30 Despite a sound assessment of repeatability, the actual fat mass (g) measurement was not compared with other techniques. Therefore, it should be concluded that more studies are necessary to draw conclusions on the use of MRI in determining fat mass in preterm infants. Accuracy of ADP Roggero and colleagues found a small bias when comparing ADP to isotope dilution.32 However, the limits of agreement were relatively wide, resulting in poor agreement between the two methods. Of note, the accuracy of ADP was only assessed in a small subgroup of 10 preterm infants. In line with their findings Forsumet al. also found a small bias when comparing ADP with deuteriumdilution in 14 preterm infants.31 Fat mass percentage, however, had relatively wide limits of agreement and thus poor agreement. Fat free mass density (g/ml), however, agreed well. Precision was studied by Roggero in a larger group of 57 preterm infants and also showed a small bias for fat mass percentage between repeated ADP measurements. 32 Despite wide limits of agreement the authors concluded that ADP shows good agreement with isotope dilution for fat mass percentage as well fat free mass density.31,32 Likewise, carcass analyses showed small bias, with relatively wide limits of agreement.36 Nevertheless it is generally accepted that ADP is a reliable method in infants. Taking into account the relatively small study populations and small bias, it is to be expected that larger studies would yield better agreement. Therefore, we conclude that ADP is a reliable method to assess fat mass in preterm infants. Accuracy of isotope dilution Isotope dilution is a well-established method for the measurement of TBW from which the fat free mass can be derived.62 As a result there were no studies validating isotope dilution against another reference method, such as DXA, in
RkJQdWJsaXNoZXIy MTk4NDMw