2.2. Crown-Heel Length

Body length reflects skeletal growth and fat-free mass ^[23][24][32,35]. In preterm infants, it was found to be a good predictor (r² = 0.85) of ADP-determined fat-free mass shortly after birth [13]. Length assessment is useful only if measurement has been accurately undertaken ^[23][32]. Length is frequently measured inaccurately or ignored in clinical practice because of the perceived difficulty in measurement of neonates ^[24][25][35,48]. The reluctance of the observer to extend the lower limbs forcefully is a factor affecting its accuracy in full-term neonates, who are more comfortable in flexion ^{[25][26][27][28]}[48,49,50,51]. In spite of the lack of similar studies in preterm infants, the reluctance to fully extend the lower limbs as recommended seems to be widespread in these infants. Moreover, the accuracy of measured length is of utmost importance when it is squared or cubed in indices, since a small error in its measurement amplifies the distortion of final results ^[16][29][16,52].

2.3. Head Circumference

An increase in occipitofrontal circumference reflects brain growth, although it is not a sensitive or specific measure [19]. Some factors should be taken into account when interpreting HC measurements in neonates. During the first postnatal week, HC may decrease by about 0.5 cm due to extracellular fluid space contraction [19]. If HC values are above or below the reference limits it may represent a variant of normal, and therefore, the parents’ HC should be measured and plotted [16]. In infants born prematurely, a non-expected deviation of head growth in the presence of good weight and linear growth should be investigated for causes other than inadequate nutrient intake. These may be related with prematurity-related morbidity, including post-hemorrhagic hydrocephalus or brain atrophy ^[19][30][19,31]. In addition, magnetic resonance imaging measurements recently revealed that in “encephalopathy of prematurity”, increased size of the extra-axial spaces may artificially increase HC ^[31][54]. In preterm infants, postdischarge head growth seems to be a better predictor of cognitive outcomes than intra-hospital head growth ^[32][55].

2.4. Mid-Upper Arm Circumference

The MUAC reflects the combined arm muscle and fat; a decrease in its value indicates a reduction in body muscle and/or fat mass ^[23][33][32,56]. As the upper arm is less affected by fluid status changes than other areas of the body, in the presence of edema, the MUAC may be a more accurate estimate than other measurements ^[23][32]. In preterm infants, measurements of MUAC are reproducible and detect changes over time ^[33][56]. In moderately preterm infants, adiposity defined by ADP-determined percent of fat mass (%FM), accounts for 60.4% of the variation of MUAC ^[34][57].

2.5. Skinfolds

Skinfolds measurement is an inexpensive and suitable method for bedside nutritional assessment in neonates ^[10][16][10,16]. The triceps and biceps skinfolds are used mainly to assess peripheral subcutaneous fat, whereas subscapular and suprailiac skinfolds are used to assess central subcutaneous fat ^[35][61]. The use of skinfolds to estimate body fat assumes that the thickness of subcutaneous fat reflects a constant proportion of total body fat and the sites used for the measurements reflect the average thickness of the subcutaneous fat layer. Although these assumptions have been questioned ^[36][62], a good correlation was reported between skinfolds and DXA- ^[37][38][63,64] and ADP-determined body fat ^[34][57]. Contrarily, skinfolds seem to overestimate total body fat comparing with body water dilution measurements ^[39][65]; however, this method assumes a constant of lean tissue hydration that in fact is not constant and varies with age [10]. Some reasons have been reported for the discrepant results between studies. Skinfolds do not reflect intra-abdominal fat; the hydration status affects skinfold compressibility; and reproducible measurements require skill and practice of the observer ^{[15][16][36][40]}[15,16,62,66]. Particularly in preterm infants, the rapidly changing distribution of fat accretion makes it difficult to generate an accurate equation for predicting total body fat ^[36][62].

2.6. Weight-to-Length Based Equations

Indices based on weight and length commonly used to assess body proportionality and body composition in neonates include the weight-for-length ratio, the body mass index (BMI) (weight/length²), and the ponderal index (weight/length³) ^[12][13][16][12,13,16]. The reliability of these indices is highly dependent on the accuracy of length measurement [16]. It has been described that accurate crown-heel length measurement is difficult to obtain at least in term neonates ^[28][51]. Inaccurate length squared in the BMI magnifies the error while leading to the index decreasing its ability to differentiate over- from underestimation. When cubed to obtain ponderal index, the inaccuracy of the length measurement is further magnified, despite the fact that it still differentiates overestimation from underestimation ^[29][52]. Ponderal index has been the traditional measure used to assess proportionality at birth and to distinguish between asymmetrical and symmetrical types of intrauterine growth restriction ^[41][69]. More recently, BMI has been reported to be more appropriate to assess body proportionality than either weight-for length ratio or ponderal index ^[42][70]. When used to assess body composition, BMI accounts for more than 81% of the variance of DXA-determined lean mass in AGA term infants measured shortly after birth ^[38][64]. Compared with ADP-determined %FM (adiposity), all weight-to-length based equations were found to be poor predictors of adiposity in preterm infants ^[10][13][10,13], even though the BMI z-score predicts adiposity better than the ponderal index (r² = 0.43 vs. 0.29) ^[43][71].

2.7. Mid-Upper Arm Circumference to Head Circumference Ratio

The MUAC:HC ratio may contribute to the estimation of body composition; its usefulness as a complementary index was assessed, comparing with DXA measurements ^[38][64]. This index was proposed to identify acute intrauterine malnutrition at birth, assuming that in acute protein-energy deprivation the brain is spared in relation to muscle and fat ^[44][73]. It may be particularly useful to identify symptomatic malnourished AGA neonates who are not diagnosed based exclusively on birth weight ^[45][74]. Longitudinal MUAC:HC measurements were reported to be useful for monitoring growth, seeming to not overestimate malnourishment in apparent protein-energy sufficiency ^[46][75].

2.8. Upper-Arm Cross-Sectional Areas

Upper-arm fat and muscle areas have been used in the assessment of nutritional status in infants ^[47][76]. For their calculation, it is assumed that the upper arm is cylindrical, the subcutaneous fat is a concentric ring evenly distributed around a circular core of muscle, the fat thickness is half the triceps skinfold, and the muscle includes the humeral diameter ^[48][77]. Two equations derived from MUAC and triceps skinfold have been proposed relying on different geometrical assumptions ^[48][49][77,78]. Upper-arm fat and muscle areas measurements should be interpreted with caution in neonates, since their ability to predict total body fat and muscle has been questioned ^[38][64]. In term infants assessed shortly after birth, the added value of cross-sectional arm areas contributed little to detect the variation of DXA-determined body lean and fat mass, compared with weight and length alone ^[38][64]. The ability of cross-sectional arm areas to predict mid-upper arm muscle and fat was also questioned. In full-term infants, a poor correlation of arm areas with ultrasound measurements was found, leading to overestimation of muscle and underestimation of fat ^[50][79]. In preterm infants, arm muscle and fat areas were inaccurate predictors (r² < 0.56) of magnetic resonance imaging measurements ^[51][59]. These poor correlations may be explained by the limited reliability of MUAC ^[34][57] and triceps skinfold ^[39][65] measurements and by the oversimplification of geometrical assumptions used for the calculation of cross-sectional arm areas [16].

3. Biochemical Markers

In preterm infants, some biochemical markers are useful in the assessment of nutritional status, helping to detect nutritional deficiencies before the appearance of clinical signs [17]. These markers should be interpreted with caution and used to complement other nutritional data, including anthropometric measurements [17]. Markers for metabolic and electrolyte, iron, protein, and bone status have been reviewed previously ^[8][17][52][8,17,81]. Markers to monitor protein status and bone status are summarized in Table 1.