Background: Neonatal body composition is an important outcome in diabetes in pregnancy trials. However, gold-standard three- or four-compartment models are not feasible in neonates, and current neonatal body composition techniques have limitations. We assessed the feasibility of using ultrasound to measure fat and muscle depth in a research context.

Methods: Neonates were recruited from a feasibility trial of continuous glucose monitoring for type 2 diabetes in pregnancy (ACTRN12625000370404). Following brief training and support from an experienced sonographer, a research midwife used a portable ultrasound device (Vscan, GE Healthcare) to obtain triplicate measures of biceps, quadriceps and upper abdominal skin fat depth, and biceps and quadriceps muscle depth. Whole-body fat and fat-free mass index (FI, FFI, kg/m²) were determined from air displacement plethysmography. Correlations with customised birthweight z-score (cBWz, referenced to fetal growth curves and expected maternal constraint) and between measures were assessed.

Results: Of 22 ultrasound assessments attempted, 1 was unsuccessful, 1 had partial images and 20 (91%) were completed successfully.  Mean (SD) gestation was 36.9 (1.1) weeks and cBWz was 0.61 (1.82). Six (29%) infants were large for gestational age and 2 (10%) were small for gestational age. Mean intra-subject coefficient of variation for biceps, quadriceps and abdominal fat depth was 10%, 14% and 8%, and for biceps and quadriceps muscle depth was 6% and 7%. Quadriceps and abdominal skin fat depth were strongly correlated with cBWz (both r=0.69), but biceps skin fat depth was only moderately correlated with cBWZ (r=0.47). Abdominal skin fat depth was strongly correlated with FI (r=0.71), but biceps and quadriceps skin fat depth were only moderately correlated with FI (r=0.55 and 0.60). Biceps muscle depth was moderately correlated with cBWz (r=0.37) and FFI (r=0.36). Quadriceps muscle depth showed no association with cBWz or FFI. FI was strongly correlated with cBWZ (r=0.85), but FFI was only moderately correlated with cBWz (r=0.48).

Conclusions: Relatively precise ultrasound measures were obtained by a novice operator after brief training. Ultrasound measurement of abdominal skin fat depth is a potentially useful indicator of neonatal subcutaneous adipose mass. Assessing muscle mass in neonates remains challenging.