Reference values for MRI‐derived psoas and paraspinal muscles and macroscopic fat infiltrations in paraspinal muscles in children

Abstract

Background

Sarcopenia, defined as loss of skeletal muscle mass, is a novel term associated with adverse outcomes in children. Magnetic Resonance Imaging (MRI) is a safe and precise technique for measuring tissue compartments and is commonly used in most routine paediatric imaging protocols. Currently, there is a lack of MRI‐derived normative data which can help in determining the level of sarcopenia. This study aimed to introduce reference values of total psoas muscle area (tPMA), total paraspinal muscle area (tPSMA), and total macroscopic fat infiltrations of the PSMA (tMFI).

Methods

In this retrospective study, the local database was searched for abdominal and pelvic region MRI studies of children aged from 1 to 18 years (mean age (standard deviation (SD)) of 9.8 (5.5) years) performed in the years 2010–2021. Children with chronic diseases and a history of surgical interventions were excluded from the analysis. Finally, a total of 465 healthy children ( n = 233 girls, n = 232 boys) were enrolled in the study. The values of the tPMA, tPMSA, and tMFI were measured in square centimetres (cm ²) at the level of the L4/L5 intervertebral disc as the sum of the left and right regions. Age‐specific and sex‐specific muscle, fat, and body mass index percentile charts were constructed using the LMS method. Inter‐observer agreement and intra‐observer reproducibility were assessed using the Bland–Altman plots.

Results

Both tPMA and tPSMA showed continuous increases in size (in cm ²) throughout all age groups. At the age of 18, the median tPMA areas reached 26.37 cm ² in girls and 40.43 cm ² in boys. Corresponding tPSMA values were higher, reaching the level of 40.76 cm ² in girls and 56.66 cm ² in boys. The mean value of tMFI within the paraspinal muscles was 5.0% (SD 3.65%) of their total area in girls and 3.5% (SD 2.25%) in boys with the actual difference between sexes up to 0.96 cm ². Excellent intra‐observer reproducibility and inter‐observer agreement were noted. Actual mean differences for tPMA were at the level of 0.43 and 0.39 cm ², respectively. Mean bias for tPSMA was 0.1 cm ² for inter‐observer and 0.05 cm ² for intra‐observer measurements.

Conclusions

Our findings demonstrate novel and highly reproducible sex‐specific MRI‐derived reference values of tPMS, tPSMA, and tMFI at the level of the L4/L5 intervertebral disc for children from 1 to 18 years old, which may guide a clinician in the assessment of sarcopenia, a prognostic outcome marker in children.