Percentiles for paraspinal muscle parameters at the L3 vertebra level in patients undergoing lumbar computed tomography scanning

Abstract

Computed tomography (CT) is the gold standard for analyzing muscle parameters. To clarify sex-specific paraspinal muscle area (PMA), paraspinal muscle index (PMI), and muscle fat infiltration (MFI) percentiles. This was a cross-sectional study of 760 individuals (45% men; age range = 20-92 years; mean age = 53.4 ± 21.1 years) with a body mass index (BMI) in the range of 16.4-38.1 kg/m2. CT scans were retrospectively used to establish PMA, PMI, and MFI at L3 level using a deep-learning (DL) tool. Sex-specific distributions for these parameters were assessed based on associations between age/BMI and individual muscle parameters, after which age- and BMI-specific percentile estimates were determined. The 5th percentile was regarded as the cutoff for PMA/PMI, and the 95th percentile was regarded as the cutoff for MFI. Sex-specific PMA, PMI, and MFI cutoffs in the paraspinal muscles group were 52.9 cm2, 15.0 cm2/m2, and 33.3%, respectively, in men, and 33.2 cm2, 9.5 cm2/m2, and 41.2% in women. Age was moderately negatively correlated with PMA and was strongly negatively correlated with PMI, but age was strongly positively correlated with MFI. BMI was moderately positively correlated with PMA/PMI in men and strongly positively correlated in women; BMI was weakly positively correlated with MFI, thus enabling the establishment of age- and BMI-specific cutoff percentiles. Sex-specific PMA, PMI, and MFI percentiles and age- and BMI-specific cutoff values for these parameters were successfully established for an outpatient population.