The effect of posture on lumbar muscle morphometry from upright MRI.

Abstract

To assess the effect of upright, seated, and supine postures on lumbar muscle morphometry at multiple spinal levels and for multiple muscles. Six asymptomatic volunteers were imaged (0.5 T upright open MRI) in 7 postures (standing, standing holding 8 kg, standing 45° flexion, seated 45° flexion, seated upright, seated 45° extension, and supine), with scans at L3/L4, L4/L5, and L5/S1. Muscle cross-sectional area (CSA) and muscle position with respect to the vertebral body centroid (radius and angle) were measured for the multifidus/erector spinae combined and psoas major muscles. Posture significantly affected the multifidus/erector spinae CSA with decreasing CSA from straight postures (standing and supine) to seated and flexed postures (up to 19%). Psoas major CSA significantly varied with vertebral level with opposite trends due to posture at L3/L4 (increasing CSA, up to 36%) and L5/S1 (decreasing CSA, up to 40%) with sitting/flexion. For both muscle groups, radius and angle followed similar trends with decreasing radius (up to 5%) and increasing angle (up to 12%) with seated/flexed postures. CSA and lumbar lordosis had some correlation (multifidus/erector spinae L4/L5 and L5/S1, r = 0.37-0.45; PS L3/L4 left, r =  - 0.51). There was generally good repeatability (average ICC(3, 1): posture = 0.81, intra = 0.89, inter = 0.82). Changes in multifidus/erector spinae muscle CSA likely represent muscles stretching between upright and seated/flexed postures. For the psoas major, the differential level effect suggests that changing three-dimensional muscle morphometry with flexion is not uniform along the muscle length. The muscle and spinal level-dependent effects of posture and spinal curvature correlation, including muscle CSA and position, highlight considering measured muscle morphometry from different postures in spine models.