Increasing BMI increases lumbar intervertebral disc deformation following a treadmill walking stress test

Abstract

High body mass index (BMI) and obesity have been implicated as risk factors for lumbar degenerative disc disease and low back pain. Despite this, there is limited in vivo data to quantify how obesity influences the mechanical function of intervertebral discs (IVD) in response to activities of daily living. Recently, our lab has developed methodologies to non-invasively measure in vivo IVD deformation resulting from activities of daily living using magnetic resonance (MR) imaging and solid modeling techniques. This pilot study expands on these methodologies to assess how BMI influences IVD deformation following treadmill walking in eight asymptomatic individuals. Ordinary least squares regression analyses revealed a statistically significant relationship between BMI and compressive deformation (strain (%)) in the L5-S1 IVD (R² = 0.61, p < 0.05). This relationship was weaker in the L3-L4 (R² = 0.28, p > 0.05) and L4-L5 IVDs (R² = 0.28, p > 0.05). Importantly, no relationship between pre-exercise disc height and BMI was identified (p > 0.05). Therefore, the results of this study suggest that BMI may alter the mechanical response of lumbar spine IVDs, particularly at the L5-S1 level. Furthermore, the observed relationship between increased BMI and IVD compressive deformation, in the absence of a detected relationship between pre-exercise disc height and BMI, suggests that changes in IVD mechanical function may be more sensitive to alterations in disc health than static clinical imaging alone. This finding highlights the importance of quantifying disc mechanical function when examining the relationship between BMI and IVD degeneration.

Keywords: Gait; Imaging; In vivo; MRI; Obesity; Spine; Strain.

Figure 1:. IVD segmentation processing.

Visualization of the directions of the 1^st, 2^nd and 3^rd principal component vectors are denoted by the X, Y and Z axes, respectively. The Z-axis demonstrates the axis along which height is calculated. (A) The outer contours of each IVD were segmented on each slice of the MR images. (B) Segmentation yielded a 3D wireframe model of each IVD. (C) Following alignment of the 3^rd principal component vector with the z-axis of the cartesian coordinate system, a 3D triangulation mesh of the disc was created for subsequent analyses.

Figure 2:. IVD surface generation, height calculation and regional disc divisions.

(A) An automatic perimeter detection algorithm was applied to the triangulated disc meshes (Figure 1C) and the mesh elements within the perimeter were then used to create uniformly sampled superior (yellow) and inferior (blue) surfaces. (B) Each disc was subsequently divided into five regions to obtain measures of disc height: Nucleus Pulposus (NP), Anterior Annulus Fibrosus (AF), Posterior AF, Left Lateral AF, and Right Lateral AF. Starting at the centroid of the disc, the NP Region was defined by all points within [0–50%] of an individual disc’s semi-axes length (i.e., X/Y-axis), while the AF was defined by all points within [50–100%] of disc semi-axes length. Lateral divisions of the AF form an arc of 60°, while anterior-posterior divisions form an arc of 120°. Disc height within each region was then calculated as the mean difference between the points in the superior disc region minus the points in the corresponding inferior region.

Figure 3:. Subject IVD maps.

Subject-level IVD maps showing mean disc height (all levels) and regional strain measurements. In rows one (pre) and two (post) warmer colors are indicative of greater disc thickness, whereas cooler colors indicate smaller disc thicknesses. In row three compressive strain (%) is shown in red. Subjects are organized in order of increasing BMI from left to right across the page.

Figure 4:. OLS regression results:

IVD strain increased with increasing BMI at the L3-L4 (A) and L4-L5 (B) levels, although these relationships were not statistically significant at p<0.05. (C) A strong statistically significant (p<0.05) relationship between BMI and IVD strain was detected in the L5-S1 level.

Figure 5:. OLS regression results:

Mean pre-exercise disc height was not related to BMI at any level. Baseline disc height for each level and subject was defined as the unweighted average of disc heights in the 5 regions described in Figure 2.