The relation between bulk (external) and internal measures of spinal stiffness

TL;DR

This study shows that external spinal stiffness measurements do not reliably reflect internal spinal stiffness, even after adjusting for body and muscle factors.

Contribution

The study demonstrates that bulk spinal stiffness measurements are not valid proxies for internal spinal stiffness, even after normalization.

Findings

Unnormalized bulk stiffness had a weak correlation with internal spinal stiffness (R2 = 0.1883).

Normalizing bulk stiffness with body weight and muscle volume did not improve the correlation with internal stiffness.

Post hoc analysis suggested en bloc movement of the lumbar spine.

Abstract

While spinal stiffness is thought to be an important factor in the diagnosis and management of various spinal conditions, it is notoriously difficult to measure directly. As a result, clinicians often rely on posteroanterior palpation to estimate bulk stiffness as a proxy for the stiffness of internal spinal tissues. Unfortunately, the validity of this proxy remains uncertain. To investigate this, we posed two key research questions: (1) How do measurements of bulk stiffness correlate with direct measures of spinal stiffness? and (2) Can bulk stiffness measurements be normalized to more accurately reflect internal spinal stiffness?

This cross-sectional measurement study investigated the relation between bulk and internal spinal stiffness in a young, asymptomatic cohort. Bulk stiffness defined as external resistance of the spine measured through mechanical indentation at the L3 vertebra, while internal spinal stiffness was assessed concurrently using fluoroscopic imaging. Linear regression was used to analyze the relation between bulk and internal spinal stiffness measures. Bulk measures were then normalized using physical measurements (e.g. height, weight) and tissue volume measures of the multifidi obtained by MRI (i.e. muscle volume) and reanalyzed.

Twenty-six persons (26) participated, with data from 7 of those being excluded due to fluoroscopic movement artifacts. Unnormalized bulk stiffness was found to correlate poorly with internal spinal stiffness (R2 = 0.1883). Normalization of bulk stiffness using factors such as body weight and multifidus muscle volume did not improve R2 values. Our results were further validated through post hoc analysis, suggesting en bloc movement of the lumbar spine.

Raw bulk spinal stiffness values should not be used as a proxy for internal spinal stiffness as they measure unrelated constructs. Our results may help explain why bulk stiffness measures of the spine may not always align with clinical outcomes. Attempts to normalize bulk spinal stiffness to various human factors such as weight and paravertebral muscle volume did not improve the correlation between bulk spinal stiffness and internal spinal stiffness.

Figures