Measuring the neutral zone of spinal motion segments: Comparison of multiple analysis methods to quantify spinal instability.

PURPOSE:Neutral zone (NZ) parameters in spinal biomechanics studies are sensitive to spinal instability, disc degeneration, and repair. Multiple methods in the literature quantify NZ, yet no consensus exists on applicability and comparability of methods. This study compares five different NZ quantification methods using two different load-deflection profiles. METHODS:Rat caudal and lumbar motion segments were tested in axial rotation to generate load-deflection curves with profiles exhibiting prominent distinction between elastic and NZ regions (ie, triphasic) and profiles that did not (ie, viscoelastic). NZ was quantified using five methods: trilinear, double sigmoid (DS), zero load, stiffness threshold (ST), and extrapolated elastic zone. Absolute agreement and consistency of NZ parameters were assessed using intraclass correlation (ICC), Bland-Altman analyses, and analysis of variance. RESULTS:For triphasic profiles, NZ magnitude exhibited high consistency (methods correlate but differ in absolute values), and only some methods exhibited agreement. For viscoelastic profiles, NZ magnitude showed limited consistency and no absolute agreement. NZ stiffness had high agreement and consistency across most methods and profiles. For triphasic profiles, the linear NZ regions for all methods were not well-described by a linear fit yet for viscoelastic profiles all methods characterized a linear NZ region. CONCLUSION:This NZ comparison study showed surprisingly limited agreement and consistency among NZ parameters with approximately 5% to 100% difference depending on the method and load-deflection profile. Nevertheless, the DS and ST methods appeared to be most comparable. We conclude that most NZ quantification methods cannot be applied interchangeably, highlighting a need to clearly state NZ calculation methods. Future studies are required to identify which methods are most sensitive to disc degeneration and repair in order to identify a "best" method.