Measures of complexity during walking in chronic non-specific low back pain patients

Abstract

Summary Objective Chronic low back pain is costly both to society and to the individual. Although patient reported outcomes are common in the low back pain literature, objective measures of function are rare. Some authors have shown that low back pain affects gait, but these have predominantly used traditional average measures such as gait speed, stride/step length and the amount of variability. Recent developments in the use of non-linear methods of investigating biological time series have supported the notion that ‘normal’ biological systems are stable because they contain degrees of noise. A critical amount of noise in these time series appears to reflect adaptability and therefore stability. However, it is not the ‘amount’ of noise, but the structure that appears to be important for systems to adapt to perturbation. This has been shown in human gait in the context of elderly fall risk and neurodegenerative disorders, with emerging research suggesting that such measures reflect complex neural feedback loops involved in the control of gait. To date little if any research has looked at these concepts in the context of gait changes in non-pathological scenarios such as non-specific low back pain. The aim of this study is therefore to ascertain any possible differences in the complexity of stride interval correlations (fractal structure) between subjects with chronic non-specific low back pain and control subjects. Method Twenty-four subjects, 12 with low back pain and 12 controls participated, recruited from a UK chiropractic teaching clinic. Subjects walked on an instrumented treadmill at their preferred walking speed for a period of 8 min. Gait parameters including walking speed, step cycle, step length, coefficient of variation and ambulation index were calculated. A measure of the complexity of the step intervals was ascertained through detrended fluctuation analysis. Results Low back pain subjects had a slower walking velocity, smaller step length and reduced complexity (lower fractal scaling indices) compared to control subjects. Conclusions This study is the first to show differences in the fractal indices of gait between low back pain patients and matched controls. It is possible that further development of this outcome measure may result in a useful diagnostic and prognostic tool for the musculoskeletal pain patient. In addition, it could plausibly be developed as an objective monitor of treatment effectiveness adjunctive to more traditional patient reported outcomes.