Thomas Ertelt

Low back pain affects trunk as well as lower limb movements during walking and running.

Up to now, most gait analyses on low back pain concentrate on changes in trunk coordination during walking on a treadmill. Locomotion on uneven ground as well as lower limb changes receives little attention in association with low back pain. The present study focuses on how chronic non-specific low back pain causes modifications in lower limb and trunk movements, in level and uneven walking and running. We found that trunk as well as lower limb movement was influenced by chronic non-specific low back pain. A consistent finding across all gaits and ground level changes is that patients with chronic non-specific low back pain show less pelvis and unchanged thorax rotation as compared to healthy controls. Furthermore, in chronic non-specific low back pain patients the trunk rotation decreased only during level and uneven running whereas the sagittal trunk inclination at touchdown increased only during uneven walking as compared to healthy controls. Besides significant changes in the upper body, in chronic non-specific low back pain patients the knee joint angle at touchdown was more extended during level walking but also during uneven walking and running as compared to healthy controls. We assume that trunk movements interact with lower limb movements or vice versa. Therefore, we recommend that further investigations on low back pain should consider both trunk (primarily pelvis) and lower limb (primarily knee) movements.

Leg adjustments as a key: initial insights into quick-release trials between healthy controls and chronic non-specific low back pain patients.

We state that all authors have made substantial contributions to all of the following: (i) the conception and design of the study, or acquisition of data, or analysis and interpretation of data; (ii) drafting the article or revising it critically for important intellectual content; and (iii) final approval of the version to be submitted. All contributors who do not meet the criteria for authorship will be listed in an acknowledgements section of the manuscript. Back pain has turned out be a diagnostically difficult area because the pain can be classified and distinguished in a wide variety of ways and even today the actual causes are largely unknown. It is generally recognized that patients with chronic non-specific low back pain (CNSLBP) have poor balance and a delayed response to different tasks (Luoto et al. 1998), as well as shorter strides (Khodadadeh et al. 1988) and an increase in contact time during walking (Keefe & Hill 1985). Vogt et al. (2003) observed significant differences in the activation behaviour of the m. biceps femoris. Still, it is nearly impossible to say whether the findings are causative or compensatory mechanisms of back pain. This study therefore examined whether, and how, provoked disturbances during walking lead to different and specific patterns of behaviour in healthy individuals compared to patients with CNSLBP. The classification and matching of the subjects were conducted by the Institute for Sports Medicine (FSU Jena) and were described in Ertelt et al. (2015). The subject’s task was to walk along a 17-m running track with a self-selected walking speed under loading. The loading was a constant traction force equivalent to 13% of the body weight. The traction force acted axially in the direction of walking. The subjects were told that at a specific time (contact with the force plate), an abrupt and total release would occur when the perturbation was applied. The experimental arrangement and the applied measurement systems as well as the disturbance application are described in detail in Ertelt et al. (2015). The study conforms to good publication practice in physiology (Persson 2015). The parameters analysed revealed some significant differences between the two groups studied. Except for differences resulting from specific time-parameters in ground reaction force, the analysis of knee joint acceleration also revealed a significant difference between healthy subjects and CNSLBP patients (Table 1; Fig. 1) when measured in a special way using pre-adjustment processes, measuring the residuals of knee acceleration over 100 ms from before contact up to touchdown (KJResTD). Compared to the healthy group, CNSLBP patients showed a lower adjustment process in the knee joint just before contact (up to 27%; Table 1, Fig. 1). Immediately before and after touchdown, a difference in behaviour was observed between the groups in knee joint acceleration. While in healthy people, knee joint acceleration decreased before touchdown, paused and then accelerated, in patients with CNSLBP their applied strategy was maintained beyond touchdown (Fig. 1, Table 1). The movement strategy plays a key role in the processing of loading and the transfer of load on structures of the human body. With the focus on patients with CNSLBP, these initial findings point to differences in kinetics as well as kinematic parameters. Some of our findings correspond to a known strategy in CNSLBP patients (prolongation of the contact time, lower unloading during contact; Luoto et al. 1998, Khodadadeh et al. 1988). Shortening the swing phase reduces the vertical deflection of the COM (centre of mass), while simultaneously increasing the contact time. Both adjustments lead to a reduction of total loading. The observed behaviour in CNSLBP patients partly refers to a lack of adjustment or maladjustments in the various leg joints, especially the knee joint. In the present case,

GROUP SPECIFIC BEHAVIOR OF BIARTICULAR UPPER LEG MUSCLES EXEMPLIFIED BY SLEDGE

The biarticular skeletal muscles represent a much considered topic in current research. Concerning these muscles, not only Lombards paradox is of special interest. Some authors refer to their energy-transferring function and others have observed an increase of activity in M. biceps femoris after a special training. By means of so-called sledge jumps, we have experimentally examined the behavior of M. biceps femoris for two groups, jumpers and non-jumpers, whose task was to push the sledge with different loads and different speeds. The results clearly show a specific behavior of the muscles for each group — which seems to affect the muscle coordination and has crucial influence on the shape of the ground reaction force. Due to nearly equal kinematic conditions, Lombards paradox could apply for both groups. However, the amount of training seems to be an essential parameter. The jumpers activate the M. biceps femoris during the entire contact phase and during the shortening as well as the stretching of the leg. Thus the force shape becomes smoother and shows fewer fluctuations. Despite positive geometrical conditions, not all subjects use this muscle from the beginning of the movement. In this context, neural pathways, inhibitions or coupling processes are decisive factors. All that indicates that the controlling of the muscle can only be achieved by an appropriate training.

THE GEOMETRY-CRITICAL FUNCTIONAL DEPENDENCE OF M. GASTROCNEMIUS

The change in length of a muscle usually leads to a change in segment position. If a muscle is shortened the joint is being flexed. However bi-articular muscles are able to bend a joint as well as to stretch it by its contraction because of varied geometrical conditions. This phenomenon has already been analyzed in different surveys on M. biceps femoris. Until today such a consideration cannot be found for M. gastrocnemius, which is a bi-articular muscle, too. The weight- and speed-dependent activities of this muscle during fast and slow hopping, and also during running and walking, suggest that a movement-relevant function exists. This behavior pattern was the reason for a closer look on the actions of M. gastrocnemius from a geometrical point of view. The results clearly show a knee angle-dependent function. Moreover, a so-called critical angle area of the knee joint, from which the shortening of M. gastrocnemius leads from joint flexion to joint stretching, can be specified for the first time. At this, the position of the ankle joint seems to have little influence on the function of M. gastrocnemius.

Describing force-patterns: A method for an analytic classification using the example of sledge jumps

To recognise and classify movement patterns correctly can be a difficult task. Nevertheless, movement analysts are working on it on a daily basis. Therefore, we have developed and evaluated a method to do the classification by using contact forces during hopping in a sledge system. Here, experiments showed that reaction-forces of different subjects on a sliding sledge could be divided into four major types. These types are symmetric single-modal (type I), positive mono-modal (type II), negative mono-modal (type III), and multi-modal associated with plateau formation (type IV). Up until now, an exact determination of these types was not possible. However, the new method helps to approximate those four types with well established mathematical functions. With this approach, the measured reaction-force will be reproduced by particular coefficients. Subsequently, the coefficients are subjected to a discriminant-analysis. The result is a three-dimensional function-coefficient, which allows the classification of the actual force-pattern on the one of the four types.

Walking with chronic non-specific low back pain-a failed strategy: What can we learn from sports?

The present theoretical consideration is focused on the different behaviours of patients with chronic non-specific low back pain and healthy people when walking. The ground reaction force pattern and its direct influence on the success of the applied strategy are of special interest. This consideration does not clarify whether or not and to what limit the indicated pattern is the main cause or compensatory effect of chronic non-specific low back pain. It is known that pain patients exhibit divergent walking patterns. In addition to a shorter stride, the relationship between swing-to-stance phase shifts in the direction of stance. This strategy should contribute to an improvement in the control. Simultaneously, the impact and the forces should be reduced. Recent studies refer to different activation behaviours of the bi-articular Musculus biceps femoris. According to our analysis, this muscle owns a key function concerning force transmission. The type of force transmission immediately influences system loading. The strategy exhibited by patients with chronic non-specific back pain fails due to false activations, in particular false activations of the M. biceps femoris. The essential objective of the strategy for reducing the maxima of the applied forces, as well as to distribute the force over a longer period of time, was implemented; however, it could be shown that the increase in the maximum force is clearly reached early. Such a fast development in force is equivalent to higher loadings to the human system. This concerns visco-elastic structures in particular. Due to the fast increasing forces, these structures possess low or insufficient damping properties. If our hypothesis of the failed strategy holds true, it might be the basis for new therapeutic approaches, as well as diagnostics involving non-specific low back pain.

Enhancement of force patterns classification based on Gaussian distributions

Description of the patterns of ground reaction force is a standard method in areas such as medicine, biomechanics and robotics. The fundamental parameter is the time course of the force, which is classified visually in particular in the field of clinical diagnostics. Here, the knowledge and experience of the diagnostician is relevant for its assessment. For an objective and valid discrimination of the ground reaction force pattern, a generic method, especially in the medical field, is absolutely necessary to describe the qualities of the time-course. The aim of the presented method was to combine the approaches of two existing procedures from the fields of machine learning and the Gauss approximation in order to take advantages of both methods for the classification of ground reaction force patterns. The current limitations of both methods could be eliminated by an overarching method. Twenty-nine male athletes from different sports were examined. Each participant was given the task of performing a one-legged stopping maneuver on a force plate from the maximum possible starting speed. The individual time course of the ground reaction force of each subject was registered and approximated on the basis of eight Gaussian distributions. The descriptive coefficients were then classified using Bayesian regulated neural networks. The different sports served as the distinguishing feature. Although the athletes were all given the same task, all sports referred to a different quality in the time course of ground reaction force. Meanwhile within each sport, the athletes were homogeneous. With an overall prediction (R = 0.938) all subjects/sports were classified correctly with 94.29% accuracy. The combination of the two methods: the mathematical description of the time course of ground reaction forces on the basis of Gaussian distributions and their classification by means of Bayesian regulated neural networks, seems an adequate and promising method to discriminate the ground reaction forces without any loss of information.