Ligaments can do work during repetitive straining

Abstract

We have found that when the longitudinal ligaments of the spine are strained, they can continue to store energy after recoil; they can release some of this energy if they are strained and recoil again within a sufficiently short period (up to about 3 h). Thus ligaments which have had work done on them can store strain energy and subsequently do work themselves. Longitudinal ligaments are biological examples of composite materials in which crimped collagen fibres are surrounded by a weak matrix, often referred to as "ground substance" [1-4]. However , ligaments also contain cells, known as fibroblasts, which contain the contractile protein actin. In some ligaments (e.g. the collateral ligaments of rat knees) the fibroblasts are known to form extensions (processes) which contain actin filaments and surround the collagen fibres [5]. The results reported here show that the ligaments have unusual mechanical properties which may be associated with the incoporation of a contractile material within their structure. Most of the experiments reported here were performed on longitudinal ligaments dissected from the lumbar spines of mini-pigs which had been frozen immediately after death. However , further experiments were performed on human tissue and on ligaments which had been dissected from the spines of pigs which had not been frozen. A total of 20 pig ligaments (3 anterior longitudinal ligaments and 17 posterior longitudinal ligaments) and 1 posterior human ligament were used for these experiments. The anatomy of these ligaments within t h e spine and their functions are described elsewhere [6]. Specimens were moistened with physiological saline (9.5 g1-1 NaC1) throughout the experiments which were performed at room temperature (20 °C) and body temperature (37°C). All specimens showed qualitatively the same effects. That temperature had little effect on our results is consistent with the observation that the mechanical properties of other ligaments and tendons are independent of temperature in this range [7-9]. Mechanical testing was performed using computer-centrolled apparatus which has been described elsewhere [10]. Specimens were clamped to the apparatus in the same way as in our previous experiments on longitudinal ligaments [3, 4]. Video recordings were made of the specimens during testing so that the strain could be independently checked [11]. Strains were applied at rates of either 0.03 or 0.2 s -1, to a maximum value of typically 0.15. When the maximum strain had been attained in a cycle, the ligament was allowed to recoil, i.e. the clamps attached to its ends were returned to their original separation. The rate of recoil was controlled and was simply the negative strain rate. The experimental procedures were also carried out using steel springs and rubber bands as specimens, to check that they behaved as expected, i.e. to ensure that there was not a phase delay between the load-cell reading and the signal to the stepper motor used to apply the strain. In all our experiments the strain was controlled rather than stress. Each ligament was stretched to some predetermined strain and the clamps were always returned to their original separation. The stress developed was then a consequence of the applied strain. Usually, a material is returned to the original stress in a cyclic test. However, during joint movement , it is tissue strain which is controlled; muscles contract until the required posture is attained. This posture then defines the strains in the deformable joint tissues [12]. Fig. 1 shows the result of cyclically stretching a longitudinal ligament. Hysteresis is most clearly apparent in the first cycle, where, as expected, more work was done in stretching the ligament than was done by it in recoil. In principle this effect could be accounted for by energy dissipation (which occurs with all viscoelastic materials) or by energy storage. In practice only the former mechanism for hysteresis