Instability of the spinal column is thought to be a contributory factor to the development of low back pain. Local short-latency stretch reflexes have been evoked in the erector spinae (ES) muscles in man (Dimitrijevic et al. 1980) and it is known that the supraspinous ligaments contain sensory receptors and are innervated by both spinal and autonomic nerves (Solomonow et al. 1998). In this study we have investigated reflexes in ES muscles evoked to mechanical stimulation of either the ES muscle itself or to the adjacent spinal processes along the length of the thoracic spine.

With local ethical approval and informed consent eleven healthy individuals (aged 24Ð46 years, 3 females) were recruited and seated on a low-backed chair. A Perspex probe (diameter 5 mm) was used to prod the spinal process (PP) at each of thoracic spinal levels T2, T7, T11 and lumbar segment L4 or the ES muscle (PM) adjacent to these segments. An electromagnetic servo applied the probe at a frequency of 1 Hz and a throw of 3 mm. Surface electromyographic (EMG) recordings were made from the right ES 3Ð6 cm lateral to the same four spinal processes. Trials of 50 mechanical prods were conducted at each probe site and reflex responses in ES were recorded and averaged from each of the four recording sites. Amplitude and latency of the reflex responses were recorded.

Evoked reflexes were seen at short (mean ± S.E.M. 7.54 ± 0.57 ms; n = 121) and long (mean 39.21 ± 0.76 ms; n = 199) latencies in response to both PP or PM at all four segmental levels. Short latency responses were seen with the following incidence: at prod site PP 100 %, PM 100% one recording site away PP 30 %, PM 24% two recording sites away PP 7 %, PM 4% three recording sites away PP 0 %, PM 0 %. The mean amplitude of the response was smaller (ANOVA on ranks, P < 0.05) at recording sites further from the prod site for both PP and PM stimulation. Long-latency responses were seen less frequently to PM than to PP (Student’s paired t test, P < 0.05) having the following incidence: at prod site PP 75 %, PM 45% one recording site away PP 71 %, PM 45% two recording sites away PP 64 %, PM 25% three recording sites away PP 27 %, PM 14 %. The mean amplitude of the response was smaller (ANOVA, P < 0.05) at recording sites further from the prod site for both PP and PM stimulation. There was no consistent change in latency of either reflex with either prod site or recording site.

The stronger and more frequently observed long-latency reflex to PP suggests that receptors responsible might lie in the supraspinous ligaments. The rather localised short-latency reflex suggests that mechanical deformation at a site remote from the prod is not evoking the long-latency reflex and that it may be mediated via the motor cortex. Further investigation using conditioning transcranial magnetic stimulation will clarify this.

This work was supported by the ARC and ISRT.

All procedures accord with current local guidelines and the Declaration of Helsinki.