Shoulder posture and median nerve sliding

Longitudinal nerve movement was measured using high frequency ultrasound imaging, as previously described by Dilley et al. [5, 6]. A Diasus ultrasound system (Dynamic Imaging, Livingston, Scotland, UK) was used to collect sequences of ultrasound images at 10 frames/second for 50 to 70 seconds, running at 10–22 MHz and using a 26 mm linear transducer. A cross-correlation algorithm was used to determine relative movement between adjacent frames in sequences of images [5]. The maximum correlation coefficient (r) was calculated for each pixel shift determining the relative movement between frames. To account for probe movement the same method was employed on deep stationary structures (eg. bone or interosseous membrane) and the result subtracted from the nerve excursion values.

Fourteen healthy subjects, 5 male and 9 female, aged 25–38 years (mean = 32 years) were screened to exclude upper limb or cervical spine pathologies, rheumatological or neurological conditions. In each subject the nerve bed length was estimated, from the C6 spinous process to the tip of the index finger (mean = 97.0 cm (SD, 5.9)), and used to normalise the ultrasound transducer position between individuals. All measurements were taken from the right upper limb only.

The median nerve was imaged in longitudinal section in the forearm during forward head position, trunk flexion and protraction and in the forearm and upper arm during contralateral neck side flexion (CNSF). Each movement was repeated three times, including some reverse trials.

For each procedure the range of movement was determined from pictures obtained using a digital camera. Changes in joint angle and distance for each movement were determined from skin surface markers and measured using either CorelDraw (Kodak Digital Science, USA) or "tpsDig" (F. James Rohlf, Department of Ecology and Evolution, State University of New York). Measurements for the individual components of slumped sitting are summarised in figure 1. The posterior-anterior shift of the acromion was used as a measure of protraction during CNSF.

Strain is defined by the difference in the amount of elongation that occurs at two points along a nerve divided by the distance between these two points. In practice strain was determined by using regression lines fitted to plots of nerve movement against the distance along the arm. Note that the strain estimates represent the additional strain produced by the movement rather than the total nerve strain.

Comparisons of nerve movement and strain in scapulothoracic neutral and protraction during CNSF were performed using paired t-tests.

In 11 subjects the median nerve moved in a proximal direction during 35° CNSF when the scapulothoracic joint was in neutral, as reported previously [6]. The movement increased at the more proximal location (scapulathoracic neutral, mean in upper arm = 2.3 mm (SEM, 0.2) and forearm = 1.5 mm (SEM, 0.2)). With the shoulder protracted, there was a 60% reduction in nerve movement in both upper arm and forearm locations (p < 0.05 for both locations) (mean in upper arm = 0.9 mm (SEM, 0.2) and forearm = 0.6 mm (SEM, 0.1)) (figure 5). The mean extent of protraction was 48.0 mm (SEM, 4.3). The additional strain on the median nerve was 0.3% (SEM, 0.1) in scapulothoracic neutral. There was a significant reduction in strain in protraction (0.1% (SEM, 0.1); p < 0.05, paired t-test).

Median nerve excursion was measured in 4 subjects with simultaneous measurement of CNSF. The results revealed no obvious delay in the onset of nerve movement in protraction compared to scapulothoracic neutral. Despite less movement in protraction, the pattern of nerve movement mimicked that observed in scapulothoracic neutral (figure 6).

Nine of 11 subjects reported paraesthesia in the distribution of the median nerve dermatone once the shoulder girdle was sustained in protraction. The onset of symptoms ranged from 1 to 4 minutes. Symptoms disappeared when the shoulder was repositioned in scapulothoracic neutral.

Nerves are designed to slide and stretch to accommodate joint movement. Using the method of Dilley et al. [5, 6], median nerve sliding was examined during the individual components of slumped sitting. Both forward head position and trunk flexion produced only minimal nerve movement in the forearm. The only examined component to produce substantial nerve movement was shoulder protraction. The median nerve strain in the forearm with protraction was 0.7%, which was well below the limits that cause changes to nerve function (reviewed in Grewel et al. [7]).

As the shoulder girdle is protracted there is a delay in nerve movement, which is followed by a steady increase in nerve excursion. During this initial toe region the median nerve appears bowed in the upper arm. The nerve trunk appears to straighten as the range of protraction progresses. It therefore seems that with the upper limb in scapulothoracic neutral and the glenohumeral joint in 90° flexion, the median nerve is unloaded. If this is the case, the strain value of 0.7% will represent the total strain.

The results for CNSF provide evidence for a possible restriction within the shoulder region during shoulder protraction. With the shoulder protracted there was a 60% reduction in the transmission of nerve movement through the upper limb. Consistent with a reduction in movement, there was also significantly less strain in the forearm. The possibility that the nerve becomes unloaded when the shoulder is protracted is unlikely since it had been found that protraction itself causes some median nerve stretch. In addition, there was no obvious delay in nerve movement in response to CNSF (Figure 6). The evidence for a restriction is consistent with previous suggestions that shoulder protraction may cause a neurovascular impingement within the shoulder region resulting in pain [2, 8]. This suggestion was further supported by the experience of paraesthesia within the median nerve distribution during sustained protraction in 82% of subjects. These symptoms indicate the presence of a vascular restriction, which in turn affects neural function.

Scapular protraction is a complicated movement, often resulting in the combined movement of numerous other structures within the shoulder girdle, including anterior displacement of the head of the humerus. It is therefore difficult to establish the precise cause of a neurovascular entrapment. Shortening of pectoralis minor and the downward displacement of the coracoid process might affect sliding of the cords of the brachial plexus. Alternatively, elevation of the first rib during full protraction (due to its soft tissue attachments with surrounding structures) might reduce the space between the clavicle and the first rib, restricting nerve sliding.

The components of slumped sitting (i.e. forward head position, trunk flexion and protraction) are associated with poor posture [2, 9‐11], and are often adopted by office workers. Shoulder protraction is the only component of this posture to tension the median nerve, although the level of nerve strain in the forearm with the shoulder at 90° flexion and elbow extension, is not sufficiently high to result in direct neural injury.

Problems are more likely to result from local effects of shoulder protraction on the chords of the brachial plexus. The present study shows that protraction restricts nerve sliding through the shoulder region. Most subjects also experienced paraesthesia when maintaining shoulder protraction plus elbow extension and shoulder abduction. Therefore, sustained shoulder protraction may place the median nerve at enhanced risk of injury and possibly cause a vascular compromise. This may in turn explain for the trend that a high number of NSAP patients have poor shoulder posture. (e.g. [2]).