Experimental shoulder pain models do not validly replicate the clinical experience of shoulder pain

Background and aims People with shoulder pain often present with abnormal shoulder muscle function. It is not known whether shoulder pain causes or is the result of muscle dysfunction. If pain leads to muscle dysfunction, therapeutic interventions that produce shoulder pain may be contraindicated. Experimentally induced nociception can be used to investigate a causal relationship between shoulder pain and muscle dysfunction. However, the validity of current experimental shoulder pain protocols has not been established. The aim of this study was to determine whether current experimental shoulder pain protocols validly replicate the clinical experience of shoulder pain with respect to pain distribution, quality and behaviour. Methods Nine pain free participants received two injections of hypertonic saline, one into the subacromial space and one into supraspinatus, in random order, at least 1 week apart. Investigators blind to the injection site assessed pain distribution, pain response to clinical tests which provoke shoulder pain and pain quality assessed using the McGill Pain Questionnaire. Results Following hypertonic saline injection into both the subacromial space and supraspinatus: pain was most commonly reported in the deltoid region and did not extend beyond the elbow; the most common response to clinical tests which provoke shoulder pain was a decrease in pain; and the highest rating of pain quality was in the sensory domain with very few responses in the affective domain. Conclusions Experimental shoulder pain induced by injection of hypertonic saline into either the subacromial space or supraspinatus produced a pain distribution similar to that observed in clinical shoulder pain, but neither experimental pain protocol could reproduce the increases in pain intensity following shoulder provocation tests or the emotional distress commonly observed in people with clinical shoulder pain. Implications Pain induced by local shoulder nociception produced by hypertonic saline injection into shoulder structures has significant limitations as a model of clinical shoulder pain. While it is perhaps unsurprising that short duration, chemically-induced experimental pain does not replicate the quality of the clinical experience of shoulder pain, the validity of experimental shoulder pain models which produce the opposite response to provocation testing to clinical shoulder pain must be questioned.