The frequency of restricted range of movement in individuals with self-reported shoulder pain: results from a population-based survey.

Prevalence studies suggest that shoulder pain is very common (7-20%) in the adult population, though little is known about the severity and impact of such pain. Disability results, in part, from the restriction of movement and we therefore determined the frequency of restricted shoulder movement in individuals in a general population sample reporting shoulder pain and associated disability. In all, 232 individuals were interviewed about shoulder pain and related disability, and their range of shoulder movement in the following planes was measured: elevation, forward flexion, backward flexion, external rotation and internal rotation, together with the minimum difference achieved between the tip of a thumb and the spinous process of C7. In total, 48 (21%) subjects reported current pain and disability in one or both shoulders. Using receiver operating characteristic (ROC) analysis, cut-offs for restricted movements were selected at the point of maximal discrimination between painful and pain-free shoulders. There was considerable overlap in the distribution of range of movement, at all six sites, between these two groups. The highest positive predictive value for pain was observed in restriction of the thumb to cervical spine movement (53%), but this also had the lowest negative predictive value. By contrast, the highest negative predictive value was observed for restriction of external rotation (96%). Restriction in any plane was observed in 77% of those individuals with pain, but was also present in half of those without pain. In summary, most people with self-reported shoulder pain in the community do not have widespread restriction of movement. Reduction in external rotation was the most discriminatory, but it is necessary to examine movement in multiple planes to assess the true burden of shoulder pain in the community.