Risk factors of hemiplegic shoulder pain in stroke patients

Objectives: This study is designed to investigate the degree of correlation of hemiplegic shoulder pain to some clinical parameters including electrodiagnostic testing. Methods: Fifty post-stroke patients were divided into 3 groups: Group A patients had spontaneous hemiplegic shoulder pain, Group B patients had shoulder pain only during movement, and Group C patients had no shoulder pain. The correlational relationships between shoulder pain in the hemiplegic side and some clinical parameters including age, sex, side affected, duration of stroke, motor function, sensory impairment, spasticity, shoulder subluxation, shoulder contracture, peripheral nerve conduction, and the presence of spontaneous electromyographic [EMG] activity was analyzed statistically. Results: Data analysis revealed no significant correlation between the hemiplegic shoulder pain and age, sex, side affected, or duration of stroke. However, the presence of shoulder subluxation or limitation of passive shoulder range of motion [ROM] showed significant correlation to the hemiplegic shoulder pain. The mean median nerve motor nerve conduction velocity, which was significantly slower in the hemiplegic side than in the normal one, and the mean median sensory nerve conduction, which was normal in the hemiplegic side as compared to the normal side, were not significantly correlated with the hemiplegic shoulder pain. F-wave conduction was prolonged in patients with shoulder pain [Groups A and B], but was normal in patients without pain [Group C]. In needle EMG examination, spontaneous activity was observed in 94% of Group A patients, in 71% of Group B patients, and only 28% of Group C patients [P < 0.001]. There was a negative correlation between spontaneous EMG activity and muscle strength in the hemiplegic arm [P < 0.025]. Conclusions: There is a significant assocation of hemiplegic shoulder pain with shoulder subluxation, restricted passive ROM, slow F-wave conduction, and the presence of spontaneous EMG activities in stroke patients. Hemiplegic shoulder pain may be predicted and prevented or treated in an early state if there is any evidence of shoulder subluxation, restricted passive ROM of shoulder, slow F-wave conduction, or spontaneous activities in EMG examination in a stroke patient.