Although innervation level of the glenohumeral joint was reported, innervation level of subacromial bursa was not clarified. The purpose of this study was to evaluate the dorsal root ganglion neurons innervating the glenohumeral joint and subacromial bursa. In this manner, dichotomizing axons projecting to the glenohumeral joint and subacromial bursa were also evaluated. 6 Wister rats were used in this study. 2 kinds of neurotracers such as fluorogold (FG) and dioctadecyl-tetramethyl-indocarbocyanine-perchlorate (DiI) were used to detect the glenohumeral joint and subacromial bursa separately. After anesthesia, FG was injected into the glenohumeral joint and DiI were injected into the subacromial bursa. Dorsal root ganglia were harvested between C1 and T1 7 days after injection. The number of FG labeled-neurons, DiI-labeled neurons and double-labeled neurons were counted. FG-labeled neurons were distributed from C2 to C7. The glenohumeral joint was mainly innervated by C4 and C5 dorsal root ganglion neurons. DiI-labeled neurons were distributed from C2 to C7. Although subacromial bursa was innervated from C2 to C7 dorsal root ganglion neurons, there was no main innervation level for subacromial bursa. Dorsal root ganglion neurons which double-labeled with FG and DiI were 19.0 ± 6.4% of FG-labeled neurons and 20.3 ± 6.2% of DiI-labeled neurons. Dichotomizing axons projecting to the glenohumeral joint and subacromial bursa were approximately 20% of dorsal root ganglion neurons. The glenohumeral joint was innervated from C2 to C7 and subacromial bursa was innervated from C2 to C7 dorsal root ganglion neurons. Obscurity of shoulder pain might be caused by which dichotomizing axons projecting to the glenohumeral joint and subacromial bursa approximately 20% of both sides. Injection into either the glenohumeral joint or subacromial bursa would be somewhat effective with the existence of dichotomizing axons.
