The unsupervised hierarchical analysis showed three clusters within the patients. One subgroup (n = 18; proximal arm type) corresponded to patients with higher IntraMAT content of proximal arm muscles; supraspinatus (32.7 ± 4.3%, p < 0.01), deltoid (28.5 ± 4.4%, p < 0.01), infraspinatus (28.6 ± 3.9%, p < 0.001), subscapularis (37.4 ± 4.7%, p < 0.01), biceps (31.0 ± 3.5%, p < 0.001), and triceps (47.8 ± 5.1%, p < 0.001). The second subgroup (n = 11; upper leg type) corresponded to patients involving IntraMAT content mainly in quadriceps muscles. The third subgroup (n = 16; dysphagic type) corresponded to patients with dysphagia having cricopharyngeal bar (100%, p < 0.001), high SDQ scores (16.8 ± 2.3, p < 0.001), less IntraMAT content of rectus femoris (1.4 ± 0.7%, p < 0.01), vastus lateralis (23.5 ± 6.6%, p < 0.05), vastus intermedius (13.9 ± 4.7%, p < 0.05), vastus medialis (9.5 ± 3.7%, p < 0.01), supraspinatus (12.4 ± 3.5%, p < 0.01), deltoid (5.5 ± 1.4%, p < 0.001), infraspinatus (7.6 ± 2.1%, p < 0.001), subscapularis (11.6 ± 3.9%, p < 0.01), biceps (6.6 ± 1.8%, p < 0.001), and triceps (13.1 ± 3.0%, p < 0.001).
This study indicates the clinical subsets of IBM to expand the knowledge of the heterogeneity of the patients.