An Insertion-Based Anatomical Reappraisal of the Lumbar Multifidus: Evidence From Gross Anatomical Dissection and...

Lumbar multifidus (LM) muscle dysfunction is a common cause of low-back pain (LBP). Training the LM muscles improves LBP; however, understanding the shape, muscle bundle structure, and origin-insertion of the LM muscle is essential for designing an effective training program. Although the LM muscle anatomy has been largely explored, controversies persist regarding its shape, muscle bundle structure, origin, and insertion, and a detailed understanding of its anatomy remains lacking. The current study sought to elucidate the shape and bundle structure of the LM muscle, focusing on the muscle insertion, to provide a scientific basis for developing training methods for the LM muscle. The formalin-fixed remains of nine cadavers (five males and four females; mean age at death, 79.6 ± 12.9 years) were examined for gross anatomical observation of the origin and continuity of the insertion of the LM muscle. The attachment area was also observed and recorded, with the periosteum remaining. The dissection findings were corroborated by the histological observations of the origin and muscle bundle structures of the two donors. In addition, ultrasonographic evaluation of the right LM at the level of the L5 spinous process was performed in four healthy adult males and two healthy adult females (mean age: 47.3 ± 16.0 years), and the findings were compared with the gross anatomical observations. Dissection revealed two muscle bundles with unique running patterns in the LM that inserted at one of the spinous processes in the lumbar vertebrae. One muscle bundle originates from the inferior lumbar mammillary process and the lateral sacral crest. The other is a pennate muscle that originates from two directions, with some muscle bundles originating from the thoracolumbar fascia. In this study, the LM muscle was dissected from its insertion, according to the fiber run, and its shape, muscle bundle structure, origin, and insertion were clarified, which have been reported inconsistently in previous studies. These results may facilitate training and evaluating the LM muscle.