This paper presents methods of decomposition of radiography into multiple individual muscle and bone structures. While existing solutions require the dual-energy scan for the training dataset and are mainly applied to structures with high-intensity contrast such as bones, we focused on multiple superimposed muscles with subtle contrast, in addition to bones. The decomposition problem is formulated as an image translation problem between 1) a real x-ray image and 2) multiple digitally reconstructed radiographs, each of which contains a single muscle or bone structure, and solved by unpaired training based on the CycleGAN framework. The training dataset was created by automatic computed tomography (CT) segmentation of muscle/bone regions and virtually projecting them with geometric parameters similar to the real x-ray images. Two additional features were incorporated in the CycleGAN framework to achieve a high-resolution and accurate decomposition: hierarchical learning and reconstruction loss with the gradient correlation similarity metric. Further, we introduced a new diagnostic metric for muscle asymmetry directly measured from a plain x-ray image to validate the proposed method. Our simulation and real image experiments using real x-ray and CT images of 475 patients with hip disease suggested that each additional feature significantly enhanced the decomposition accuracy. The experiments also evaluated the accuracy of muscle volume ratio measurement, which suggested a potential application to muscle asymmetry assessment from an x-ray image for diagnostic and therapeutic assistance. The improved CycleGAN framework was able to apply for the decomposition of musculoskeletal structures from single radiography.