Year 2: First, we anticipate that the predictors (individual information and the degrees of fatigue and symptom distress, exercise involvement, and dietary intake) and mediator (exercise involvement) of dependent variables (muscle wasting and QoL) will be identified. This will help healthcare providers to design exercise programs. Second, we anticipate that the longitudinal trends (symptom clusters and the modes of the recovery trajectories of cancer symptoms, muscle wasting, and QoL) can be determined. We thus can identify the patients who are in the risk of muscle wasting and design appropriate and effective interventions for patients who need the supportive care.