It is difficult to use multidisciplinary design optimization (MDO) method in the complex product optimization because of the existence of coupled factors. In order to help solve this problem, the components and affection of coupled factors are analyzed first in this paper. Then the coupled factors are divided into three classes. According to the global sensitivity equations (GSE), the properties of couplings between sub-systems are analyzed. The criterion that can confirm a redundant coupled factor has been derived and is applied to collaborative optimization by handling the couplings on the basis of coupled properties. An example is given to show the results achieved.