The traditional centralized optimization method of distribution networks has encountered heavy communication burden and large amount of computational cost. Moreover, it is difficult to meet the privacy protection requirements of multiple stakeholders in distribution networks. New distributed optimization methods for distribution networks are urgently needed. This paper proposes a distributed optimization method for distribution networks considering data privacy protection. Firstly, a distributed optimization model of distribution network is constructed based on the neighbor-side cluster coordination algorithm, and the global optimization solution problem is transformed into a distributed interactive solution problem. Then, the homomorphic encryption algorithm is employed to encrypt the boundary variables of the model, which strengthens the data barrier between multiple regions of the distribution network and protects the privacy of the interactive data. Finally, the proposed method is verified on a practical example, and the feasibility and effectiveness of the proposed method are illustrated.