Confidentiality of information should be preserved in outsourcing data. To this purpose, encryption techniques are currently used in many scenarios where an overlying layer of encryption is applied on data before outsourcing. Dealing with encrypted data, however, highly reduce query evaluation efficiency. Also, they are based on the assumption that outsourced data are all equally sensitive, and therefore encryption is a price to be paid for their protection. However, in many scenarios, what is sensitive is the association among data, more than data items per se. In recent years, novel proposals resort to fragmentation as a way to break sensitive associations among attributes while using encryption as an underlying measure for making data unintelligible. But those techniques have problems that the database capacity increase due to duplicate encryption data of attributes in order to improving query evaluation efficiency. In this paper, we address this issue and propose a fragmentation and encryption scheme to protect privacy in outsourcing data with reducing storage load.