In an object-oriented database, classes are organized according to three types of relationship, namely generalization, association and aggregation relationships. These relationships, which impact the query language in different ways, should be considered if we want to design a suitable indexing structure for an object-oriented database. Many researchers focus on generalization and association relationships when trying to find indexing structures for the OODB. For the aggregation hierarchy, there is no deep research in this field. Though we can use the same indexing architectures as have been proposed for association relationships, we must consider the special features of the aggregation relationship. These particular features of aggregation relationships are discussed in this paper in detail considering both the similarities and differences with the association relationship. Next the paper presents index nesting techniques to support aggregation relationships in composite objects. A comparison between this technique and other indexing techniques is also discussed in this paper.