Energy consumption has become a key issue in the design of communication systems. Indeed, both for economic and green reasons, the concept of energy saving appears at an early stage of projects, fully integrating the list of expected performance, as well as throughput or security. In multihop wireless networks, several energy-aware approaches have been proposed with specific goals, such as network's lifetime or stability. To reduce the global energy consumption of such networks, we proposed in a previous work an optimal solution with interference consideration, based on mixed integer linear programming, to route a set of flows over a minimal number of nodes. Thus, without degrading flow rates, inactive nodes can be put in a sleep mode or be turned off to maximize energy savings. Indeed, the energy consumption related to communication is just a fraction of the total consumption of a node In this article, we are going a step further, by providing a metric to efficiently aggregate flows with classical shortest path algorithms. Thanks to a theoretical comparison and simulations, we demonstrate the significant gains can be obtained with our approach. Finally, we discuss on ways to implement our solution in existing routing protocols, in a fully distributed manner, and other considerations that may need to take on.