In Machine-to-Machine (M2M) communications, machines are wirelessly connected to accomplish collaborative tasks without human intervention, and provide ubiquitous solutions for real-time monitoring. The real-time monitoring application is one of the killer applications for M2M communications, where M2M nodes transmit sensed data to an M2M gateway, and then the M2M gateway can have real-time monitoring for each sensing region. In real-time monitoring application, the energy consumption for the M2M nodes to send sensed data to the M2M gateway is an important factor that significantly affects the performance of the system. In this paper, we first consider the energy consumption as well as the validity of sensed data to design either centralized or distributed energy-efficient reporting mechanisms. We then analyze the complexity of the reporting mechanisms. Simulation experiments are conducted to investigate the performance of the proposed mechanisms, and show that the distributed mechanism outperforms the centralized mechanism when the M2M nodes are mobile.