Mobile ad hoc network (MANET) consists of a group of self-operated mobile nodes that form a network without considering predefined infrastructure. In the network, each node poses less energy due to battery-operated sensor nodes. Numerous methods are devised for routing, but each method suffered from the energy constraint. This paper proposes an energy-efficient multipath routing protocol for exploiting different routes between source and destination to mitigate energy constraints. The key idea is to determine the optimal path from a set of paths available between the source and the target node. To improve the security in routing protocol, the factors, like energy, trust, and mobility are considered as major components. Here, the trust is modeled by considering trust factors, like average encounter rate (AER), successful cooperation frequency (SCF), direct trust, integrity factor, and forwarding rate factor. Moreover, secure routing is performed between the nodes using the proposed Adaptive-Atom Whale Optimization algorithm (Adaptive-AWOA). The proposed Adaptive-AWOA) is designed by incorporating the Adaptive concept in the Atom Whale Optimization algorithm (AWOA). The fitness function is newly modeled considering mobility, energy, and trust factors. The proposed Adaptive-AWOA outperformed the other methods with a minimal end-to-end delay of 0.009, maximal throughput of 85.802% and maximal packet delivery ratio (PDR) of 97.391%, respectively. © 2020 World Scientific Publishing Company.