The maintenance decision making considering various types of coupling effect within system components has been a focus in the field of operation and maintenance (O and M) due to its wide applications in engineering systems with potentially significant economic benefits. Various opportunities as the coupling effect can provide, it leads to the difficulty in system modeling and policy establishment when multiple types of dependence exist in the system. This paper presents a novel approach to reflect economic dependence and stochastic dependence in the Markov model so that an optimal maintenance policy can be selected through policy performance comparison. Failure rates of system components change according to the health states of related components and combined maintenance actions introduce O and M time and cost savings. A case study has been used to demonstrate the developed method and further illustrate the benefit of combined maintenance actions in different O and M scenarios considering failure couplings among system components.