<p>This paper aims to investigate the issue of adaptive event-triggered recursive state estimation (RSE) for a class of nonlinear complex networks with periodic hybrid cyber attacks (PHCAs) and random mixed couplings (RMCs). The phenomenon of RMCs is characterized by both random outer coupling and random inner coupling, with the former modeled by a sequence of random variables uniformly distributed over a given interval and the latter by multiplicative noise. Taking into account periodic denial-of-service (DoS) attacks and periodic false data injection (FDI) attacks, a novel periodic hybrid attack model is introduced. In this model, the DoS and FDI components have their own attack cycles, each comprising an active period and a dormant period. To alleviate the communication burden, an adaptive event-triggered mechanism is adopted to regulate the data transmission frequency in the node-to-node channel. Accordingly, a novel adaptive event-triggered RSE strategy is proposed, accounting for both RMCs and PHCAs. Specifically, an upper bound on the estimation error covariance is presented, and then the estimator gain that minimizes this upper bound is constructed. Furthermore, the upper bound on the estimation error covariance is proved to be uniformly bounded in mean-square sense via rigorous theoretical analysis. Finally, simulation experiments are carried out to reveal the availability of the developed RSE scheme.</p>