In this paper， an event-triggered model predictive control scheme based on a variable horizon strategy is proposed for constrained continuous linear time-invariant systems with additive disturbances. Firstly， an event-triggered mechanism without Zeno behavior is designed based on the deviation between the optimal state trajectory and the actual state trajectory to reduce the frequency of solving optimization problem. Next， in order to reduce the computational complexity of solving optimization problem when the actual state approaches the terminal set， a more efficient adaptive prediction horizon update mechanism in the form of exponential shrinkage is designed. Then， based on the dual-mode control strategy， an adaptive event-triggered model predictive control algorithm is proposed， and sufficient conditions are provided to ensure the feasibility of the algorithm and the stability of the closed-loop system. Finally， the effectiveness of the proposed algorithm is verified based on a mass-spring-damper system， and the results show that the proposed scheme can effectively reduce system resource consumption and computational complexity for solving optimization problem without losing control performance.