Cell-free massive multiple-input multiple-output (MIMO) is a promising technology to address user service inequity and frequent cell re-selection. However, most of the existing distributed detectors are based on the assumption that all users are served by each access point (AP) with simple linear detection methods, resulting in extremely high fronthaul overhead and performance degradation. In this paper, we propose a novel hybrid distributed detector for the user-centric cell-free massive MIMO system. Each AP pre-defines a load threshold to determine whether to use the low-complexity maximum ratio combining (MRC) algorithm or the high-performance expectation propagation (EP) algorithm to ensure detection accuracy while reducing the computational complexity. Moreover, we propose a sparsified EP (SEP) algorithm to further improve the detection performance, from the perspective of user interference suppression via channel sparsified transformation. Based on the properties of the distributed MRC-EP/MRCSEP detection scheme, an efficient hybrid fusion approach is proposed. Numerical results illustrate that the proposed hybrid distributed detectors achieve a good performance-complexity trade-off compared to the existing distributed detectors.