3.1 Characteristics of the included studies
The 19 included literatures are all in English, and a total of 370 animals were included, including 180 animals in the experimental group and 190 animals in the control group (Figure 3) . Animal species include Sprague-Dawley rats(SD rats)(Y. Deng et al., 2016; J. Guo et al., 2010; Xiong et al., 2016; F. Zhang et al., 2015; L. Zhang et al., 2017), C57BL mice(H. Li et al., 2018; Shao et al., 2015; Tao et al., 2013; Wu et al., 2011; Xie et al., 2018; C. Q. Xu et al., 2010), BALB/c mice(L. Li et al., 2014; P. Ma et al., 2015; Wei et al., 2015; J. Zhou et al., 2011), Wistar rats(Fu et al., 2018; Y. Hu et al., 2016), MRL/lpr mice(Su et al., 2018) and Minipigs(X. Zhang et al., 2018). Inflammatory diseases studied include local inflammation(L. Li et al., 2014; Wu et al., 2011; X. Zhang et al., 2018), circulatory system inflammation(Fu et al., 2018; Y. Hu et al., 2016), urinary system inflammation(P. Ma et al., 2015; Su et al., 2018; Xie et al., 2018; L. Zhang et al., 2017), nervous system inflammation(Y. Deng et al., 2016; J. Guo et al., 2010; H. Li et al., 2018; Xiong et al., 2016), respiratory system inflammation(Wei et al., 2015; C. Q. Xu et al., 2010), Digestive system inflammation(Tao et al., 2013; F. Zhang et al., 2015), Motion system inflammation(Shao et al., 2015) and tumor-associated inflammation(J. Zhou et al., 2011) (Table 1).
According to the inclusion and exclusion criteria, all animal models were models of inflammation-related diseases. The experimental group received monotherapy with icariin or its derivatives, and the control group received only normal saline or distilled water. In order to evaluate the effect of icariin and its derivatives on the treatment of inflammatory diseases, 12 literatures reported the level of TNF-α(Fu et al., 2018; J. Guo et al., 2010; Y. Hu et al., 2016; H. Li et al., 2018; L. Li et al., 2014; Shao et al., 2015; Su et al., 2018; Tao et al., 2013; Wu et al., 2011; Xie et al., 2018; C. Q. Xu et al., 2010; F. Zhang et al., 2015), 10 literatures reported the level of IL-1β(Y. Deng et al., 2016; Fu et al., 2018; J. Guo et al., 2010; H. Li et al., 2018; Shao et al., 2015; Su et al., 2018; Xie et al., 2018; Xiong et al., 2016; L. Zhang et al., 2017; X. Zhang et al., 2018), and 6 literatures reported the level of IL-6(Fu et al., 2018; Y. Hu et al., 2016; Shao et al., 2015; Wei et al., 2015; Xie et al., 2018; F. Zhang et al., 2015), 5 literatures reported the level of TGF-β1(Y. Deng et al., 2016; Fu et al., 2018; Wei et al., 2015; Xiong et al., 2016; L. Zhang et al., 2017), 5 literatures reported the level of SOD(Y. Hu et al., 2016; H. Li et al., 2018; P. Ma et al., 2015; Xie et al., 2018; F. Zhang et al., 2015), 4 literatures reported the level of MDA(Y. Hu et al., 2016; H. Li et al., 2018; P. Ma et al., 2015; Xie et al., 2018), 5 literatures reported the level of IκBα(Y. Deng et al., 2016; Fu et al., 2018; L. Li et al., 2014; Xiong et al., 2016; L. Zhang et al., 2017), 4 literatures reported the level of Bcl-2 Level(H. Li et al., 2018; P. Ma et al., 2015; Xie et al., 2018; F. Zhang et al., 2015), 3 literatures reported the level of IFN-γ(Tao et al., 2013; X. Zhang et al., 2018; J. Zhou et al., 2011), 2 literatures reported the level of Bax-2(H. Li et al., 2018; Xie et al., 2018), 2 literatures reported the level of Caspase-1(Su et al., 2018; L. Zhang et al., 2017), 2 literatures reported the level of PPARα(Y. Deng et al., 2016; Xiong et al., 2016), and 2 literatures reported the level of PPARγ(Y. Deng et al., 2016; Xiong et al., 2016), 2 literatures reported the level of NLRP3(Su et al., 2018; L. Zhang et al., 2017), and 2 literatures reported the level of p65 NF-κB(Y. Deng et al., 2016; L. Zhang et al., 2017).
3.2 Study quality
Among the 19 articles included in the study, all articles describe methods for randomizing animals into experimental and control groups. 6 articles described possible prognostic features and animal factors, 16 articles described methods of concealing the order of allocation, all studies described random allocation of animal rooms, and 16 articles reported blinding of trial participants, 12 articles described random selection of animals for outcome assessment, 15 articles blinded outcome assessment, and 8 articles reported the information of experimental attrition and sample exclusion, and 10 articles reported the completeness of the data, and no other sources of bias and selective reporting were found in these 19 articles(Figure 4).