Chemokines in post-traumatic stress disorder: A network meta-analysis.

• • Network meta-analyzed chemokines system in PTSD. • • Concentrations of CCL3, CCL4 and CCL5 in the PTSD were higher in the controls. • • SUCRA probabilities showed that CCL5 had the highest rank in PTSD for all chemokines. Previous studies on the association between chemokines concentrations and post-traumatic stress disorder (PTSD) yielded inconsistent results. Therefore, the purpose of this network meta -analysis was to summarize these results. The databases of PubMed, Web of Science, Psyc-ARTICLES, Embase and Cochrane Library were searched for relevant articles published not later than January 15, 2020. Then, eligible studies were selected based on predefined study selection criteria. Standardized mean differences (SMDs) with 95% confidence intervals (CIs) were calculated as group differences in chemokines concentrations. Moreover, network meta-analysis was used to rank chemokines effect values according to their respective surface under cumulative ranking curve (SUCRA) probabilities. A total of 18 eligible studies that investigated the association between 9 different chemokines and PTSD were identified. They involved 1,510 patients and 2,012 controls. Results of the meta-analysis showed that the concentrations of CCL3, CCL4 and CCL5 in the PTSD patients were significantly higher than that in the controls (SMDs of 4.12, 6.11 and 1.53 respectively). However, although not statistically significant, concentrations of CCL2 tended to be lower in PTSD patients than in the controls (SMD = -0.76); whereas concentrations of CXCL12 tended to be higher in PTSD patients than in the controls (SMD = 0.37). SUCRA probabilities showed that, among all the chemokines studied, the effect of CCL5 was the highest in PTSD patients. Concentrations of CCL3, CCL4 and CCL5 may be associated with a trauma and/or PTSD. Also, CXCL12 and CCL2 may be the underlying biomarkers for trauma and/or PTSD. Thus, future studies with large population based samples are needed to further assess these associations. In addition, future research should explore possible mechanisms underlying these associations, with the aim to develop new diagnostics for PTSD. PROSPERO CRD42019147703. [ABSTRACT FROM AUTHOR]