This review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [16]. The review protocol for this study was prospectively registered at PROSPERO (registration number 179382).