Invited Commentary: Benefits of Heterogeneity in Meta-analysis of Data from Epidemiologic Studies | Synapse