Sex differences in the distribution of entheseal changes: Meta-analysis of published evidence and its use in Bayesian paleopathological modeling

Objectives We studied the sex differences in the distribution of entheseal changes (EC) in an archeological population through a Bayesian approach that allows incorporating existing knowledge while controlling for confounder factors that may affect EC development. Materials and methods We performed a meta-analysis of published research on sex differences in EC frequencies from archeological populations. Also, EC were assessed for fibrocartilaginous entheses following the "New Coimbra Method" in a Spanish population that dates from the 15th to the 18th century. Data were analyzed with multivariate generalized linear mixed models (MGLMM). Results Meta-analysis showed a consistent but small effect of males usually manifesting higher EC frequencies. Similarly, our MGLMM analysis showed that bone formation and erosion is unequally distributed in the archeological population we studied, with bone formation more present in male lower limbs and erosion more frequent in male upper limbs. Discussion Bayesian inference makes it possible to assess more complex models than traditional frequentist methods, and can be informed by meta-analysis to reflect the current state of knowledge on any given topic. MGLMM are an appropriate technique for the study of EC as they can accommodate several response variables in a single model, controlling for well-known confounders of EC formation to infer sex differences that could be attributed to daily behavior.