“…We restricted analyses to adult individuals free from reported major pathology, specifically those that could have affected growth (including cribra orbitalia, porotic hyperostosis and indications of leprosy) or prevented accurate measurement (including femoral fractures, osteomalacia and severe osteoarthritis). Ultimately, from approximately 4000 published aDNA samples (Antonio et al, 2019; Brace et al, 2019; de Barros Damgaard et al, 2018; Fernandes et al, 2018; Fu et al, 2016; Furtwängler et al, 2020; González‐Fortes et al, 2017; Krzewińska et al, 2018; Lipson et al, 2017; Margaryan et al, 2020; Martiniano et al, 2016; Mathieson et al, 2015, 2018; Mittnik et al, 2018, 2019; Narasimhan et al, 2019; Olalde et al, 2018, 2019; Schiffels et al, 2016; Schroeder et al, 2019; Sikora et al, 2017), we compiled metric data for 182 individuals (Alciati, 1967; Alpaslan‐Roodenberg, 2001; Andrews & Thompson, 2016; Auerbach, 2004; Auerbach & Ruff, 2004; Berthold et al, 2008; Boroneant, 2010; Caffell & Holst, 2012; Cairns, 2015; Dunwell, 2007; Fokkens et al, 2017; Frei et al, 2019; Kitti, 2008; Kjellström, 2005; Köhler et al, 2017; Malmström et al, 2019; Massy, 2018; Pardini, 1977; Price et al, 2016; Rosenstock et al, 2019; Saag et al, 2020; Schiffels et al, 2016; Szczepanek, 2013; Tebelskis & Jankauskas, 2002). We removed 28 samples with more than 95% missing genetic data and one individual with an unusually short femur (AITI_95, see below), bringing the sample size to 153.…”