“…As shown in Table 1 , the methodological quality of the observational studies was rated high, with 10 studies scoring nine of nine on the NOS ( Alves Guimaraes et al, 2018 ; Celli et al, 2010 ; Chen & Liao, 2018a ; Desai et al, 2019 ; Genao et al, 2015 ; Short et al, 2012 ; Tomioka et al, 2019 ; Wang et al, 2018b ; Warnier et al, 2010 ; Xiao et al, 2019 ), six scoring eight of nine ( Hirayama et al, 2018 ; Hu & Lin, 2018 ; Koskela et al, 2014 ; Lainscak et al, 2009 ; Nadeem et al, 2015 ; Volchkova et al, 2015 ), three scoring seven of nine ( Ganga et al, 2013 ; Liao & Chen, 2017 ; Wilchesky et al, 2012 ), and one scoring six of nine ( Rusinowicz, Zielonka & Zycinska, 2017 ). The overall risk of bias was rated low; however, the included studies used different disease definitions, which influenced the population selection.…”