“…The overall quality of evidence was low, with only one study rated at low risk [40] and the remaining rated at serious (n = 4) [27,29,30,37,39,41] to critical (n = 3) risk of bias [28,31,38]. RoB was particularly high for the confounding domain where three studies were judged at critical risk [28,31,38], four at serious [27,29,37,41] and two at moderate risk [30,39] and one at low risk [40]. RoB for the selection of participants was mostly moderate (n = 4) [29,30,37,39], with four studies rating at low [27,37,40,41], one at serious [31] and the remaining one at critical risk [28].…”