“…Studies were generally of acceptable quality with quality ratings ranging from 5 to 9 on a ten point scale (see Supplemental Table 4 for an overview). Five studies (Fahy, Stansfeld, Smuk, Lain, van der Horst, Vickerstaff & Clark, 2017;Fergusson, McLeod & Horwood, 2013;Harkonmäki, Korkeila, Vahtera, Kivimäki, Suominen, Sillanmäki & Koskenvuo, 2007;Mullen, Martin, Anderson & Romans, 1994;Mullen, Martin, Anderson, Romans & Herbison, 1996;Vinnerljung, Sundell, Löfholm, & Humlesjö, 2006) scored five, five scored six or seven (Covey, Menard & Franzese, 2013;Lund, Andersen, Winding, Biering, & Labriola, 2013;Pinto Pereira, Li & Power, 2017;Strøm, Thoresen, Wentzel-Larsena, Hjemdal, Lien & Dyb, 2013;Tanaka, Jamieson, Duku, Boyle & MacMillan, 2011) and two scored eight (Currie & Widom, 2010) or nine (Mersky & Topitzes, 2010). Study sample size ranged from 492 to 11,874, with an average of c4250 participants.…”