“…First, the association between PE and suicidal ideation and behaviour has been investigated mostly in cross-sectional studies (Capra et al, 2015;DeVylder et al, 2015b;Jang et al, 2014;Kelleher et al, 2012c;Nishida et al, 2010;Saha et al, 2011;Temmingh et al, 2011), disabling causal inferences. Second, several studies included psychiatric diagnoses in the regression models to control for confounding (Calkins et al, 2014;DeVylder et al, 2015b;Fisher et al, 2013;Koyanagi et al, 2015;LewisFernandez et al, 2009;Olfson et al, 2002;Saha et al, 2011;Sharifi et al, 2015). However, it is conceptually impossible to examine a confounder in the association between PE and suicidal behaviour when that confounder, for example depression, is fundamentally associated with the outcome (Miller and Chapman, 2001).…”