“…Consistent with this view, we estimated less mediation for the quadratic effect of sleep duration on depression and anxiety risk in the university sample, suggesting the mediation effect in students is stronger for short sleep duration compared to long sleep. Sleep debt and daytime sleepiness can decrease certain behaviors (e.g., physical activity, leisure activities) (McAllister et al, 2009;McClain, Lewin, Laposky, Kahle, & Berrigan, 2014) and alter neurochemical (e.g., serotonin and norepinephrine) (Chellappa, Schroder, & Cajochen, 2009;Walker & van der Helm, 2009) and neuroanatomical processes (e.g., amygdala activity) (Motomura et al, 2013(Motomura et al, , 2014 Table S5 and Table S6). Social activity can also influence chronotype (Leonhard & Randler, 2009), and growing research suggests that the contribution of social relative to biological (circadian and sleep) factors on eveningness may vary in different social contexts (Natale, Adan, & Fabbri, 2009;Randler, Faßl, & Kalb, 2017).…”