Fig 1. Population-wide and within-family associations between violent victimisation and subsequent common psychiatric disorders, suicidal behaviours, and premature mortality in Finland and Sweden.

Notes: The “Crude” model adjusted for sex and birth year. The “Adjusted” model additionally accounted for birth order, parental immigrant background, low family income at offspring birth, single mother at offspring birth, parental psychiatric, and violent crime histories at offspring birth, as well as pre-victimisation psychiatric disorders, alcohol use disorders, drug use disorders, and self-harm. The “Sibling-comparison” model refers to within-family estimates comparing biological full-siblings differentially exposed to violent victimisation and is adjusted for all time-invariant unmeasured familial confounders shared between the siblings as well as the following measured confounders that vary within families: sex, birth year, birth order, and parental characteristics at birth (low family income, single mother, psychiatric history, and violent crime history) as well as any preexisting psychiatric disorders and self-harm events. CI refers to confidence intervals.