This study has many strengths and some notable limitations. Strengths include the high participation rate of the large population-based sample, the 8–10 years of follow-up time between aggression ratings in early adolescence and psychiatric data collection in young adulthood, and the fact that attrition was not selective on aggression ratings. Additionally, while the dataset is comprised of twins, previous studies have shown the twins to be representative of the population in general, being no different than their singleton peers regarding externalizing and internalizing problems [39]. Important to note, however, is that while our data came from a population-based sample, the current study sample was derived from a more intensively studied subset of that population. Furthermore, while this is a sample based on a general population, it is necessary to understand the culture and context of the population utilized. Finland has a low incarceration rate compared to most of the world, and has had specific policies in place for decades to reduce the adolescent prison population. Thus, our sample may be comprised of a slightly different mix of individuals than countries
