Predictors of female criminality:: Findings from the northern Finland 1966 birth cohort

Objective: Females still commit fewer criminal offenses than males, but the percentage of female offending has been increasing during the past few decades. Thus there is a need for original studies into the perinatal contribution to the etiology of female offending. Method: A large, prospectively collected birth cohort database of female members (N = 5,056) was available. Information on perinatal biological and psychosocial risks as well as data from the National Crime Registers up to 32 years of age were collected and analyzed by logistic regression and a chi(2) automatic interaction detection (CHAID) analysis. Results: The absence of the father during childhood was the strongest risk factor in predicting female criminality (odds ratio 2.5; 95% confidence interval 1.4-4.3). Furthermore, in the families in which the father was present, maternal smoking during pregnancy together with being born unwanted correlated with an increased prevalence of criminal offending significantly up to 7.2%. Conclusions: CHAID analysis proved to be a useful statistical method in predicting female adult criminality after preceding perinatal risks. It revealed that the risk factors were mainly familial, the paternal factor being the most important one in determining the probability of daughters committing criminal offenses.