Childhood risks for young adult symptoms of personality disorder: Method and substance

This article employs data gathered prospectively over a 17 year period to assess 36 childhood risk factors for personality disorder in young adults. Measurement issues are discussed, and the overall extent of prediction is assessed in younger and older boys and girls. Prediction is shown to be better for girls than for boys and, among girls, better over a younger age span than over an older age span covering an equivalent time. Risks are shown to predict independently of the continuity of symptoms over time. The effects of individual risks are examined by means of a) correlation coefficients, b) standardized mean differences, c) odds ratios for scaled measures, and d) attributable fractions.