Year: 2002 Source: Acta Psychiatrica Scandinavica, v.106, no.5, (November 2002), p.365-372 SIEC No: 20080778

Birth cohort effects in New South Wales suicide data, 1865-1998, were examined. Aggregate suicide data were examined after controlling for age & period using descriptive statistics & Poisson regression modelling. 3 approaches were applied to address the non-identification of age, period, & cohort in age-period-cohort models. Apparent cohort effects in male & female suicide using a minimum constraints age-period-cohort model were no longer present in female suicide when a surrogate period age-period-cohort model was used, & were absent in both male & female suicide when identifiable cohort effects were isolated using a drift adjusted approach to age-period-cohort modelling. It is concluded the increases in youth suicide occurring since the 1960s are due to causes operating contemporaneously. (38 refs.)