Background. Psychiatric co-morbidity is thought to be an important problem in suicide, but it has been little investigated. This study aims to investigate patterns of co-morbidity in a group of male suicide completers.
Method. One hundred and fifteen male suicide completers from the Greater Montreal Area and 82 matched community controls were assessed using proxy-based diagnostic interviews. Patterns of co-morbidity were investigated using latent class analysis.
Results. Three subgroups of male suicide completers were identified (L2=171·62, df=2012, P<0·05), they differed significantly in the amount of co-morbidity (Kruskal–Wallis χ2=71·227, df=2, P<0·000) and exhibited different diagnostic profiles. Co-morbidity was particularly found in subjects with disorders characterized by impulsive and impulsive–aggressive traits, whereas subjects without those traits had levels of co-morbidity which were not significantly different from those of controls (χ2=8·17, df=4, P=0·086).
Conclusions. Suicide completers can be divided into at least three subgroups according to co-morbidity: a low co-morbidity group, a substance-dependent group and a group exhibiting childhood onset of psychopathology.