ACTING ON DELUSIONS .1. PREVALENCE

Associations between delusions and abnormal behaviour were retrospecti vely assessed in a sample of 83 consecutively admitted deluded subject s. All were interviewed about events in the previous month using a new measure of delusional phenomenology and action. For 59 subjects this information was supplemented by informant interviews. Clinical consens us was reached concerning the probability that actions reported by inf ormants were linked to delusions. Half of the sample reported that the y had acted at least once in accordance with their delusions. Violent behaviour in response to delusions was uncommon. Information provided by informants suggested that some aspect of the actions of half of the sample was either probably or definitely congruent with the content o f their delusions. However, there was no link between self-reports and informants' reports of such action. A latent class analysis of self-r eported delusional action suggested three classes of action, namely ag gressive to self or other, defensive action, and either none or single action. Self-reported action was associated with delusions of catastr ophe. Informant data suggested that persecutory delusions were the mos t likely to be acted upon, but in contrast delusions of guilt or catas trophe appeared to decrease the chance of delusional behaviour. Action s associated with abnormal beliefs are more common than has been sugge sted.