It remains unclear whether early-onset and late-onset schizophrenias (EOS and LOS, respectively) are distinct disorders. Both EOS and LOS are characterized by the presence of positive symptoms, par-ticularly systematized delusions. In late-onset cases delusions most often are of the persecutory or partition type, with partition delu-sions being almost unique to that group. Partition delusions refer to the belief that persons or forces (such as gas or electricity), usually of a threatening nature, are operating or entering the home through the walls, ceilings, floors or doors, attempting to interfere. Hallucinations can also be seen in both EOS and LOS. Multimodal hallucinations (i.e. auditory, visual, sensory, gustatory and tactile) are characteris-tic of the late-onset group², while early-onset cases typically have auditory hallucinations only. Schneiderian first-rank symptoms, e.g. hearing a running commentary or two or more voices conversing, can be seen in both groups³.

Important differences between early-and late-onset cases include infrequent formal thought disorder and fewer negative symptoms, such as affective flattening, in late-onset cases, as well as better premorbid functioning, including better occupational and marital histories. Mental functioning and personality remain relatively well preserved in late-onset cases. Jeste etal. compared three groups – younger persons with EOS, older persons with EOS, and LOS – and found a similar overall pattern of neuropsychological impairment, but with some slight differences favouring the late-onset group⁴. Likewise, a recent review by Rajji and Mulsant compared older patients with EOS versus LOS and found them to be equally cognitively impaired, exhibiting deficits in executive functioning, visuo-spatial ability and verbal fluency⁵. Both groups respond well to neuroleptics but later onset cases have a more favourable short-term prognosis than those with onset before 40. In spite of the many similarities, the preservation of personality and lack of deterioration in function lead some experts to challenge the inclusion of later onset cases in the traditional definition of schizophrenia.

Epidemiological studies of schizophrenia and associated disorders in the elderly face several challenges. First, terminology has changed over time. In 1912 Emil Kraepelin introduced the term ‘paraphrenia’ to describe a condition characterized by schizophrenia-like delusions and hallucinations not associated with mood disorder, but in the last edition of his text book classified such individuals as having schizophrenia because the course of their illness followed that dis-ease. Roth later revived and modified the phrase ‘late paraphrenia’ to describe a late-life-onset disorder characterized by prominent para-noid delusions but an absence of social and functional deterioration⁶.

Second, diagnostic criteria pertaining to age cut-offs have changed over time. Prior to DSM-III-R, schizophrenia could not be diagnosed in individuals who had onset of symptoms after age 44. Beginning with DSM-III-R, onset after 44 was allowed but cut-offs for what could be defined as ‘late-onset’ schizophrenia have varied. More recently, a consensus statement proposed by an international group of experts in the field defined late-onset schizophrenia as arising after 40 years of age and very-late-onset schizophrenia-like psychosis (VLOSLP)as arising after 60 years of age⁷.