While antipsychotics are used extensively in older patients to treat a variety of symptoms, including psychotic symptoms as features of behavioural and psychological symptoms in dementia, much of their use in treating late-life psychosis is inferred from studies of treatment for psychosis in younger patients. This is particularly problematic since older patients are more likely to suffer from major physical illness as a co-morbid feature, and a number of studies have drawn into focus concerns about the over-use of antipsychotic medication, particularly in nursing homes⁶, and excess mortality associated with antipsychotic prescribing in the elderly⁷.

While the majority of people with late-life psychotic symptoms will have hallucinations and delusions from delirium secondary to a physical cause, as a behavioural and psychological symptom of dementia, or indeed as part of an affective illness, those with schizophrenia or other late-life psychosis comprise 0.1-0.5% of the population over the age of 65⁸. Of all people over the age of 65 with schizophrenia, the majority will have developed their illness prior to age 45⁹. While there are benefits from extrapolating data derived from adults under 65, in the absence of large multicentre double-blind placebo- controlled studies into treatment of late-life psychosis, there are a number of problems associated with this approach. Primarily the difficulties fall into areas including metabolic side effects, increased risk of extrapyramidal side effects and movement disorder, and the problems of polypharmacy, co-morbid major physical illness and pharmacokinetic changes associated with ageing.

Typical antipsychotics have been used extensively in treating older people with psychosis despite the fact that their high affinity for dopamine D2 receptors is associated with increasing extrapyramidal side effects in older people¹⁰. In addition these medications tend to be sedating and associated with quite marked anticholinergic side effects, which can cause additional confusion and disorientation in patients with mild cognitive impairment or dementia. By contrast, atypical antipsychotics are associated with less in the way of significant side effects in older people¹¹ and less hospitalization by comparison to typicals¹². Both olanzapine and quetiapine are associated with a lower incidence of extrapyramidal side effects than typical antipsychotics¹³, though sedation can be problematic and there are additional issues in relation to olanzapine and increased risk of diabetes¹⁴. Risperidone has been used extensively principally because of the low incidence of sedation or anticholinergic side effects. Clozapine has been used extensively in younger patients, though hypersalivation, neutropaenia and increased risk of agran- ulocytosis, particularly in older patients, limit its use¹⁵. Much of the evidence for efficacy of antipsychotics in late-onset psychosis consists of case reports and open studies though these are at times conflicting and there is not clear evidence to suggest that atypical antipsychotics are any more effective in this group of patients. However, compliance is better¹⁶.

In the absence of robust double-blind placebo-controlled data for meta-analysis, guidelines for management of late-onset psychosis focus primarily on the expert consensus approach. Focusing on the management of schizophrenia, the use of atypical antipsychotics is clearly favoured, with 93% of experts rating risperidone as first-line treatment in this group of patients, perhaps because of its low incidence of anticholinergic or sedative side effects. While quetiapine and olanzapine were also rated as first-line treatments by 67% of experts and aripiprazole by 60%, the doses recommended in this group of patients were higher than the treatment doses used in psychosis in dementia¹⁷.

A recent study of attitudes to antipsychotic prescribing in Australian specialists working within old age psychiatry identified a preference for the use of atypical antipsychotics as opposed to typical antipsychotics in an elderly population although no clear preference for any individual atypical antipsychotic was apparent¹⁸. The principal reasons for this centred on the impression that not only were atypical antipsychotics less likely to be associated with adverse events, particularly in an elderly population who were more likely to be suffering from co-morbid physical conditions, but they were also perceived as being clinically more effective than typical antipsy- chotics. This was supported by clinical global impression of change data within the clinicians’ own patient groups, where 33% of patients with schizophrenia were rated as having improved from moderately ill to mildly ill while on atypical antipsychotics. This is similar to rates of reduction in PANSS¹⁹ total score for elderly patients with psychotic illness treated with amisulpride and risperidone²⁰ though a more recent open label study of amisulpride treatment for very-late- onset schizophrenia-like psychosis showed significant improvement over a variety of measures and 46.6% reduction in PANSS total score²¹.

Taking into account the likelihood of co-morbid physical or psychiatric symptoms in elderly patients and the altered pharmacodynamics and pharmacokinetics of older people, the limited evidence available suggests that treatment with atypical antipsychotic medication may be useful. Although there are recent concerns in relation to increased cerebrovascular risk in older patients with dementia treated with atypical antipsychotic medication²² and also concerns in relation to excess mortality associated with these drugs⁷, their cautious utilization continues to be appropriate in the absence of any clear alternative treatments.

The use of depot medication may not lead to additional advantages. Reeves et al.¹ found that treatment response was not increased by the use of depot and concluded that regular contact with a community psychiatric nurse, an intervention likely to increase social exposure, may be as important in maintaining treatment response.

The most common types of delusions are those of reference, control or hypochondriasis²³. Partition delusions²⁴ are also over-represented in the late-life onset group. A detailed description of cognitive models and cognitive behavioural interventions lies outwith the scope of this chapter, but an understanding of the development of delusions comes from cognitive behavioural work with younger adults. Cognitive models have been used to explain delusion formation and maintenance²⁵. Similar errors in attribution have been proposed as mechanisms leading to hallucinations²⁶, although some exploratory work to examine the cognitive aetiology of delusions in people with late-life psychosis^27,28 suggests that the wider range of cognitive distortions described in younger adults with schizophrenia may not be as pertinent in older people.

Bentall et al.²⁹ describe the development of persecutory thinking as a way of the person protecting themselves from perceived assaults on their self-esteem by allowing them to attribute negative events to external causes. Since many patients with late-life onset psychoses have a history of such attribution during most of their adult life, it is easy to understand the continuum of a person feeling that others are in some way hostile towards them, through the development of more eccentric beliefs with associated social withdrawal and finally explicit persecutory delusions in an intensely isolated individual. Protective behaviours become increasingly explicit and the picture of a chronically isolated, elderly lady sitting in a cold, dark house, unheated because she believes the neighbours are stealing her electricity through some tenuously understandable process, and covering surfaces of her furniture or cooking utensils with silver foil to prevent the effects of toxic spray used by the same neighbours will not be unfamiliar to clinicians treating this group of people.