Intended for healthcare professionals


Treating behavioural and psychological signs in Alzheimer's disease

BMJ 1999; 319 doi: (Published 17 July 1999) Cite this as: BMJ 1999;319:138

The evidence for current pharmacological treatments is not strong

  1. Clive Ballard, MRC clinical scientist-senior lecturer in old age psychiatry (c.g.ballard{at},
  2. John O'Brien, Senior lecturer in old age psychiatry
  1. Newcastle General Hospital, Newcastle upon Tyne NE4 6BE

    Dementia is a prominent healthcare issue for primary care physicians and specialist services. Over 90% of patients with dementia experience a “behaviour disturbance,”1 often referred to as behavioural or psychological signs in dementia in accordance with the recommendation of the International Psychogeriatric Association. These symptoms are distressing to patients and troublesome to carers and often precipitate admission to residential facilities.1 What is the evidence that any of the several drugs that are currently used to treat these symptoms are effective?

    Managing the behavioural and psychological signs of dementia is a major problem for healthcare professionals. Neuroleptic drugs are the mainstay of pharmacological treatment, although their use is justified largely on the basis of clinical anecdote, and they have many harmful side effects. These include parkinsonism, drowsiness, tardive dyskinesia, falls, accelerated cognitive decline,2 and severe neuroleptic sensitivity reactions.3 It is therefore not surprising that the chief medical officer has recommended judicious use of these agents in patients with dementia.4

    In 1990 Schneider published a landmark study showing the paucity of large, placebo controlled, double blind trials of neuroleptic agents in treating behavioural and psychological signs in dementia.5 Since then research in the subject has increased, but most treatment studies have used an open or active comparison design, a major methodological flaw given the high placebo response rates (40%).5 Two large multicentre studies with risperidone have recently been completed,6 showing a significant advantage over placebo for overall reduction of behaviour disturbances, although in one of the studies psychotic symptoms did not improve significantly. In addition, psychosis and aggression responded preferentially to different doses. The studies of neuroleptic agents are summarised in the table on the BMJ's website (

    There is a tendency in clinical trials to group together different behavioural and psychological signs, although they are likely to have separate neurochemical or neurophysiological bases. For example, there is evidence linking visual hallucinations to cholinergic depletion,7 but delusions are linked to relative preservation of the parahippocampal gyrus,8 while aggressive behaviour is related to abnormalities of serotonergic function.9 As these phenomena are disparate, they may respond to different agents or different doses of the same agents. In practice many patients with dementia have more than one type of behavioural or psychiatric symptom, but it should be possible to undertake separate studies focusing on common disturbances such as aggression or psychosis when they are clinically important and represent the most problematic behavioural or psychological symptom for that individual patient.

    Most intervention trials for behavioural or psychological signs in dementia have a duration of less than three months. In this time frame low doses of neuroleptic drugs are well tolerated, but the issues of longer term efficacy and safety are not addressed. For example, one prospective study following a research cohort of patients with dementia has suggested that neuroleptic agents might hasten cognitive decline,2 while another found no differences in the persistence of psychotic symptoms over one year in patients taking or not taking antipsychotic medication.10 Furthermore, the high placebo response rates5 and the spontaneous resolution of psychotic symptoms within three months10 among many sufferers indicates that a substantial proportion of patients may not require pharmacological treatment.

    Clinical trials, mainly with an open design, have also been undertaken with a wide range of non-neuroleptic treatments including anticonvulsants and trazadone, as well as non-pharmacological treatments such as carer training or sensory inputs.11 Again these trials have generally been small and have clustered together different behavioural and psychological signs. In the few preliminary studies which have been conducted carbamazepine and trazadone appear to be as effective as neuroleptics,11 although the main areas of improvement are agitation, aggression, and restlessness, rather than psychosis.

    Encouraging data are starting to emerge from the secondary analysis of trials involving cholinesterase inhibitors, suggesting that 50% or more of patients with psychotic symptoms experience improvement.11 Placebo response rates are, however, high, and the patients were recruited because of cognitive impairment not because of psychotic symptoms. Although this does not constitute evidence of efficacy, it indicates the need for specific intervention trials.

    Dementia is a common condition, and sufferers have a particularly high risk of adverse treatment responses, so it is important to have clear evidence that treatments are both effective and safe. There is an urgent need for double blind trials focusing on specific behavioural or psychological signs in dementia, with designs that allow for the high rates of spontaneous remission, especially studies using cholinesterase inhibitors and non-pharmacological intervention strategies.

    Given current knowledge, unless symptoms are extremely distressing it would seem appropriate to monitor the disturbances for at least one month before starting pharmacological treatments. The monitoring period allows time for spontaneous resolution, while psychosocial interventions may facilitate better practical management or evoke key elements from the very pronounced “placebo” response. Better evidence is required before individual pharmacological agents are licensed specifically for managing behavioural or psychological signs in dementia.


    • JOB has received a research grant from Pfizer while JOB and CB have jointly received a research grant from Eli Lilly. CB acted as principal investigator for a recent trial of quetiapine for Zeneca and acted as a member of an advisory committee for Janssen.

    • website extra A table showing the results of randomised trials of neuroleptics is available on the BMJ's website


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.