GM 2013, April  

Dr Jaya Joy House Officer in Geriatric Medicine Department of Geriatric Medicine, Countess of Chester Hospital Chester, UK

Dr Joe John VattakatucherConsultant Psychiatrist, Department of Psychiatry, Hollins Park Hospital, Warrington 




Elderly patients are less likely to be aggressive or violent than younger people and they are more often victims than perpetrators of violence.2 However, dementia is associated with aggression and violence. Incidents of aggression and violence were reported in up to 20% of Alzheimer’s disease patients.3  

Burns et al reported that over half the patients with a diagnosis of dementia were verbally or physically aggressive.4 Aggression and violence are the most serious behavioural disturbances associated with dementia and cause great distress to carers.5 Such behaviour is thought to be more frequent in hospitals than in the community.4 This may be explained by the fact that patients with such behaviour are concentrated in psychogeriatric wards as aggression and violence are the most common causes for admission to hospital or nursing homes.4,6

Several theories have been proposed to explain aetiology of aggression and violence in dementia. They could be grouped into three main categories: biological, psychological and environmental.5 Biological theories tend to be either damage to particular areas of the brain or malfunction of specific neurotransmitter systems. Damage to amygdale, hippocampi, temporal lobes and frontal lobes have been implicated.7 Changes to 5-HT, GABA, cholinergic and catacolamine levels in the brain have been suggested.5 There is a possibility that some of the aggressive behaviour in dementia may be due to ictal activity.

This could explain the effectiveness of carbamazepine in a select group of patients with aggression and dementia. There are several reports of aggressive behaviour secondary to acute confusional states superimposed on dementia.5 Psychological factors such as depressive symptoms, psychotic symptoms, exacerbation of premorbid personality traits and frustration from impaired ability to communicate have been suggested.2,5,8 Environmental factors such as overcrowding, inadequate lighting, poor nursing techniques and inappropriate ward design have been suggested.1,5,9

The aim of this article is to systematically review literature about aggression and violence in patients with dementia in psychogeriatric wards and hospitals.



A comprehensive search was carried out using the following key words: aggression, violence, behavioural disturbance, dementia, psychiatric units, mental health units. Medline, EMBASE, PsychInfo, CINHAHL and Cochrane Library were searched using the above-mentioned key words. Abstracts of the articles were reviewed and relevant articles were identified and requested. All articles requested were analysed. Reference lists were examined and relevant articles requested. All articles that reported on aggression and violence in dementia patients in the hospital setting were included. Studies in community samples were excluded. Seven studies met the criteria and the important findings are reported below.



Nilsson et al reported a prospective study of six weeks duration on aggressive behaviour in hospitalised psychogeriatric patients in Stockholm, Sweden.10 The patients selected were inpatients (n=40) from two psychogeriatric wards. All patients were observed using a special aggression report. 33 (82.5%) patients had a diagnosis of dementia with marked cognitive impairment, four had chronic schizophrenia with cognitive impairment and two suffered from unspecified chronic psychosis. 266 aggressive incidents were reported from 27 patients. 88% were directed towards staff and 75% were provoked when the staff helped the patients with activities of daily living. To our knowledge, this was the first study to report on this topic. However, the study did not specifically investigate aggression and violence in dementia patients. From the results provided, it is not possible to ascertain what proportion of incidents involved patients with dementia.

Haller et al studied violent behaviour in geriatric inpatients with dementia. They used a retrospective study design and examined medical records of 52 dementia patients admitted from July 1 1979 to September 1 1987 in an acute ward of a university hospital.11 The incidents were grouped into "fear inducing behaviour" and "physical attacks". 30% of females and 70% of males exhibited fear-inducing behaviour, 51.9% of females and 48.1% of males physically attacked others. Married patients and those with families were over represented in the group of violent patients. No significant association were found between violence and gender, social class or ethnic group. The study tabulated only the most serious description if more than one incident of violence had occurred. The results do not give all the incidents of violence, only the most serious ones. This is an important drawback of the study.

Meyer et al compared geriatric patients (n=78) admitted during the three-year period from January 1 1987 to December 31 1989 to non-geriatric sample (n=1904) admitted to Hasting Centre, Nebraska. 60% of geriatric subjects had organic mental disorder.12 The information was collected from "physical intervention reporting form" that was used to record incidents that required seclusion, restraint, mechanical support or behavioural control. The most common antecedent for both geriatric and non-geriatric staff was being asked to do something. Geriatric patients showed more aggression to other patients but significantly less towards staff, self or objects compared to non-geriatric patients. Patients who had diagnosis of organic mental disorder or schizophrenia exhibited the majority of aggressive behaviour. The percentage of aggressive episodes by organic mental disorders (27.3%) was comparable to schizophrenia (30.7%). This is a large study and a substantial number of patients were studied for a considerable period of time. However, only the incidents that led to physical intervention were taken into account and less severe incidents were not considered.

Shah studied patients admitted to a 12-bedded acute psychogeriatric assessment ward between January 1989 to December 1990.13 The incidents of violence were categorised into four groups: assault against staff, patients, self and property. The study used a retrospective design and the nature of the incidents was ascertained from incident forms completed by the nursing staff. 39 incidents were recorded during this period. There were 18 incidents of violence towards staff, 18 incidents of violence towards other patients, one incident of self-harm and one incident of violence against property. The study reported that violent patients were more likely to be demented (p<0.005), not to be depressed (p<0.05) and under the age of 75 (p<0.001). 91% of assaults did not result in any detectable injury and the remainder resulted in minor physical injuries such as bruising, abrasions and small lacerations. Incident forms do not capture all the events especially when used in a retrospective study and significant underreporting of violence in incident forms is well established.14  

Patel and Hope studied 90 patients on one assessment and five long stay psychogeriatric wards in Oxford, UK.15 The data was collected using Rating Scale for Aggressive Behaviour in the Elderly (RAGE) over a three day period. 71% of patients had a diagnosis of dementia and 13% had a diagnosis of schizophrenia. The dementia group had a higher level of aggressive behaviour (p<0.05) compared to patients with other diagnoses. Unlike instruments used in other similar studies, RAGE has been demonstrated to possess high inter-rater reliability, test-retest reliability, internal consistency and clinical validity.

Shah studied patients with dementia admitted to a 12-bedded acute psychogeriatric ward from January 1989 to December 1991.16 A retrospective design was used and details of events were ascertained using a standard incident form similar to their previous study. 29 violent incidents were recorded; 14 were against staff and five against other patients. Patients with ICD-9 diagnosis of senile dementia were over represented in the violent group (p<0.005) compared to other forms of dementia. No differences were detected with regard to age, sex, legal status, previous psychiatric history, source of admission or place of discharge. Although the author did not specifically comment on the site of the study or sample of patients, it appears that the study was conducted in the same site as the previous study and there may be a considerable overlap of patients and incidents in both studies.13 This raises the question of if this study could be considered as a separate study or a part of the previous study.

Freyne and Wrigley conducted a prospective study on patients who were admitted to a 40-bedded dementia ward in north Dublin.17 They introduced a form to record aggression and violence; only incidents against staff were recorded. The study was conducted over a period of three months. The authors reported a total of 47 incidents over a three-month period. The study found that 89% of aggressive incidents took place during some form of nursing intervention (66% when dressing/undressing, 15% when washing). 80% of incidents were rated as "not so serious", 20% as "serious" and no incidents were rated as "very serious". The study also used "staff impact questionnaires" which were completed by 11 staff nurses, two nursing assistants, and four domestic staff. One nurse rated a reaction of "very upset", six nurses reported "quite upset", five responded "upset a little", and was "not upset". Only two of 17 respondents felt that their training to deal with such behaviour was adequate.

The study used a prospective design and thus reduced the scope for recall bias. The study also investigated the impact on the staff, which had not been investigated in the other studies. However, the study did not report aggression towards other patients, property or self. This was a significant drawback of the study. The authors did not comment on the total number of admissions or bed occupancy rate in the ward over the period of the study, which made it difficult to get a clear picture of the prevalence of aggression/violence in the hospital setting.



Dementia often starts as a problem with cognition but almost always evolves into a problem with behaviour. However, no comprehensive guidelines exist on the assessment of risk to others posed by patients with dementia.2 What constitutes aggression or violence amongst dementia patients is not straightforward. Some nurses would rate behaviour as "aggressive" only if the patient intended to harm.5 The concept of "aggression" and "violence" can be problematic if it depends on "intention to harm" as this may be difficult to assess in this group of patients.

Significant variations in aggression and violence have been reported between countries, between those living in their homes and those living in nursing homes.18 Swearer et al reported a significant relationship (p<0.01) between disease severity and violence.19 No significant differences were found between patients with Alzheimer’s dementia, multi infarct dementia and mixed dementia. However the relationship between disease severity and violence may not apply to fronto-temporal dementia as research suggests that fronto-temporal dementia patients are more likely to be aggressive or violent early in the course of disease and they become apathetic and inactive as the disease progresses.7








The prevalence of aggression and violence among inpatients with dementia was comparable to other diagnoses such as schizophrenia. The majority of the incidents were directed at staff and occurred during some form of nursing intervention. Staff working with this group of patients felt that their training was inadequate to deal with aggression and violence. More research is needed in this area and such behaviour should be taken into consideration when training staff and planning services for this group of patients.


Conflict of interest: none declared




1. Owen C, Tarantello C, Jones M, Tennant C. Violence and aggression in psychiatric units. Psychiatric Services 1989; 49(11); 1452–57

2. Hindley N, Gordon H. The elderly, dementia, aggression and risk assessment. International Journal of Geriatric Psychiatry 2000; 15: 254–59

3. Swearer JM, David A, Drachman, et al. Troublesome and disruptive behaviours in dementia. Journal of American Geriatrics Society 1988; 36(9): 784–90

4. Burns A, Jacoby R, Levy R. Psychiatric phenomena in Alzheimer’s disease IV: Disorders of behaviour. British Journal of Psychiatry 1990; 150: 86–94

5. Patel V, Hope, T. Aggressive behaviour in dementia: A review. International Journal of Geriatric Psychiatry 1993; 8: 457–72

6. Shah A. Violence and psychogeriatric inpatients. International Journal of Geriatric Psychiatry 1992; 7; 39–44

7. Miller BK, Darby A, Benson DF, et al. Aggressive, socially disruptive and antisocial behaviour associated with fronto-temporal dementia. British Journal of Psychiatry 1997; 170; 150–55

8. Constantine G, Lyketsos, Steele S, Galik E, et al. Physical aggression in dementia patients and its relationship to depression. American Journal of Psychiatry 1999; 156(1), 66–71

9. Ng B, Kumar S, Ranclaud M, et al. Ward crowding and incidents of violence in acute psychiatric unit. Psychiatric Services 1990; 52(4), 521–25

10. Nilsson et al. Aggressive behaviour in hospitalised psychogeratric patients. Acta Psychiatrica Scandinavia 1988; 78; 172–75

11. Heller E, Binder MD, McNiel DE. Violence in geriatric patients with dementia. Bulletin of academic Psychiatry and Law 1989; 17(2): 183–88

12. Meyer J, Schalock R, Genaidy H. Aggression in psychiatric hospitalised geriatric patients. International Journal of Geriatric Psychiatry 1991; 6; 589–92

13. Shah A. Violence and psychogeriatric inpatients. International Journal of Geriatric Psychiatry 1992; 7; 39–44

14. Lion JR, Snyder W, Merrill GL. Underreporting of assaults on staff in a state hospital. Hospital and Community Psychiatry 1981; 32: 497–98

15. Patel V, Hope RA. Aggressive behaviour in elderly psychiatric inpatients. Acta Psychiatrica Scandinavia 1992; 85; 131–35

16. Shah, A. Violence among psychogeriatric inpatients with dementia. International journal of Geriatric Psychiatry 1995; 10; 887–91.

17. Freyne and Wrigley. Aggressive incidents towards staff by elderly patients with dementia in a long stay ward. International Journal of Geriatric Psychiatry 1996; 11: 57-63

18. Skovdahl K, Kihlgren AL, Kihlgren M. Different attitudes when handling aggressive behaviour in dementia- narratives from two caregiver groups. Ageing and Mental Health 2003; 7(4): 277–86

19. Swearer JM, David A, O’Donell BF, Mitchell AL. Troublesome and disruptive behaviours in dementia. Journal of American Geriatrics Society 1988; 36(9); 784–90