Intended for healthcare professionals


Treating agitation in dementia

BMJ 2011; 343 doi: (Published 17 July 2011) Cite this as: BMJ 2011;343:d3913
  1. Paul B Rosenberg, associate professor of psychiatry and behavioral sciences1,
  2. Constantine G Lyketsos, Elizabeth Plank Althouse professor1
  1. 1Department of Psychiatry and Behavioral Sciences, Johns Hopkins Bayview, Johns Hopkins Medicine, Johns Hopkins University, Baltimore, MD, 21224, USA
  1. kostas{at}

A systematic pain management protocol may help

Neuropsychiatric symptoms are nearly universal in dementia,1 and agitation is among the most distressing for patients and caregivers. Currently no drug has clearly been shown to be of value in the treatment of agitation in dementia. The most commonly used class of drugs (antipsychotics) may be neither safe nor effective.2 3 Although antidepressants may have a better risk to benefit ratio, no sufficiently powered trials have evaluated their safety and efficacy in dementia, and only one is currently in progress.4

Patients may be unable to communicate their sensation of pain because of impaired memory or lack of expressive language. Underdiagnosed and undertreated pain has been associated with agitation in dementia.5 Thus, better pain management may decrease agitation in dementia.

To test this hypothesis, the linked randomised controlled trial by Husebo and colleagues (doi:10.1136/bmj.d4065) assessed an eight week trial of pain treatment in 253 nursing home residents with dementia and agitation.6 Treatment allocation used a cluster approach at the nursing home level because staff needed to be trained in assessing pain and because of concerns about spillover effects. The authors compared the stepwise protocol for treatment of pain (SPTP) with usual care. Seventy per cent of participants received high dose paracetamol (acetaminophen) 3 g daily, about 20% received a buprenorphine patch because of swallowing difficulties, and fewer than 10% received morphine or pregabalin. The cohort was typical of people with advanced dementia, with a mean age in their late 80s and mean mini-mental state examination scores between 7 and 8. Outcomes were measured with two well established agitation measures: the Cohen-Mansfield agitation inventory (CMAI) and the neuropsychiatric inventory-nursing home version (NPI-NH), as well as with an observational rating of pain severity (mobilisation-observation-behaviour-intensity-dementia-2 pain scale; MOBID-2).

The intervention produced a clinically and statistically significant reduction in agitation and pain. CMAI decreased 7 points more with the intervention than with placebo (from a baseline of 56). NPH-NH decreased 9 points more from a baseline of 31-35, and MOBID-2 decreased 1.3 more points from a baseline of 3.7-3.8. These effects compare favourably with the equivocal results of treatment for agitation using antipsychotics,2 acetylcholinesterase inhibitors,7 and memantine.8

The trial shows that treating underdiagnosed comorbidity in dementia can greatly reduce agitation, and as such should give clinicians a new strategy for tackling agitation in subgroups of patients. The intervention is straightforward to implement in practice because of the simplicity of the protocol, which follows recommendations developed by the American Geriatrics Society.9 Agitation is both most prevalent and most problematic in patients with dementia and pain, and this type of intervention could greatly improve their care.

The results seem to be broadly generalisable to typical nursing home residents with advanced dementia, and they reinforce the principle that optimal dementia care starts with optimal medical care,10 including management of medical comorbidities. Greater medical comorbidity is associated with problem behaviours in dementia, including agitation,11 so it may be possible to reduce agitation in dementia by treating other medical comorbidities, such as occult infection and constipation. Most importantly, the results suggest that the management of agitation in dementia is a part of complete “dementia care” and that agitation should not be thought of as an isolated symptom. These observations hold true for other important neuropsychiatric comorbidities of dementia, including depression and apathy.

However, there are caveats to the results. Firstly, although four different treatments were available, most patients received a substantial regular dose of paracetamol. The beneficial results could largely be due to paracetamol, which is a particularly safe and effective analgesic in frail elderly patients; the specific safety and efficacy results of the other study drugs are not reported. Secondly, although the study provides no details of the pain assessment, the interventions were all drug based. Increasing evidence points to the efficacy of systematic non-drug based interventions for behavioural disturbance in dementia.12 Hence, it is not clear to what extent patients received a systematic evaluation of pain sources with tailored non-drug based interventions. It would be interesting to know how effective the combination of interventions (targeted pain evaluation, non-drug based interventions, and SPTP) would be, and whether treatment of agitation should start with an empirical trial of paracetamol.

Current and future research will continue to search for effective disease modifying treatments for Alzheimer’s disease, targeted using biomarkers, but given the challenges it will probably take decades to bear fruit. In the meantime, Husebo and colleagues’ study highlights the opportunities for providing effective dementia care by identifying and treating pain with low technology safe interventions that are currently available.


Cite this as: BMJ 2011;343:d3913


  • Research, doi:10.1136/bmj.d4065
  • Competing interests: All authors have completed the ICMJE uniform disclosure form at (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; CGL has received grant support from Forest, Glaxo-Smith-Kline, Eisai, Pfizer, Astra-Zeneca, Lilly, Ortho-McNeil, Bristol-Myers, Novartis, National Football League, and Elan; has been a consultant/adviser for Astra-Zeneca, Glaxo-Smith Kline, Eisai, Novartis, Forest, Supernus, Adlyfe, Takeda, Wyeth, Lundbeck, Merz, Lilly, Pfizer, Genentech, Elan, NFL Players Association, and NFL Benefits Office; and received travel support or honorariums from Pfizer, Forest, Glaxo-Smith Kline. PBR has received research support from Merck, Lilly, Elan, Janssen, and Pfizer.

  • Provenance and peer review: Commissioned; not externally peer reviewed.


View Abstract