There is no single best medication for agitation in dementia, but one drug now stands apart as the only FDA-approved option specifically for this use: brexpiprazole (brand name Rexulti), approved in 2023 for agitation associated with Alzheimer’s disease. Before that approval, every medication prescribed for dementia-related agitation was used off-label, and many still are. The right choice depends on how severe the agitation is, what other health conditions are present, and whether non-drug approaches have already been tried.
Why Non-Drug Approaches Come First
Every major guideline, from NICE in the UK to the American, Canadian, and European geriatric psychiatry associations, recommends starting with non-drug strategies before turning to medication. Music therapy, massage, therapeutic touch, and structured daily activities can reduce mild-to-moderate agitation in the short term. These approaches carry no risk of the side effects that come with nearly every medication used for this condition.
The limitation is that these interventions tend to lack lasting benefits over weeks and months, and they do little for severe agitation. When someone with dementia is in distress, at risk of harming themselves or others, or when behavioral strategies have been genuinely tried and aren’t enough, medication becomes a reasonable next step.
Brexpiprazole: The Only FDA-Approved Option
Brexpiprazole is an atypical antipsychotic that earned FDA approval based on two clinical trials measuring agitation using a standardized rating scale (the Cohen-Mansfield Agitation Inventory) over 12 weeks. In both trials, patients taking 2 mg or 3 mg daily showed statistically significant improvements compared to placebo. The target dose is 2 mg once daily, reached gradually: 0.5 mg for the first week, 1 mg for the second week, then 2 mg from day 15 onward. If needed, the dose can go up to 3 mg after at least two more weeks.
Common side effects include headache, dizziness, urinary tract infections, cold-like symptoms, and sleep problems (both drowsiness and insomnia). Like all antipsychotic medications used in elderly patients with dementia, brexpiprazole carries a boxed warning noting that these drugs are associated with an increased risk of death. Across 17 controlled studies of various antipsychotics in elderly dementia patients, those taking the medications were 1.6 to 1.7 times more likely to die than those on placebo. That warning applies to brexpiprazole as well, which is why it’s reserved for cases where the benefits clearly outweigh the risks.
Off-Label Antipsychotics: Risperidone and Others
Before brexpiprazole’s approval, risperidone was the most studied antipsychotic for dementia-related agitation, and it remains widely used off-label. A meta-analysis of 15 placebo-controlled trials found small but significant benefits for risperidone in reducing agitation and psychosis. In one large comparative study, patients on risperidone stayed on the medication an average of 26.7 weeks before it stopped working, compared to just 9 weeks for placebo. Olanzapine showed a similar pattern at 22.1 weeks.
Quetiapine is frequently prescribed in practice, often because it’s perceived as gentler, but the evidence for it is thin. Research reviews have been unable to draw conclusions about whether quetiapine actually helps with dementia-related agitation. Aripiprazole has somewhat better evidence, with the same meta-analysis showing small but significant effects similar to risperidone.
All of these carry the same class-wide mortality warning as brexpiprazole. Additional risks in older adults include falls, sedation, movement problems, and metabolic changes. When an antipsychotic is started, guidelines emphasize using the lowest effective dose and reassessing regularly to determine whether the medication is still necessary.
Citalopram: An Antidepressant Alternative
Citalopram, a common antidepressant, showed genuine benefit for agitation in a well-designed trial called CitAD. At 30 mg daily over nine weeks, 40% of patients on citalopram showed moderate or marked improvement, compared to 26% on placebo. Patients also improved on measures of overall behavioral symptoms and caregiver distress.
The catch is safety. Citalopram at that dose caused a meaningful change in heart rhythm, prolonging the QT interval by about 18 milliseconds on average. Some patients exceeded thresholds considered risky for dangerous heart rhythms. The medication also worsened cognition slightly, and patients experienced more falls, appetite loss, and diarrhea than those on placebo. Because of the heart rhythm concern, the FDA has warned against using citalopram above 20 mg in adults over 60, which is below the dose that showed benefit for agitation. This creates a real clinical tension: the dose that works best may not be safe for many older adults.
Despite these limitations, citalopram or similar antidepressants are sometimes tried when antipsychotics are too risky or unwanted, particularly for patients with milder agitation or co-existing depression.
Medications That Don’t Work
Valproate, a mood stabilizer sometimes tried for agitation, has been studied in multiple trials and consistently fails to outperform placebo. A Cochrane review found no meaningful effect on agitation scores or overall behavioral symptoms across several studies. Worse, patients on valproate experienced roughly double the rate of side effects, including sedation, nausea, vomiting, and diarrhea. The UK’s NICE guidelines explicitly state that valproate is no more effective than placebo for this purpose and causes more harm. It is not a recommended option.
Benzodiazepines like lorazepam are sometimes used in emergencies to calm someone quickly, but guidelines recommend against their routine use for agitation in dementia. They increase fall risk, worsen confusion, and can paradoxically make agitation worse over time. The American Psychiatric Association specifically recommends against benzodiazepines in patients with delirium or pre-existing cognitive impairment unless there is a very specific reason to use them.
What’s in the Pipeline
AXS-05, a combination of dextromethorphan and bupropion, is being studied in phase 3 trials for Alzheimer’s-related agitation. Early clinical data suggest it may reduce agitation and appears generally well tolerated. Because bupropion is an antidepressant, this combination could be particularly useful for patients who also have depression alongside their agitation. It has not yet been approved.
A synthetic cannabinoid called nabilone showed promise in a small pilot trial of 38 patients with moderate-to-severe Alzheimer’s, significantly improving agitation, overall behavioral symptoms, and caregiver distress over six weeks. Larger trials are needed before it could become a standard option.
How Medications Are Typically Chosen
In practice, the decision often comes down to severity. For mild agitation, clinicians may try non-drug strategies alone or consider a cautious trial of an antidepressant like citalopram at a lower dose, especially if depression is also a factor. For moderate-to-severe agitation that causes real distress or safety concerns, brexpiprazole is now the leading option given its FDA approval and supporting evidence. Risperidone remains a common alternative when brexpiprazole isn’t available, isn’t covered by insurance, or isn’t tolerated.
Regardless of which medication is chosen, the goal is the lowest dose that meaningfully reduces distress, reassessed every few weeks. Agitation in dementia often fluctuates with the environment, time of day, pain levels, infections, and other factors. Addressing those root causes, when they can be identified, sometimes reduces or eliminates the need for medication entirely.