Before starting antipsychotics in dementia, the questions you need answered center on three critical areas: whether the behavior reflects an underlying medical problem that’s being missed, whether non-drug strategies have actually been tried and exhausted, and whether the specific cardiovascular and neurological risks make sense for this person’s condition and life expectancy. Antipsychotics carry genuine dangers for people with dementia—increased risk of stroke, heart attack, pneumonia, and earlier death—so the threshold for prescribing them must be high and the reasoning documented.
Consider a common scenario: an 84-year-old woman with Alzheimer’s disease suddenly becomes aggressive and refuses meals. Her family requests medication to manage her behavior. Before the doctor prescribes an antipsychotic, these questions must be asked and answered: Is she in pain from untreated arthritis or dental disease? Does she have a urinary tract infection, which can cause dramatic behavioral changes in dementia? Is she having a reaction to a recently started blood pressure medication? Is her agitation a response to being moved to a new environment or having a new caregiver? Each of these underlying problems can masquerade as behavioral problems that seem to demand psychiatric medication—but treating the actual cause often resolves the behavior without drugs.
Table of Contents
- Has a Medical Workup Identified the Real Cause of the Behavior?
- Have Evidence-Based Non-Pharmacological Strategies Been Tried and Found Insufficient?
- What Are the Specific Cardiovascular and Mortality Risks for This Individual?
- Does the Person’s Remaining Life Expectancy Align with the Medication’s Risk and Benefit Timeline?
- What Other Medications Might Interact Dangerously, and What Are the Fall and Sedation Risks?
- Is There a Specific Plan for Monitoring, Re-evaluation, and Stopping the Medication if Indicated?
- Have the Person (If Capable) and Their Family Been Fully Informed of the Risks?
- Frequently Asked Questions
Has a Medical Workup Identified the Real Cause of the Behavior?
behavioral symptoms in dementia are almost always a sign that something is wrong, but antipsychotics treat the symptom, not the cause. Delirium—acute confusion and agitation—is one of the most common triggers for inappropriate antipsychotic prescriptions in older adults with dementia, yet delirium is usually reversible if the underlying cause is found. Common culprits include untreated infections (particularly urinary tract infections, pneumonia, and skin infections), uncontrolled pain, medication side effects, hypoxia, hypoglycemia, constipation, or acute cardiac events. A 79-year-old man with mild-to-moderate Alzheimer’s presented to the ER with sudden onset hallucinations, paranoia, and refusal to eat.
His daughter reported that he had been fine until three days ago. Before any psychiatric medication was considered, his doctors found that he had an infected catheter, a high fever, and sepsis—a medical emergency unrelated to his dementia. Once he was treated with antibiotics, his psychiatric symptoms resolved completely. If he had been given an antipsychotic, the medication would have masked the infection while making him more sedated and less likely to recover. The behavioral assessment must be thorough: When exactly did this behavior start? Did something change—a new medication, a UTI, a death in the family, relocation, loss of a familiar caregiver? What does the behavior look like specifically—is it agitation, refusal to cooperate, hallucinations, paranoia, or violence? Which times of day is it worse? Has the person been examined for pain, fever, infection, and metabolic problems? If these questions haven’t been answered, an antipsychotic prescription is premature.
Have Evidence-Based Non-Pharmacological Strategies Been Tried and Found Insufficient?
The strongest evidence for managing behavioral symptoms in dementia comes from environmental and behavioral modifications, not medication. These include consistent routines, meaningful activities tailored to the person’s interests and abilities, adequate sleep and exercise, management of pain, validation of feelings even when the content isn’t accurate, and reducing overstimulation and confusion in the environment. Staff or family training in these approaches can often prevent escalation of behavioral crises entirely. Yet the reality in many care settings is that these non-drug strategies are not tried first, or they’re tried inadequately. A nursing home might have insufficient staffing to provide one-on-one engagement with a resident who is becoming agitated, so a sedating medication becomes the path of least resistance.
A family caregiver, exhausted and unsupported, might feel that medication is their only option without ever consulting an occupational therapist or dementia specialist about environmental or activity-based solutions. A physician who has only 15 minutes with a patient in clinic might write a prescription rather than spend time understanding the triggers of the behavior. The evidence is clear: rushed antipsychotic prescriptions often happen because the alternative—time, attention, structure, and creativity—is in short supply. The warning here is significant: using antipsychotics as a substitute for adequate dementia care and staffing is a form of over-medication that exposes the person to real harm. If non-drug strategies haven’t been implemented consistently for at least several weeks before escalating to medication, and if they haven’t involved a dementia specialist or behavioral health professional, the prescription may be unjustified.
What Are the Specific Cardiovascular and Mortality Risks for This Individual?
Antipsychotics carry a black-box warning from the FDA for increased mortality in older adults with dementia. The risk includes sudden cardiac death, stroke, and other cardiovascular events. In clinical trials, older people with dementia taking antipsychotics had mortality rates roughly 1.6 to 1.8 times higher than those not taking them. This risk is not theoretical—it translates into real deaths, often occurring within months of starting the medication.
The risk is not equal for all antipsychotics. Risperidone and aripiprazole carry the highest documented mortality signals in dementia, while quetiapine and citalopram have less robust evidence but still carry concern. Before prescribing, a doctor should assess the specific person’s cardiovascular status: Has there been a recent heart attack? Does the person have a history of stroke, atrial fibrillation, or heart failure? Is blood pressure already difficult to control? Does the person have diabetes or other metabolic risk factors? Is the person taking other medications that increase the risk of sudden cardiac events or interact with antipsychotics? For a frail 91-year-old with a history of atrial fibrillation and previous stroke, the cardiovascular risk of an antipsychotic might be unacceptably high compared to the benefit of reducing agitation. For a 65-year-old with robust cardiac function and early-stage dementia, the risk calculation might be different—though still not trivial. The point is that the risk must be calculated for the specific person, not assumed to be the same for everyone.
Does the Person’s Remaining Life Expectancy Align with the Medication’s Risk and Benefit Timeline?
An important but often overlooked question is whether the person’s remaining lifespan and quality of life are compatible with the goals of antipsychotic therapy. Antipsychotics are often prescribed with the expectation that they will provide long-term benefit, but this assumption must be questioned in dementia. Someone with late-stage dementia—minimal verbal communication, minimal ability to eat or drink without assistance, bedbound, near the end of life—may receive more harm than benefit from starting an antipsychotic medication. The medication might reduce agitation that stems from the distress of dying, but the drugs themselves can accelerate decline, increase the risk of aspiration and pneumonia, and make comfort care more difficult.
The person may not live long enough to experience any behavioral benefit, and the time they do have remaining may be made worse by sedation, falls, or medical complications. In contrast, a person with early-stage dementia and a projected lifespan of many years might have a different risk-benefit profile, though the mortality risk must still be weighed seriously. This question forces a conversation about prognosis, priorities, and what “improvement” actually means for this person. If the goal is to reduce distressing behaviors in someone with months to live, there are often better approaches—comfort medications like low-dose opioids for pain, lorazepam for acute anxiety during specific events, and intensive non-drug strategies focused on dignity and comfort. If the goal is to manage persistent behavioral symptoms in someone with years ahead, the calculation is more complex and warrants specialist input.
What Other Medications Might Interact Dangerously, and What Are the Fall and Sedation Risks?
Most people with dementia are taking multiple medications for other conditions—blood pressure drugs, heart medications, diabetes medications, pain medications, sleep aids. Antipsychotics interact with many of these, sometimes in dangerous ways. Some combinations increase the risk of severe sedation, confusion, and cognitive decline. Others can cause or worsen cardiac arrhythmias. Some increase the risk of sudden drops in blood pressure, dizziness, and falls. Falls are a particular concern in older adults with dementia taking antipsychotics.
The medications impair balance and coordination, and dementia already impairs judgment and spatial awareness. A fall in an older person can mean a hip fracture, which often leads to hospitalization, surgery, mobility loss, and cascading decline. For a person on anticoagulation therapy (for atrial fibrillation or other reasons), a fall can mean intracranial bleeding. The question that must be asked is: What is the comprehensive medication review showing? Has the prescriber or pharmacist reviewed all the medications together to identify dangerous interactions? A 78-year-old man on a beta-blocker for heart failure, a diuretic for fluid overload, and several other medications was prescribed risperidone for agitation. Within days, he became severely sedated, fell in the bathroom, and hit his head. The combination of medications had caused profound sedation and orthostatic hypotension—drops in blood pressure when standing—that made him unsteady. He was hospitalized for a head injury that might have been prevented by a more careful medication review before the antipsychotic was added.
Is There a Specific Plan for Monitoring, Re-evaluation, and Stopping the Medication if Indicated?
Too often, antipsychotics are prescribed and then continued indefinitely without clear reassessment. The question should always be: How will we know if this medication is working? When will we reassess? Under what circumstances would we stop it? Without these answers in writing, the medication becomes a default therapy rather than a treatment with clear goals and endpoints.
A reasonable monitoring plan includes baseline documentation of the specific behaviors being targeted, a trial period (typically 4 to 6 weeks) with clear goals and measurable outcomes, reassessment at the end of the trial, and a decision about continuing, changing the dose, or stopping. If the medication has not reduced the targeted behavior by at least 25%, continuing it exposes the person to ongoing risk for no clear benefit. The plan should also specify how often the person will be seen (ideally by the prescriber and not just by nursing staff), what vital signs and other health parameters will be monitored, and when a follow-up assessment with a dementia specialist or geriatrician is appropriate.
Have the Person (If Capable) and Their Family Been Fully Informed of the Risks?
Informed consent for antipsychotics in dementia requires that the person (if they have capacity to understand) and their family members be told about the mortality risk, the cardiovascular risks, the risk of stroke, the risk of falls, and the goal of the medication. Many families report that they were never told about these risks—they were simply informed that their loved one would be given medication for behavioral problems. This is not adequate informed consent.
The conversation should include specific discussion of alternatives, the specific target symptoms, the timeline for reassessment, and what “success” looks like for this person. A family member should be able to ask: “So there’s about a 1.6 times higher risk of death compared to not taking this drug?” and hear an honest answer. They should be told how long the person typically lives after starting the medication in similar situations, and what the most common serious side effects are. Only with this information can a family make a genuine choice about whether to proceed with treatment or pursue alternatives.
Frequently Asked Questions
What should I do if I think my father’s agitation is actually a sign of a medical problem and not a psychiatric issue?
Request a comprehensive medical evaluation, including urinalysis and urine culture (to rule out UTI), blood tests, physical examination for signs of pain or infection, review of all medications, vital signs including temperature, and evaluation for delirium. Write down the exact timeline of when the behavior started and what changed in his environment or health around that time. Bring this information to the appointment.
Can antipsychotics ever be the right choice for someone with dementia?
Yes, but only after other options have been exhausted and the person’s specific situation justifies the risks. This typically means late-stage dementia with severe behavioral symptoms that pose safety risks, after non-drug strategies have been genuinely attempted, and when a geriatrician or dementia specialist has reviewed the case and agreed that the benefits likely outweigh the risks.
How long should we trial a non-medication approach before considering antipsychotics?
At least 4 to 6 weeks of consistent, well-structured intervention focused on environmental modification, activity engagement, and addressing any underlying medical or social triggers. This requires adequate staffing, training, and commitment to implementation—which is why “trying” for a few days before giving up and reaching for medication is not a valid trial.
What should I ask the doctor about monitoring if my family member starts an antipsychotic?
Ask how often the person will be seen, what specific behaviors the medication is meant to target, how the doctor will measure whether it’s working, when the person will be reassessed to decide whether to continue, and what the plan is if the medication isn’t helping or is causing problems.
If my mother is already on an antipsychotic for dementia, should I stop it immediately?
No—stopping abruptly can cause rebound agitation or other withdrawal effects. Instead, ask for a medication review with the prescriber, bring up your concerns about the risks, and discuss whether the medication is still necessary and working as intended. If there’s disagreement, request a second opinion from a geriatrician or dementia specialist.
Are all antipsychotics equally risky for someone with dementia?
No. Risperidone and aripiprazole have the strongest evidence for mortality risk in dementia. Quetiapine and citalopram have less robust evidence but still carry concern. However, none of them are risk-free, and all of them should be used only when absolutely necessary and with careful monitoring.
