Antipsychotics for Sundowning in Dementia — Risks, Alternatives, and When They're Warranted

The Conversation Nobody Wants to Have

Your parent's sundowning has escalated past what routines and lighting changes can manage. The geriatric psychiatrist mentions an antipsychotic. You've read enough to know these drugs carry serious risks. But the nightly screaming, the aggression during evening care, the 3 AM hallucinations that leave you both shaking — those carry risks too.

This isn't a simple decision, and anyone who tells you otherwise is selling something. Here's what the evidence actually says.

The FDA Black-Box Warning

All atypical antipsychotics carry a black-box warning — the FDA's most serious label — for increased risk of death in elderly patients with dementia-related psychosis. The risk is real: the research summarized for this guide reports approximately a 1.6- to 1.7-fold increase in mortality risk compared with placebo, as well as elevated stroke risk; cited harms include cardiovascular events and aspiration pneumonia.

The absolute risk depends on the medication, study population, and the person's clinical situation. Ask the prescriber to explain the absolute and relative risks for the proposed drug rather than relying on a single trial percentage.

The warning also doesn't tell you what happens if severe sundowning remains unmanaged. Aggression and wandering can create serious safety risks — falls, poor intake or dehydration, caregiver injuries, and crisis hospitalization.

The Medication Landscape

Sundowning medications exist on a spectrum from lower-risk sleep aids to higher-risk antipsychotics. The research-supported approach is stepwise — begin with non-pharmacological strategies and escalate only if needed:

Melatonin. May be considered as a low-dose sleep aid when sleep-wake disruption is part of the problem. Discuss whether it is appropriate with the medical team after non-pharmacological strategies have been exhausted; do not choose a dose or formulation from this article.

Trazodone. A low-dose antidepressant that a geriatric psychiatrist may consider when non-pharmacological measures are insufficient. It is not risk-free; ask the clinician or pharmacist about sleepiness, dizziness, disorientation, and fall risk before any trial.

Other antidepressants. The choice of an antidepressant or sleep aid depends on the target symptom, other medicines, and the person's medical condition. Discuss options with the prescribing clinician rather than treating a class as a general-purpose sundowning solution.

Atypical antipsychotics (including risperidone; brexpiprazole is another example discussed in the research). These may be considered for severe evening agitation, physical aggression, hallucinations, or paranoid delusions when safety is compromised and non-pharmacological strategies are insufficient. They require an individualized risk-benefit discussion because of the FDA warning above.

Review sedating and anticholinergic medications. The research identifies high-risk anticholinergic or sedative medications as possible contributors to acute delirium and behavioral worsening. Ask a clinician or pharmacist to review them rather than treating any medication change as a self-directed decision.

When the Risk-Benefit Calculus Shifts

The clinical threshold for considering an antipsychotic is specific, not subjective:

• Physical aggression during care tasks that has caused injury (to the patient or caregiver)
• Persistent hallucinations or paranoid delusions that cause visible distress and don't respond to environmental interventions or redirection
• Nighttime agitation severe enough that the person with dementia is at genuine risk of wandering outside, falling, or harming themselves
• Caregiver collapse — when sleep deprivation and stress have degraded the caregiver's ability to provide safe care, and no respite alternative is available

If these safety concerns are absent, the research supports continuing non-pharmacological approaches and discussing any medication trial with the medical team.

Questions to Ask Before Starting an Antipsychotic

If you and the prescribing physician decide a trial is warranted:

1. "What's the lowest effective starting dose, and when will we reassess?" The prescriber should select the dose. Set a specific follow-up date; the research script uses a six-week re-evaluation to discuss whether tapering or de-prescribing is safe.

2. "What's the specific target symptom?" "Agitation" is too vague. "Physical aggression during bathing that caused a bruise on Tuesday" gives the clinician a measurable outcome to track.

3. "What's the exit plan?" Antipsychotics for sundowning should be trialed, not prescribed indefinitely. Ask for a six-week re-evaluation and a safe tapering or de-prescribing plan if the symptoms that triggered the prescription are controlled.

4. "What monitoring is needed?" Metabolic effects (weight gain, blood sugar changes) and movement side effects (stiffness, tremor) require regular assessment. Ask what signs to watch for at home.

5. "Have we ruled out delirium?" A urinary tract infection, constipation, uncontrolled pain, or a recent medication change can cause sudden behavioral worsening that looks like disease progression but responds to treating the underlying cause — not to adding an antipsychotic.

The Medication Is One Component

Even when an antipsychotic is working, it works better alongside the non-pharmacological interventions it doesn't replace: consistent evening routines, environmental modifications, adequate lighting, and structured activities during the risk window. Medication reduces the intensity of episodes; environmental management reduces their frequency.

The Managing Sundowning Toolkit includes a medication tracking ledger and a clinical appointment worksheet that helps you document target symptoms, side effects, and dose changes — the kind of systematic tracking that turns "I think it's working" into data your parent's care team can act on.