Clinical bottom line
Antipsychotic medications are frequently prescribed to manage behavioral and psychological symptoms of dementia (BPSD) in older adults, yet the evidence supporting their efficacy is limited and safety concerns are significant. Nonpharmacologic behavioral interventions remain the preferred first-line approach. When antipsychotics are used, careful patient selection, close monitoring for adverse effects, and regular attempts at deprescribing are essential to minimize harm. Emerging data suggest that deprescribing can be feasible without substantial worsening of symptoms in many cases, but individualized clinical judgment is required.
What the evidence shows
The use of atypical antipsychotics for BPSD has been extensively studied, with systematic reviews indicating only modest benefits in symptom reduction and a high risk of adverse events (Hirsch et al., 2005). These medications may reduce agitation and psychosis in some patients but do not improve underlying dementia or cognitive decline. Serious risks include increased mortality, cerebrovascular events, sedation, extrapyramidal symptoms, and metabolic disturbances (Annals of Internal Medicine, 2007).
Behavioral interventions based on cognitive and environmental models have demonstrated feasibility and preliminary efficacy in reducing BPSD without pharmacologic risks (Ekiz et al., 2020; Marx et al., 2019). Such approaches focus on identifying triggers and tailoring nonpharmacologic strategies to individual patient needs.
Regarding deprescribing, recent evidence suggests that discontinuation of long-term antipsychotic use in older adults with dementia is often achievable without significant symptom relapse (Van Leeuwen et al., 2018; Ricker et al., 2023). However, some patients may experience worsening behavioral symptoms, underscoring the need for close monitoring and readiness to reinstitute treatment if necessary.
Safety signals remain a concern, including rare but serious adverse events such as antipsychotic-induced hyperthermia (Nagamine et al., 2012). Alternative pharmacologic options, such as selective serotonin reuptake inhibitors (SSRIs), are under investigation but require further validation (Yang et al., 2025).
Caveats and uncertainty
Most available data derive from heterogeneous populations with varying dementia subtypes, severity, and comorbidities, limiting generalizability. Many trials have short follow-up durations, and long-term safety and efficacy remain incompletely characterized. The modest effect sizes and high placebo response rates complicate interpretation of benefit.
Behavioral interventions, while promising, require resources and training that may not be widely accessible. The optimal timing, patient selection, and protocols for antipsychotic deprescribing are not yet standardized.
Emerging pharmacologic alternatives lack robust evidence and should be considered experimental until further data are available.
How this may change practice
Clinicians should prioritize nonpharmacologic behavioral strategies as first-line management for BPSD, reserving antipsychotics for severe, refractory symptoms posing safety risks. When antipsychotics are initiated, use the lowest effective dose for the shortest duration possible, with regular reassessment.
Routine attempts to deprescribe antipsychotics should be integrated into care plans, with careful monitoring for symptom recurrence. Multidisciplinary collaboration and caregiver education are critical to support these efforts.
Awareness of serious adverse effects, including rare events like hyperthermia, should prompt vigilance and prompt intervention. Clinicians should remain informed about emerging evidence for alternative treatments but apply these cautiously pending further validation.