Does behavioral improvement with haloperidol or trazodone treatment depend on psychosis or mood symptoms in patients with dementia?

OBJECTIVES: Several previous studies have examined the effects of pharmacological interventions for agitated behavior in patients with dementia. However, the choice of medication in clinical practice continues to be directed largely by local pharmacotherapy culture rather than empirical treatment guidelines. We examined the relationship between behavioral improvement and co-occurring delusions and mood symptoms in patients with dementia who were treated with haloperidol, an antipsychotic medication, or trazodone, a serotonergic antidepressant. DESIGN: Randomized, double-blind, parallel-group, 9-week treatment trial. SETTING: Inpatient geropsychiatry unit. PARTICIPANTS: Twenty-eight patients with dementia and agitated or aggressive behaviors. INTERVENTION: Haloxeridol 1 to 5 mg/day or trazodone 50 to 250 mg/day. MEASUREMENTS: Cohen-Mansfield Agitation Inventory (CMAI), Hamilton Depression Rating Scale (Ham-D), and delusional thoughts subscale and hallucinations subscale of the Behavioral Pathology in Alzheimer's Disease Rating Scale (BEHAVE-AD). RESULTS: CMAI scores improved in each treatment group over the 9 weeks of treatment (P < .001 in each group). Within the haloperidol treatment group, CMAI improvement was not associated with baseline delusional thoughts score or with change in delusional thoughts score over the course of treatment. Within the trazodone treatment group, CMAI improvement was associated with baseline score on total Ham-D (r = -0.60, P = .02), Ham-D items measuring subjective mood symptoms (r = -0.50, P = .07), and Ham-D items measuring neurovegetative signs (r = -0.49, P = .08). CMAI improvement was also associated with improvement in Ham-D total score over the course of treatment (r =0.62, P = .02). CONCLUSIONS: Mild depressive symptoms in patients with dementia and agitated behavior are associated with greater behavioral improvement by trazodone-treated patients. In contrast, the presence of delusions in concert with behavioral disturbance does not necessarily predict greater behavioral improvement with haloperidol treatment than in subjects without signs of psychosis.