Evidence Table 13Head-to-head trials in patients with behavioral and psychological symptoms of dementia

Author, year
Trial name
(Quality Score)
NDurationStudy design SettingEligibility criteriaExclusion criteriaInterventions (drug, dose)Run-in/washout PeriodAllowed other medications/interventionsAge
Other population characteristics (diagnosis, etc)Number screened/eligible/enrolledNumber withdrawn/ lost to fu/analyzedOutcome measuresMethod of outcome assessment and timing of assessmentResultsMethod of Adverse Event AssessmentAdverse eventsTotal withdrawals/ Withdrawals due to AEs
Deberdt, 2005
Eli Lilly Clinical Study Summary F1D-MC-HGGU
49410 weeksDouble-blind, randomized, multicenter.
Nursing homes or assisted-living centers.
Age 40 or older. All patients exhibited clinically significant psychotic symptoms associated with Alzheimer’s disease, vascular, or mixed dementia. Dementia diagnoses defined by NINCDS-ADRDA or DSN-IV criteria. Patients must have scored ≥ 6 (severity X frequency) on the sum of the Hallucinations and Delusions items on the NPI or NPI-NH.Parkinson’s disease, Lewy-body dementia, Pick disease, frontotemporal dementia; or a MMSE score <5 or >24.Risperidone, flexible dose (0.5 to 2 mg) or
olanzapine, flexible dose (2.5 mg to 10 mg) or placebo
Atypical antipsychotic use was disallowed within 30 days, lithium or anticonvulsant use within 2 weeks before the placebo/washout period. Oral conventional antipsychotic use was allowed up to 3 days before randomization. 3 to 14-day placebo washout period.Anticholinergics (up to 6 mg per day benztropine-equivalents) and benzodiazepines (up to 4 mg per day lorazepam-equivalents) were permitted.Mean age 78.3
65.6% female
84.0% Caucasian, 9.5% African descent, 6.5% other race/ethnicity
Baseline MMSE score 13.7 olanzapine vs 14.7 risperidone vs 15.4 placebo (p=0.021 for overall treatment group difference)
81.4% Alzheimer’s dementia
5.7% vascular dementia
13.0% mixed
Number screened, eligible not reported/494 enrolled157 withdrawn/lost to followup NR/474 analyzed for primary outcomeNPI Psychosis Total, NPI Total, CGI-S Psychosis, BPRS Total, CGI-S Dementia, Cornell Total, PDS (Progressive Deterioration Scale), CMAI: Aggression.Patients were assessed weekly for the first 2 weeks of the study and biweekly thereafterMean change from baseline at endpoint, risperidone vs olanzapine:
NPI Psychosis Total: −4.2 vs −4.0 (p=0.747)
NPI Total: −0.64 vs −9.7 vs −11.8 (p=0.386)
CGI-S Psychosis : −0.7 vs −0.7 (p=0.593)
BPRS Total: −3.1 vs −3.5 (p=0.838)
CGI-S Dementia: −0.1 vs −0.0 (p=0.246)
Cornell Total: −1.2 vs −1.6 (p=0.596)
PDS: −2.9 vs −2.9 (p=0.867)
CMAI: Aggression: −1.5 vs −1.3 (p=0.781)

No significant difference vs placebo for any measure
Safety assessed from spontaneous reports of treatment-emergent adverse events, using the Coding Symbols for a Thesaurus of Adverse Reaction Terms (CoSTART) dictionary, and from vital signs, ECG, analysis of laboratory tests and MMSE changes. Motor symptoms were measured with the Simpson-Angus Scale, the Barnes Akathisia Scale, and the AIMSOn Simpson-Angus Scale, both groups increased more than placebo; greater increase in risperidone patients (+0.9 olanzapine vs +1.6 risperidone, p=0.02). No changes on AIMS or Barnes.
CVAEs: 2.5% olanzapine, 2.0% risperidone (NS)
Olanzapine vs risperidone vs placebo
Mortality: 2.9% vs 2.0% vs 1.1% (NS)
Falls: 11.3% vs 9.2% vs 6.4% (NS)
Pneumonia: 2.0% vs 0% vs 2.1% (NS)
Both active treatments associated with significantly higher incidences of somnolence, urinary incontinence, and hostility relative to placebo.
Overall: 31.1% risperidone, 37.7% olanzapine, 20.2% placebo
Due to adverse events: Not reported by group.
Most common AEs leading to withdrawal were agitation (N=6), psychotic symptoms, (N=6), somnolence (N=5), and accidental injury (N=5)
Ellingrod., 2002
198 weeksSingle-blind, nonrandomized. Four rural nursing care facilities in one city.Age 70 or older, not receiving any psychotropic drug, with DSM-IV criteria for Alzheimer-type dementia, multiinfarct dementia, or mixed syndrome, and clinical symptoms necessitating administration of an antipsychotic drug.Intracranial lesion or a history of severe head trauma.Risperidone 0.25 mg to 3 mg or olanzapine 2.5 mg to 15 mg.
Dosages determined by primary physicians.
NoneAdministration of other psychotropic drugs was allowed, although none of the study patients needed them.Mean age 85 years (SD 3, range 62–99)
79% female
Ethnicity not reported
Baseline MMSE score, risperidone vs olanzapine 14.09 (SD 5.48) vs 11.75 (SD 9.91)Number screened, eligible not reported/19 enrolled0 withdrawn/0 lost to followup/19 analyzedBrief Psychiatric Rating Scale, PANSS, Mini-Mental State Examination, Mattis Dementia Rating Scale, Abnormal Involuntary Movement Scale, Simpson- Angus Extrapyramidal Symptoms Scale, Barnes Akathisia Rating Scale, and Social Adaptive Functioning Evaluation; blood pressureAssessment at baseline, 1 month, and 2 months by one rater.Mean change from baseline at endpoint, risperidone vs olanzapine:
BPRS: −1.73 vs −0.25 (p=0.60)
SAPS: −0.64 vs −0.63 (p=0.99)
SANS: −1.27 vs 0.25 (p=0.27)
MMSE: −2.27 vs −1.38 (p=0.53)
Mattis: −10.55 vs −4.13 (p=0.29)
SAFE: 2.91 vs 1.13 (p=0.35)
AIMS, Simpson-Angus Extrapyramidal Symptoms Scale, Barnes Akathisia ScaleChange from baseline on AIMS at endpoint, risperidone vs olanzapine:
−0.18 vs 0.375 (p=0.32)
Change from baseline on Simpson-Angus at endpoint, risperidone vs olanzapine:
3.0 vs 3.25 (p=0.93)
Overall: 31.1% risperidone, 37.7% olanzapine, 20.2% placebo
Due to adverse events: NR
Fontaine, 2003
392 weeksDouble-blind, long-term care facilities.Residents of extended care facilities, meeting DSM-IV criteria for dementia; medically stable and able to comply with oral, nonliquid medication; Clinical Global Impressions scale score 4 or higher and an Alzheimer’s Disease Cooperative Study agitation screening scale score 25 or higher with 6 points on the delusions, hallucinations, physical aggression, or verbal aggression subscales.Previous neuroleptic malignant syndrome or known sensitivity to olanzapine or risperidone; current major depressive disorder or history or evidence of schizophrenia or bipolar disorder; people receiving amantadine, anorexics, carbamazepine, chloramphenicol, clonidine, erythromycin, guanabenz, guanadrel, guanethidine, guanfacine, ketanserin, methyldopa, metyrosine, narcotics, psychostimulants, reserpine, tryptophan, antiparkinsonian drugs, and benzodiazepines other than lorazepam.Risperidone 0.5, 1.0, or 2.0 mg or olanzapine 2.5, 5.0, or 10.0 mg3-day washout of psychotropic drugs.Allowed ongoing use of anticonvulsants (except for carbazepine), anti-depressants, and cholinesterase inhibitors if they had been in stable use for 30 days prior to drug washout. Allowed episodic use of antiemetics, cough/cold preparations (except those containing diphenhydramine), inhaled, topical, or ophthalmic steroids, zolpidem, and chloral hydrate. Lorazepam allowed in doses of 0.5 to 1 mg as needed for acute agitation.Mean age 83 (SD ~7.5)
67% female
Baseline MMSE score, risperidone vs olanzapine 9.3 SD 7.2) vs 7.2 (SD 7.0)Number screened not reported/47 “recruited”/39 enrolled33 withdrawn/# lost to followup not reported/39 analyzedPrimary outcome measures: Neuropsychiatric Inventory (NPI) and Clinical Global Impressions Scale (CGI)
Secondary measures: Empirical Behavioral Pathology in Alzheimer's Disease Rating Scale, Psychogeriatric Dependency Rating Scales, Multidimensional Observational Scale for Elderly Subjects, Mini-Mental Status Examination, and Quality of Life in Late-Stage Dementia Scale
Assessment at baseline, observation on days 1,2,3,5,8,10,12, and 15 by study nurse and study physician.Mean change from baseline to day 15, risperidone vs olanzapine (p-value, visit-by-drug group interaction effect, ANOVA):
CGI: −1.26 vs −1.31 (p=0.87)
NPI: −23.63 vs −15.0 (p=0.31)
E-BEHAVE-AD (Global Score):+0.52 vs +0.21 (p=0.45)
E-BEHAVE-AD (Total Score): −1.85 vs −2.26 (p=0.81)
PGDRS (Behavioral Symptoms): −4.26 vs −4.05 (p=0.91)
PGDRS (Orientation): +0.47 vs −0.21 (p=0.30)
PGDRS (Mobility): 0 vs −0.16 (p=0.07)
MOSES: −1.74 vs −0.74 (p=0.59)
QUALID: −3.53 vs −4.06 (p=0.88)
AIMS, Simpson-Angus Extrapyramidal Symptoms Scale, Barnes Akathisia ScaleChange from baseline on AIMS (% rating of minimal or mild), risperidone vs olanzapine:
no change on either (p=0.52)
Change from baseline on Simpson-Angus, risperidone vs olanzapine:
0.12 vs 0.17 (p=0.44)
Change from baseline on Barnes Akathisia Scale: (% with a rating of questionable or mild)
risperidone 0.5, 1.0, or 2.0 mg: no change (6% to 6%)
olanzapine 2.5, 5.0, or 10.0 mg: +5% (6% to 11%) (not analyzed, too few frequencies)
olanzapine: 1 stroke
No significant change in weight in either group.
113 adverse events, 31 patients had at least one adverse event.
Olanzapine: 1 patient had 2 serious adverse events (asystole followed by brain stem stroke 6 days later)
12 falls: 2 result of being pushed. Of 10 spontaneous falls, 6 olanzapine, 4 risperidone (p=0.62)
Overall: 20% olanzapine, 11% risperidone. Due to adverse events: 4 olanzapine (1 rash + elevated blood pressure, pulse, white blood cell count and temperature; 2 unsteady gait or falls; 1 diaphoresis, fainting, and asystole) vs 0 risperidone.
Gareri, 2004
Italy (POOR)
608 weeksDouble-blind, setting not reportedAge 65 or older, with DSM-IV diagnoses of Alzheimer's disease, vascular dementia, or a combination of both; NPI score of at least 24.NRRisperidone 1 mg, olanzapine 5 mg, or promazine 50 mg; if no clinical response after 4 weeks, dose could be increased to 2 mg risperidone, 10 mg olanzapine, or 100 mg promazine.10-day washoutConcomitant use of other antipsychotics, antidepressants, or mood stabilizers was avoided. Lorazepam (1 to 3 mg/day) could be given as needed until the end of the first 2 weeks.Mean age 78.9
55% female
Ethnicity not reported
Not reportedNR/NR/60NR/NR/60Primary outcome measure: NPIAssessment at baseline, 4 and 8 weeks.Complete regression of symptoms at 8 weeks (NPI):
risperidone: 14/20 (70%) (6 men, 8 women)
olanzapine: 16/20 (80%) (8 men, 8 women)
promazine: 13/20 (70%) (7 men, 6 women)

Partial response at 8 weeks (NPI) (defined differently for different groups):
risperidone: 2/20 (10%) (1 man, 1 woman)
olanzapine: 4/20 (80%) (3 men, 1 woman)

No response:
risperidone: 1/20 (70%) (1 woman, drug interrupted at 4th week because of hypotension and confusion)
promazine: 7/20 (70%) (2 men, 5 women)
Hoehn and Yahr Scale used for evaluating parkinsonism, administered at baseline, 4 weeks, and 8 weeks.Extrapyramidal symptoms not reported.
Main side effects:
olanzapine: somnolence and weight gain (32%), dizziness and constipation (16%), postural hypotension (8%), akathisia (4%), and worsening of glycemic levels in one diabetic patient (4%) risperidone: hypotension and somnolence (20%), dyspepsia (12%), sinus tachycardia, asthenia, constipation, EPS (8%) increase of libido and disinhibition, abdominal pain and insomnia (4%).
Not reported
Mulsant, 2004
866 weeksDouble-blind, multicenter, long-term care facilitiesOver age 55, with probable Alzheimer's disease, probable vascular dementia, or probable dementia of mixed etiology (by DSM-IV criteria); duration of illness of at least 1 year; MMSE scores at study entry between 7 and 26; definite psychotic symptoms, as defined by NPI frequency X severity score of >=4 on delusions, hallucinations, or both.Presence of delirium at the time of study entry as defined by the Confusion Assessment Method, an inability to swallow oral medication, a probable or definite diagnosis of psychosis prior to the onset of dementia, and an inability to otherwise cooperate with the study procedures.Risperidone 0.25 mg/day for the first 3 days, followed by an increase to 0.5 mg/day for days 3 through 6. Starting at day 7, dose increased to 0.75 mg/day until day 10, after which the investigator could increase the dose by 0.25 mg/day every 4 days if there was an insufficient clinical response. Total allowable dose 1.5 mg/day

Olanzapine starting dose 2.5 mg/day and the same titration schedule as above, with a maximum possible dose of 10 mg/day.
3-day washout, 7-day single-blind placebo run-in.Lorazepam allowed for 4 days in any 7- day period for the treatment of agitation, at a maximum dose of 3 mg/day.Mean age 83.8 78% female 77.6% white, 17.6% Hispanic, 5% blackBaseline MMSE score, risperidone vs olanzapine 13.7 (SD 5.05, range 7–25) vs 13.2 (SD 4.79, range 7–25)

81.2% Alzheimer's dementia
7.0% vascular dementia
11.8% mixed

Length of hospitalization
risperidone: 11.9 months (SD 13.5)
olanzapine: 27.1 months (SD 34.6)
NR/NR/8617/NR/85Primary outcome measures: Udvalg for Kliniske Undersogelser (UKU) rating scale measuring peripheral anticholinergic effects, or a site report of a somnolence adverse event.

Efficacy outcomes:
NPI; abbreviated cognitive assessment.
Assessments at screening, baseline, and then at weekly periods for the duration of the trial. Cognitive assessments occurred at baseline and weeks 3 and 6 (or early termination).NPI scores:
Statistically significant change from baseline for both olanzapine and risperidone on overall NPI frequency X severity, hallucinations and delusions, and occupational disruption items, but no between-group differences (data not reported).
Udvalg for Kliniske Undersogelser (UKU) rating scales measuring peripheral anticholinergic effects (including visual accommodation disturbances, dry mouth, constipation, micturition disturbances, and palpitations) or a site report of a somnolence adverse event. ESRSFor total ESRS scores, no statistically significant changes with either risperidone or olanzapine and NSD between the 2 treatments. Results for individual subscales were equivalent to the overall analyses (data not reported). No between-group differences in UKU scale or in somnolence adverse events.Overall: 19.8%
Due to adverse events: 4 risperidone vs 2 olanzapine (p=0.428)
Rainier, 2007
Austria (FAIR)
728 weeksMulticenter OutpatientsAge 55–85 years (female patients had to be at least 2 years postmenopausal); diagnosis of dementia of Alzheimer's, vascular, mixed, or fronto-temporal lobe type; behavioral disturbances, NPI Part 1 score in sub-items relating to delusions, hallucinations, agitation/aggression, disinhibition and aberrant motor behavior, and an MMSE total score of 10–26, be able to ingest oral medication and willing to complete all aspects of the study, either alone or with the aid of a responsible caregiver. Required to live with someone for the duration of the study or had substantial daily contact with a caregiver.Participation in any other drug trial within 4 weeks of the first study visit; known or suspected hypersensitivity to quetiapine or risperidone; evidence of chronic and/or severe disease; contraindications as detailed in the country-specific Prescribing Information; history of nonadherence, use of other antipsychotics; medical history of advanced, severe or unstable disease of any type that could interfere with study, current diagnosis of uncontrolled seizure disorder, active peptic ulceration, severe or unstable cardiovascular disease, acute or severe asthmatic conditions, clinically significant abnormalities on any of the following evaluations: cardiovascular, vital signs for their age, physical examination, ECG, having an authorized representative appointed by the responsible public authority; National Institute for Communicative Disorders and Stroke-Alzheimer's Disease and Related Disorders Association exclusion criteria (sudden apoplectic onset of dementia, focal neurological findings, and seizures or gait disturbances at the onset of or very early in the course of illness).Quetiapine (50 mg to 400 mg/day); mean dose 77 mg (SD 40 mg)
Risperidone (0.5 mg to 4 mg/day); mean dose 0.9 mg (SD 0.3 mg)
Not reportedAllowed antipsychotics prothipendyl 80 mg/day and dixyrazin 25 mg/day, or tranquilizers zolpidem 10 mg/day triazolam 0.25 mg/day, and oxazepam 15–50 mg/dayMean age 78 years
58.5% female
Race not reported
66.2% Alzheimer's dementia, 13.8% mixed type, 10.8% vascular dementia, 9.2% other dementia (multi-infarct, fronto-temporal lobe dementia syndrome, Lewy body dementia)NR/NR/726/1/65Primary outcome: change from baseline in NPI Part 1 (neuropsychiatric disturbances) and Part 2 (caregiver burden and distress).
CGI-Efficacy index
Age-adjusted Concentration Test
Baseline, week 4, week 8Quetiapine vs risperidone
No difference between groups on any of the NPI scores
NPI score at week 8 (between-group p-value not reported)
NPI Part 1: 17.5 vs 16.6 (both p<0.001 vs baseline)
NPI Part 2: 27.7 vs 26.7 (both p<0.05 vs baseline)
NPI Parts 1+ 2 sum of scores: 46.7 vs 44.1 (both p<0.001 vs baseline)
CMAI scores at week 8: 55.67 vs 48.97; p=0.412
CGI-I: 35.3% vs 38.8% rated 'improved' or 'very much improved' (NS)
CGI-Efficacy index response to treatment: 70.6% vs 71.0%
Incidence of AEs elicited by request, spontaneous report or observation from the patient, caregiver, or investigator.
Simpson-Angus scale, ECG, physical examination (including body weight) and vital signs
Quetiapine vs risperidone:
Any AE: 57.9% vs 44.1%
Serious AEs: 7.9% vs 2.9%
No deaths or CVAEs
Mean change from baseline in Simpson-Angus scale score: + 0.06 quetiapine vs +0.35 risperidone (both NS)
No significant change in body weight
10.5% quetiapine, 8.8% risperidone/
5.3% quetiapine, 2.9% risperidone
Schneider, 2001
Schneider, 2006
Ismail, 2007
Zheng, 2009
Sultzer, 2008
CATIE Trial (Phase 1) (FAIR)
421Up to 36 weeksDouble-blind, multicenter, placebo- controlled
Outpatients or assisted living facilities
Dementia of the Alzheimer's type or probable Alzheimer's disease ; MMSE score between 5 and 26 ; ambulatory, living at home or in an assisted-living facility. Delusions, hallucinations, aggression, or agitation that developed after the onset of dementia and was severe enough to disrupt their functioning and justify treatment with antipsychotic drugs. Signs and symptoms of psychosis, aggression, or agitation had to have occurred nearly daily during the previous week or at least intermittently for 4 weeks. During the week before they were randomized, a severity rating of at least "moderate" for conceptual disorganization, suspiciousness, or hallucinatory behavior on the BPRS. Alternatively, a frequency rating of "often" or "more frequently" and a severity rating oat least "moderate" for delusions, hallucinations, agitation, or "aberrant motor behavior" in the NPI. A study partner or caregiver who had regular contact with the patient was required to participate in the assessments.Diagnosis of a primary psychotic disorder, delirium, other dementia such as vascular dementia or Lewy-body dementia, or psychosis, agitation, or aggression that could be better accounted for by another medical condition, medication, or substance abuse. If they required psychiatric admission, were suicidal, were going to receive treatment with a cholinesterase inhibitor or antidepressant medication, had previously been treated with two of the three atypical antipsychotic drugs under study, or had contraindications to any of the study drugs.Doses were adjusted as clinically indicated by study physicians.
Mean daily dose (range) at last dose:
olanzapine: 5.5 mg (0–17.5)
quetiapine: 56.5 mg (0–200)
risperidone: 1.0 mg (0–2)
Not reportedTo treat difficult behaviors during the trial, study physicians could prescribe a benzodiazepine or oral or parenteral haloperidol.Mean age 77.9 (7.5)
56% female
79% white, 18% black, 3% other race
73% lived in own home, 16% in family's home, 10% assisted living facility, 2% other residence521/471/421344/0/416Time to discontinuation for any reason (primary outcome)
CGI-C at week 12
Time to discontinuation for lack of efficacy
Time to discontinuation for adverse events, intolerability, or death
Change in metabolic parameters
Not describedDiscontinuation for any reason:
olanzapine: 79/99 (80%)
quetiapine: 77/94 (82%)
risperidone: 65/84 (77%)
placebo: 118/139 (85%)

Discontinuation for lack of efficacy:
olanzapine: 39%
quetiapine: 53%
risperidone: 44%
placebo: 70%
olanzapine vs risperidone: Hazard ratio 0.84 (95% CI 0.53, 1.32)
olanzapine vs quetiapine: Hazard ratio 0.63; (95% CI 0.41, 0.96; p=0.02)

Response based on CGI-C score at week 12 (p vs placebo):
olanzapine: 32%
quetiapine: 26%
risperidone: 29%
placebo: 21%

Worsening of symptoms in Olanzapine group compared to Quetiapine in BPRS withdrawn depression factor LSM differences=0.3, 95% CI=0.1 to 0.5, p=0.009
Assessed by eliciting information about the occurrence of AEs. Weight, prolactin, glucose, cholesterol, and triglyceride levels were measured at weeks 12, 24, and 36.Olanzapine vs quetiapine vs risperidone vs placebo:
Any serious AE: 14% vs 18% vs 11% vs 13%
Cerebrovascular accident or TIA: 2% vs 1% vs 1% vs 1%
Death: 1% vs 3% vs 1% vs 2%
Any severe AE: 17% vs 26% vs 14% vs 15%
Parkinsonism or EPS: 12% vs 2% vs 12% vs 1% p<0.001

Rate of change (SE) in metabolic measures per week of treatment
Weight (pounds)
Olanzapine: 0.12 (0.06) p=0.032, Quetiapine: 0.14 (0.06), p=0.019, Risperidone: 0.10 (0.06), p=0.07
HDL Cholesterol (mg/dl)
olanzapine: −0.19 (0.07), p=0.0004, Quetiapine: −0.09(0.07), p=0.19, Risperidone=0.03 (0.07), p=0.68
Triglyceride (mg/dl)
Olanzapine: 0.42 (0.59), p=0.48, Quetiapine: 0.55 (0.63), p=0.38, Risperidone: 0.43 (0.65), p=0.51
Olanzapine: 0.10 (0.11), p=0.37, Quetipaine:0.12(0.12), p=0.32, Risperidone: 0.24(0.13), 0.06
Magnitude of weight gain (pounds) after 12 weeks
Olanzapine: 1.4, Quetiapine: 1.7, Risperidone: 1.2
82% withdrew
Discontinuation because of intolerance, adverse events, or death: Hazard ratio vs placebo (95% CI):
olanzapine: 24%: 4.32 (1.84, 10.12)
quetiapine: 16%: 3.58 (1.44, 8.91)
risperidone: 18%: 3.62 (1.45, 9.04)
placebo: 5%

From: Evidence Tables

Cover of Drug Class Review: Atypical Antipsychotic Drugs
Drug Class Review: Atypical Antipsychotic Drugs: Final Update 3 Report [Internet].
McDonagh M, Peterson K, Carson S, et al.
Portland (OR): Oregon Health & Science University; 2010 Jul.
Copyright © 2010 by Oregon Health & Science University, Portland, Oregon 97239. All rights reserved.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.