Current issues of ACP Journal Club are published in Annals of Internal Medicine


Donepezil was safe and effective in Alzheimer disease

ACP J Club. 1998 Nov-Dec; 129:69. doi:10.7326/ACPJC-1998-129-3-069

Related Content in this Issue
• Companion Abstract and Commentary: Review: Tacrine has some benefits and many adverse effects in Alzheimer disease

Source Citation

Rogers SL, Doody RS, Mohs RC, Friedhoff LT, and the Donepezil Study Group. Donepezil improves cognition and global function in Alzheimer disease: a 15-week, double-blind, placebo-controlled study. Arch Intern Med. 1998 May 11;158:1021-31.



Is donepezil safe and effective for improving cognition and global function in patients who have Alzheimer disease?


12-week randomized, double-blind, placebo-controlled trial.


23 clinical centers in the United States.


468 ambulatory patients (mean age74 y, 63% women) who were ≥ 50 years of age and had probable mild to moderately severe Alzheimer disease (Mini-Mental State Examination [MMSE] score of 10 to 26) and sufficient vision and hearing to participate in the study. Exclusion criteria included type 1 diabetes mellitus, obstructive pulmonary disease, or asthma.


Patients were allocated to donepezil hydrochloride, 5 mg (n = 157) or 10 mg (n = 158) in 2 tablets per day at bedtime, or to a matching placebo (n = 153).

Main outcome measures

Cognition (Alzheimer's Disease Assessment Scale—Cognitive Subscale [ADAS-Cog]), global function (Clinician's Interview-Based Impression of Change scale [CIBIC Plus], which included information supplied by care-giver), and adverse events (i.e., treatment-emergent signs and symptoms). Secondary outcome measures included MMSE scores.

Main results

Analysis was by intention to treat. Patients allocated to donepezil, 5 mg and 10 mg, had greater improvement in their ADAS-Cog scores at 12 weeks than did patients allocated to placebo (mean score change after adjustmentfor baseline severity -2.1 for 5-mg donepezil and -2.7 for 10-mg donepezil vs 0.4 for placebo; P < 0.001 for both comparisons). Patients allocated to donepezil showed greater improvement in global function (i.e., lower CIBIC Plus scores) at 12 weeks than did those allocated to placebo (mean CIBIC Plus score 3.9 for 5-mg donepezil and 3.8 for 10-mg donepezil vs 4.2 for placebo; P < 0.01 for both comparisons). Donepezil at both doses also led to greater improvement in MMSE scores than did placebo (mean drug-placebo MMSE score differences 1.0, P < 0.04 for 5-mg and 1.3, P < 0.001 for 10-mg donepezil). The overall incidence of adverse events was similar among treatments, but 10-mg donepezil led to more nausea, insomnia, and diarrhea than did placebo (P < 0.001 for all comparisons) (Table).


In patients who have Alzheimer disease, donepezil was safe and effective for improving cognition and global function. Donepezil at 10 mg (but not 5 mg) led to more nausea, insomnia, and diarrhea than did placebo.

Sources of funding: Eisai Inc., United States; and Eisai Co. Ltd., Japan.

For correspondence: Medical Communications, Eisai Inc., Glenpointe Centre West, 500 Frank W. Burr Boulevard, Teaneck, NJ 07666-6741 USA. FAX 201-692-9183.

Table. Donepezil, 10 mg, vs placebo at 12 weeks in Alzheimer disease*

Outcomes Donepezil Placebo RRI (95% CI) NNH (CI)
Nausea 22% 8% 174% (50 to 408) 8 (5 to 17)
Insomnia 18% 5% 239% (64 to 612) 9 (6 to 18)
Diarrhea 13% 3% 408% (88 to 1294) 10 (6 to 20)

*Abbreviations defined in Glossary; RRI, NNH, and CI calculated from data in article.


In 1986, a lead article (1) in New England Journal of Medicine reported the results of a small crossover trial of oral tetrahydroaminoacridine (THA) in patients with Alzheimer disease aged 40 years and older. The accompanying editorial explained how the improvement seen in most patients validated the hypothesis that Alzheimer disease is characterized by a cholinergic deficit. The paper and editorial helped create a general expectation and, for the public, hope that cholinergic enhancement therapy with cholinesterase inhibitors would help reverse the cognitive and behavioral impairments of Alzheimer disease, just as levodopa had revolutionized treatment of another neurodegenerative disease, Parkinson disease. The 1986 trial provided what was probably the most optimistic setting in which to show the efficacy of THA (now called tacrine).

The systematic review by López Arrieta and Rodriguez Artalejo confirms what many clinicians have known for years. Only a few patients benefit from tacrine, and most cannot tolerate it. The need for frequent dosing is also a problem in patients with cognitive impairment. With the advent of alternative agents, tacrine is no longer actively marketed in the United States.

Donepezil, a second-generation cholinesterase inhibitor, has replaced tacrine. Donepezil will probably be joined soon by other agents. Indeed, 23 drugs for dementia (19 specifically for Alzheimer disease) are currently undergoing active development (2).

Rogers and colleagues provide a good assessment of the short-term efficacy and safety of donepezil. Of carefully selected patients, 10% to 20% will have gastrointestinal side effects, increased insomnia,and even agitation. The incidence of side effects will probably be greater in everyday practice, where patients are less carefully screened. The treatment effects are modest and not nearly as dramatic as the benefits of dopaminergic agents in Parkinson disease. The scales used in research to detect improvement are difficult to apply clinically. Only an occasional patient will experience a dramatic benefit. Donepezil does not alleviate behavioral disturbances and may exacerbate or unmask some symptoms (such as agitation).

In practice, for patients and families who are interested in treatment for cognitive deficits, I recommend starting with a careful assessment of function and selection of a few focused outcomes (including general well-being) that are important to the family and patient. Start with a low dose of donepezil, 5 mg/d, for 1 month to look for side effects and assess benefits in the preselected areas. If no side effects or improvement occurs, a dose of 10 mg/d can be tried. For patients who improve at either dose, it is worthwhile to continue the medication. For patients with substantial side effects or no improvement, treatment with the drug can be stopped because it does not apparently affect the long-term course of the disease. Long-term safety is unknown.

On the basis of current research, I urge clinicians to be cautious in describing the expected benefits of cholinergic treatments. The experience of Alzheimer disease can be devastating for some patients and families and may make them particularly vulnerable to unrealistic expectations, including the mistaken belief that a “magic bullet” exists for the disease.

Eric B. Larson, MD
University of Washington Medical CenterSeattle, Washington, USA

Eric B. Larson, MD
University of Washington Medical Center
Seattle, Washington, USA


1. Summers WK, Majovski LV, Marsh GM, Tachiki K, Kling A. Oral tetrahydroaminoacridine in long-term treatment of senile dementia, Alzheimer type. N Engl J Med. 1986;315:1241-5.

2. Manning A. Treatment for dementia leads mental disorder research. USA Today. 1998 June 30; Section D:7.