Efficacy of increasing donepezil in mild to moderate Alzheimer's disease patients who show a diminished response to 5 mg donepezil: a preliminary study


Correspondence: Dr Yusuke Yatabe MD PhD, Department of Neuropsychiatry, Faculty of Medical and Pharmaceutical Sciences, Kumamoto University, 1-1-1, Honjo, Chuou-ku, Kumamoto-city, Kumamoto 860-8556, Japan. Email: yatababychan@yahoo.co.jp



With the recent approval of several new drugs, pharmacological management of Alzheimer's disease has become more complicated in Japan. The efficacy and safety of increasing the dose of donepezil to 10 mg daily were assessed in an open-label study of patients with mild to moderate Alzheimer's disease who were showing a diminished response to 5 mg daily.


The subjects included 27 patients with mild to moderate probable Alzheimer's disease whose primary caregivers had confirmed progression of symptoms during treatment with donepezil 5 mg daily. The dose of donepezil was increased to 10 mg daily, and the Alzheimer's disease assessment scale-cognitive subscale (Japanese version), Neuropsychiatric Inventory, and Zarit caregiver burden interview scores were compared before and after dose escalation. Adverse events were also investigated.


Efficacy was evaluated in 24 patients; three dropped out because of adverse reactions. The Alzheimer's disease assessment scale score showed significant improvement after dose escalation of donepezil (P = 0.006). The total score of the Neuropsychiatric Inventory and the Zarit score showed no significant changes. However, the anxiety score of the Neuropsychiatric Inventory showed a significant increase (P = 0.028). Safety assessment revealed that the dropout rate was 11.1% and adverse reactions occurred in 40.7%. Nausea (29.6%) and loss of appetite (22.2%) were common adverse reactions.


Because cognitive function showed improvement after increasing the dose of donepezil, the dosage of this drug should probably be adjusted based on the overall severity of Alzheimer's disease as well as the progression of cognitive dysfunction.


Donepezil is used to delay the progression of cognitive dysfunction in Alzheimer's disease (AD). In Japan, this drug was first marketed in 1999 with the indication of ‘mild to moderate’ AD, because the number of acetylcholinergic cells was thought to be decreased in severe AD, which results in a weaker response to donepezil. Subsequently, a clinical study of donepezil was performed in patients with severe AD using higher doses, and satisfactory results were obtained with regard to both cognitive function and general symptoms.[1] As a result, treatment with donepezil 10 mg daily was approved for severe AD in August 2007. In 2011, three new anti-dementia drugs (galantamine, rivastigmine and memantine) became available in Japan. This increased range of choices for anti-dementia therapy has enabled clinicians to treat patients who could not tolerate donepezil because of adverse reactions. However, it has also made the treatment of AD more complicated, and clinicians are now required to select the most appropriate drug for each patient. When a treatment for AD is selected, it is not only important to choose an appropriate drug type for each patient and disease stage, but also to adjust the dosage based on the progression of symptoms.

It has often been found that symptoms can be controlled for some time by donepezil 5 mg daily in patients with mild to moderate AD, but symptoms eventually begin to progress again. In such clinical situations, co-administration of memantine may be one choice for patients with moderate AD. However, because symptom progression was initially inhibited by donepezil, it could be that the dose may be too low and should be increased to 10 mg daily. In the present study, the daily dose of donepezil was increased from 5 mg to 10 mg, and its efficacy and safety were evaluated in patients with mild to moderate AD who had shown a decreased response to donepezil at 5 mg.



All procedures of this study strictly followed the Clinical Study Guidelines of the Ethics Committee of Kumamoto University Hospital (Kumamoto, Japan) and were approved by the committee. After patients or their caregivers were provided with a complete description of all procedures of this study, including the fact that administration of donepezil 10 mg daily for mild to moderate AD involved off-label use of the drug, written informed consent was obtained.

The subjects were selected from patients who presented to the dementia outpatient clinics of the Department of Neuropsychiatry at Kumamoto University Hospital and two outpatient clinics of the Kumamoto Prefectural Dementia-Related Disease Medical Center between April 2008 and March 2011. Patients were required to meet all of the following inclusion criteria:

  • patients who met the diagnostic criteria of the National Institute of Neurological and Communicative Disease and Stroke and Alzheimer's Disease and Related Disorders Association for probable AD[2]
  • patients with mild to moderate AD, which was defined by a Mini-Mental State Examination score ≥10 and a Clinical Dementia Rating ≤2[3, 4]
  • patients who had been taking donepezil 5 mg daily for at least 3 months
  • patients who had had symptom progression confirmed by family members (caregivers) living with them
  • patients whose primary caregiver lived with them and could bring them to hospital throughout the study period.

Because the study involved off-label use of donepezil, patients were excluded if the attending physician considered them unsuitable for treatment with a large dose of the drug. This meant excluding patients over 80 years old, patients with serious complications (bradycardia, asthma, gastric ulcer and hepatic dysfunction), and patients whose mental symptoms were expected to worsen after dose escalation.

A total of 27 patients (6 men, 21 women) who met the above inclusion criteria were enrolled in this study. Their mean age was 66.6 ± 7.6 years, mean duration of education was 10.8 ± 2.4 years, and mean duration of AD was 3.0 ± 1.1 years. The mean Mini-Mental State Examination score was 18.6 ± 3.3 points. The Clinical Dementia Rating was 0.5, 1 and 2 in 7 patients, 19 patients, and 1 patient, respectively. The mean treatment period with donepezil 5 mg daily was 13.9 ± 8.3 months. The mean age of their caregivers was 63.1 ± 11.7 years. Most caregivers were spouses (n = 21), followed by children (n = 3), daughters-in-law (n = 2) and parents (n = 1).

Dose escalation protocol

In this open-label study, the dose of donepezil was increased from 5 mg to 10 mg daily after consent was obtained. Donepezil was administered orally once daily after breakfast under the supervision of the patient's caregiver. Information about compliance was obtained from the caregiver by interview. If patients discontinued treatment for at least 1 week for any reason, they were classified as dropouts and excluded from the efficacy evaluation.

Evaluation of efficacy

For evaluation of efficacy, cognitive function was assessed by the Alzheimer's Disease Assessment Scale-cognitive subscale (Japanese version) (ADAS-J cog),[5, 6] behavioural and psychological symptoms of dementia (BPSD) were investigated by the Neuropsychiatric Inventory (Japanese version) (NPI),[7, 8] and the burden on the caregivers was determined by the Zarit Caregiver Burden Interview (Japanese version) (ZBI).[9, 10] These assessments were performed within 4 weeks before the dose of donepezil was increased and 8 weeks after dose escalation, and the results were compared. In addition, impressions about the patients after dose escalation were obtained from the caregivers by interview.

Evaluation of safety

Information about adverse events that occurred during the study period was obtained by interviewing patients and their primary caregivers at 8 weeks after dose escalation or at the time of dropout from the study. In addition, a 12-lead electrocardiogram was recorded within 4 weeks before and at 8 weeks after dose escalation.

Statistical analysis

The paired t-test was used to compare the ADAS-J cog score, NPI total score and ZBI score before and after increasing the dose of donepezil. Wilcoxon signed rank test was used to compare the scores for NPI subscales. The number of patients with each adverse event (including electrocardiogram findings) was determined, and percentages were also calculated. A significance level of 0.05 (two-tailed) was set for all analyses, which were carried out using SPSS for Windows, version 17.0 (SPSS, Chicago, IL, USA).


Evaluation of efficacy

Efficacy was evaluated in 24 patients (6 men, 18 women; mean age: 64.9 years; mean Mini-Mental State Examination score: 17.8), after excluding the three who dropouts as a result of adverse events. The mean ADAS-J cog total score was 17.1 ± 7.5 before dose escalation of donepezil. It improved to 15.3 ± 6.9 after dose escalation, a significant difference (P = 0.006) (Table 1). The NPI total score showed no significant change after dose escalation, but the score for anxiety became significantly worse among the NPI subscales (P = 0.028). However, scores for other NPI subscales showed no significant change after dose escalation (Table 2). The mean ZBI score was 16.1 ± 11.6 before dose escalation and 17.6 ± 13.3 points after escalation, showing no significant difference (P = 0.224). Caregivers evaluated the response to the increased dose of donepezil in a positive manner, as follows: ‘the patient has become slightly more cheerful’, ‘the patient has become more positive’, ‘the patient wants to increase the frequency of day service’ and ‘the patient feels that progression of symptoms has stopped’.

Table 1. ADAS-J cog scores before and after dose escalation of donepezil
  1. Data are shown as the mean ± SD. ADAS-J cog, Alzheimer's Disease Assessment Scale-cognitive subscale (Japanese version).
Word recall6.4 ± 1.46.0 ± 1.70.177
Word recognition3.4 ± 2.62.9 ± 2.30.352
Orientation3.4 ± 2.03.0 ± 1.90.233
Recall of test instructions0.4 ± 1.10.3 ± 0.80.083
Following commands0.9 ± 0.90.7 ± 0.70.203
Naming objects/fingers0.2 ± 0.50.2 ± 0.60.328
Word finding difficulty0.2 ± 0.40.2 ± 0.51.000
Spoken language ability0.0 ± 0.20.0 ± 0.00.328
Comprehension0.1 ± 0.40.1 ± 0.60.328
Construction0.9 ± 0.80.7 ± 0.60.213
Praxis1.3 ± 1.61.2 ± 1.50.610
Total ADAS-J cog17.1 ± 7.515.3 ± 6.90.006
Table 2. NPI scores before and after dose escalation of donepezil
  1. Data are shown as the mean ± SD. NPI, Neuropsychiatric Inventory.
Delusions0.5 ± 1.20.1 ± 0.30.102
Hallucinations0.0 ± 0.00.0 ± 0.01.000
Agitation1.0 ± 1.61.0 ± 2.10.823
Depression0.8 ± 1.31.3 ± 2.30.301
Anxiety0.4 ± 1.11.4 ± 2.60.028
Euphoria0.0 ± 0.00.0 ± 0.01.000
Apathy3.2 ± 3.93.0 ± 3.50.943
Disinhibition0.0 ± 0.20.2 ± 0.80.414
Irritability0.4 ± 0.80.9 ± 2.30.676
Aberrant motor behaviour0.9 ± 2.30.6 ± 1.60.500
Total NPI score7.1 ± 7.18.3 ± 9.50.483

Evaluation of safety

Of the 27 patients, three were classified as dropouts and the discontinuation rate was 11.1%. In two dropouts (women aged 57 and 79 years), the dose of donepezil was reduced to 5 mg daily because of the onset of nausea and loss of appetite within 1 week after dose escalation. In the other patient (a woman aged 76 years), loss of appetite occurred after dose escalation. Subsequently, she suffered a compression fracture of the lumbar spine and developed depression. At the request of her family, the dose of donepezil was reduced to 5 mg daily after 5 weeks of treatment at 10 mg. In all three dropouts, gastrointestinal symptoms resolved rapidly after dose reduction.

The following adverse events were noted in 11 (40.7%) of the 27 patients (including the dropouts): nausea occurred in eight patients (29.6%), loss of appetite in six patients (22.2%), pollakiuria in three patients (11.1%), excitement in two patients (7.4%), and diarrhoea in one patient (3.7%). In one patient (3.7%), the electrocardiogram revealed supraventricular extrasystoles after dose escalation.


In this study, the ADAS-J cog score showed significant improvement after the daily dose of donepezil was increased from 5 mg to 10 mg. It has already been confirmed that the efficacy of this drug is dose-dependent. In earlier studies of the correlation between the dose and efficacy of donepezil, the ADAS score showed significant improvement in the 10 mg group compared with the 5 mg group.[11-13] Unlike these studies, the dose-response effect of donepezil in the present study was investigated in a group of patients who had shown progression of symptoms. Their symptoms had been stable for some time on donepezil 5 mg daily, but began to worsen again over time. Therefore, this study was focused on countermeasures for a common problem in clinical practice. Our results suggest that increasing the dose of donepezil to 10 mg daily is reasonable in patients with mild to moderate disease who have become less responsive to the drug at 5 mg daily.

Although cognitive function improved after dose escalation, the NPI total score showed no improvement and the anxiety score became worse. There have been some reports about the efficacy of donepezil for BPSD in patients with mild to moderate AD.[14, 15] For example, in an earlier study, improvement of BPSD (including anxiety) was noted.[14] Also, Hori et al. reported that administration of donepezil to AD patients could improve cognitive function and lead to a better understanding of their problems, but this resulted in an increase in anxiety, a phenomenon known as ‘awakening’.[16] The aggravation of anxiety observed in the present study might have been ascribable to this phenomenon. Attention is often focused on apathy when evaluating the efficacy of donepezil for BPSD. Apathy is considered to be related to a decrease of acetylcholine.[17] It has been reported that cholinesterase inhibitors are effective for apathy and that donepezil 10 mg daily was effective for apathy in AD patients.[18-21] In the present study, improvement of apathy was not seen among the NPI subscales. However, some caregivers noted improvement in apathy in their patients. For example, they stated that the ‘patient became more cheerful’, ‘the patient became more positive’ or ‘the patient wanted to increase the frequency of day service’. Accordingly, further studies are needed concerning the influence of a higher dose of donepezil on apathy. In this study, the NPI data suggested that increasing the dose of donepezil to 10 mg should not be performed with the expectation of improving BPSD in patients with mild to moderate AD and that attention should be paid to the risk of aggravation of anxiety.

After the dose of donepezil was increased in this study, cognitive function improved, but the ZBI score did not. Germain et al. investigated the burden on the caregivers of 1091 patients with mild to moderate AD using the ZBI. They reported that the relative influence on the burden placed on the caregivers was in the following order: BPSD, difficulty in performing activities of daily living, caring time, caregiver characteristics (such as age and sex) and cognitive dysfunction.[22] This suggests that it is probably difficult to reduce the caregiver burden by improving cognitive function in AD patients. If the results of the present study are taken into consideration as well, the effect of high-dose donepezil on cognitive dysfunction is probably not strong enough to materially reduce the caregiver burden. This may be a general limitation of current anti-dementia drugs.

Evaluation of safety showed that the incidence of adverse events was 40.7% after dose escalation of donepezil, which was lower than the incidence of 47% reported in Japanese patients with severe AD receiving donepezil 10 mg daily.[1] This difference in incidence of adverse events was presumably related to the difference in the severity of AD and also to the lower mean age of the patients in our study (67 years vs 75 years). In the present study, the discontinuation rate as a result of adverse events was 11.1%, which was similar to the rate for the donepezil 10 mg group in the above-mentioned study (14%).[1] Furthermore, an earlier Japanese study of donepezil 10 mg daily for mild to severe AD showed that the incidence of adverse events requiring discontinuation was 12%.[23] These results suggest that the percentage of patients who cannot tolerate donepezil at 10 mg daily is around 10% irrespective of age and the severity of AD.

It should be noted that this study had some limitations. First, it was an open-label uncontrolled study, so the reliability of our results is lower than if we had done a placebo-controlled double-blind study. Second, progression of symptoms was assessed by the subjective evaluation of caregivers (not combined objective evaluation such as neuropsychological examinations of patients). The caregivers lived with the patients and could directly observe any changes at close range, so their impressions were likely to be suitable for determining efficacy of donepezil dose escalation. Future studies will need to use neuropsychological assessments in addition to the caregivers' impressions. Similarly, quantitative assessments such as the caregiver-rated Clinical Global Impression of Change might be necessary in the objective evaluation of caregivers' impression. Third, the ADAS-cog data needs to be assessed in the context of a practice effect influence from repeated neuropsychological testing. However, a practice effect on the ADAS has not been noted in AD patients with memory impairment, but it has been reported healthy volunteers.[24] Fourth, because administration of donepezil 10 mg for mild to moderate AD involves off-label use, we excluded patients with a higher risk of aggravation of BPSD (such as those with delusions, agitation and irritability) and patients aged over 80 years. Accordingly, our results cannot be applied to the overall population with mild to moderate AD.

Despite these limitations, the finding that ADAS-J cog showed significant improvement after dose escalation of donepezil suggests a useful approach to the treatment of cognitive dysfunction in AD patients. Therefore, we propose that the dose of donepezil should not only be adjusted on the basis of the severity of AD but also by taking into account the progression of cognitive dysfunction. In contrast, we found no improvement of BPSD when the dose of donepezil was increased to 10 mg in our patients with mild to moderate AD. Thus, if progression of BPSD occurs during treatment with donepezil 5 mg, a beneficial effect cannot be expected by increasing the dose. In conclusion, treatment of AD has become more complicated since the release of several anti-dementia drugs, so it is necessary to conduct further studies to establish the optimum regimens for using the new drugs.


The present study was undertaken with the support of grants provided by the Ministry of Health, Labour and Welfare (Research on dementia) for M.I and M.H.