Effect of activity reminiscence therapy as brain-activating rehabilitation for elderly people with and without dementia


Professor Haruyasu Yamaguchi, Department of Physical Therapy Gunma University School of Health Sciences, 3-39-15 Showa, Maebasi-City, Gunnma 371-8514, Japan. Email: yamaguti@health.gunma-u.ac.jp


Background:  Preventing the progression of dementia is a widespread challenge. However, currently there is limited evidence supporting the effectiveness of dementia rehabilitation.

Methods:  We practiced activity reminiscence therapy (ART) as brain-activating rehabilitation for both lucid and demented persons (n = 18) in a day-service setting as well as in a group home. The ART sessions were conducted 1 hour every week for 12 weeks (intervention period). We compared the results of three cognitive tests (the Mini-Mental State Examination, the Kana Pick-out test and the ‘logical memory’ component of the Wechsler Memory Scale-Revised) and four behavior and caregiver's burden scales (the Clinical Dementia Rating, the Multidimensional Observation Scale for Elderly Subjects, the Dementia Behavior Disturbance scale and the Zarit Caregiver Burden Interview) conducted during the control period with those taken during the intervention period. At the end of the intervention period, we interviewed the staff and families individually to assess whether the participants seemed to have changed after intervention and, if so, how.

Results:  In cognitive tests, only immediate and delayed recall of the Wechsler Memory Scale-Revised showed significant improvement. None of the four behavior and caregiver's burden scales showed any significant changes after intervention. However, the interviews showed improvements in subjective aspects of communication, interaction and behavior.

Conclusion:  ART uses old-style tools. The nostalgia brought about by using these familiar tools led to effective recall of experiences, in which the participants taught the staff how to use the tools, which were unfamiliar to the staff. Through this role-reversal, they gained a sense of self-worth and a desire to live. Due to the reconstructed relationship between participants and caregivers, we consider ART to be effective in maintaining and improving emotional functions, activities of daily living and memory. ART should be useful for both lucid and mildly demented persons as brain-activating rehabilitation therapy.


Preventing the progression of dementia is a widespread challenge. However, there is currently limited evidence supporting the effectiveness of cognitive rehabilitation.1,2

Brain-activating rehabilitation,3 which this study examines, is different from cognitive rehabilitation, and is expected to activate the remaining cognitive ability and positively affect motivation to live through enjoyable activities and playing a social role. Brain-activating rehabilitation provides an environment in which demented patients can laugh and be themselves, resulting in increased activity and peace of mind. In addition, through this therapy, relatives of thepatients can increase an understanding of the disease and how they can interact with the affected relative.

Our brain-activating rehabilitation has guidelines. The activities aim to be enjoyable and comfortable. They should be associated with communication and playing a social role. As effective communication is promoted, the patient gains a sense of self-worth and a desire to live through playing a social role.

A typical brain-activating rehabilitation technique is ‘reminiscence using old-style tools’, or activity reminiscence therapy (ART).4 ART can assist memory recall in demented persons because old-style tools, such as a rice kettle, beanbags for juggling and old textbooks, may activate strong memories. Another characteristic of ART is ‘role-reversal’. Elderly people, who are usually helped by young care staff in the facilities, can teach the staff how to use tools, in which they are experienced. These tools encompass many activities, from housework to recreation to handiwork. Through this process, elderly people recover their sense of social function in a natural role, which is the teaching and handing-down of knowledge from the elderly to the young.

In this paper, we practiced ART as brain-activating rehabilitation for both lucid and demented persons in a day-service setting, as well as in a group home, to examine the effectiveness of this rehabilitation technique.



Of 22 participants, four left the study due to unforeseen circumstances (sickness, change of hospital, etc.). Finally, 18 participants (82.2 ± 7.3 years) took part in nine or more out of a total of 12 ART sessions, and completed three evaluations. Our test group consisted of six residents (all women) of the Ooido Group Home and 12 users (five men and seven women) of the Day-Service Center Ooido (Table 1). Of the 18 participants, one participant was mentally healthy (the Clinical Dementia Rating (CDR)5 0), and three were forgetful, but did not satisfy the International Classification of Disease-10 (ICD-10) criteria of dementia (CDR 0, one person; CDR 0.5, two). The remaining 14 participants satisfied the ICD-10 criteria of dementia (CDR 0.5, two persons; CDR 1, nine; CDR 2, three). Of these 14 demented participants, eight had Alzheimer's disease and six had vascular dementia. Donepezil, an anti-Alzheimer medication, was taken by four participants. One demented participant (CDR 1) started taking donepezil during the intervention period, and three had started it before the trial and continued during the entire study period. Two participants took antidepressant drugs during the study with constant dosage. Antipsychotics were used for three participants. Two participants (CDR 1) started antipsychotics during the intervention period, whereas one (CDR 1) continued it during the whole study with constant dosage. No participants widely changed their environment. We obtained informed consent from all participants and their families.

Table 1.  Characteristics of participants
 TotalCDR 0CDR 0.5CDR 1CDR 2
  1. Results are expressed as: mean ± standard deviation.

  2. CDR, clinical dementia rating.

Number (M/F)18 (5/13)2 (0/2)4 (2/2)9 (3/6)3 (0/3)
Age82.2 ± 7.385.5 ± 6.478.3 ± 7.582.7 ± 8.684.0 ± 1.0
MMSE20.3 ± 4.624.5 ± 0.724.3 ± 2.219.1 ± 4.315.3 ± 2.9


Our study was divided into two three-month periods: control and intervention periods. During the intervention period, we divided all participants into three groups, because ART is more effective in a small group, consisting of less than 10 participants. Group members were basically fixed to reinforce the familiar relationship. ART sessions were conducted in each group for 1 hour each week. Table 2 shows the topics of the ART sessions. The first 10 min of each session involved a reminiscence warm-up activity. The participants watched a video in which the ART topic tools were used, in order to provide visual support for the reminiscence. Next, we asked the participants about the tool names and how to use them. The tools were passed around among the participants so each person could demonstrate how to use the tools at least once. The ART staff consisted of one leader and two co-leaders for each group. The staff from both the group home (six persons) and the day-service center (seven persons) had studied the method of ART in preceding meetings, and they played the role of leader and co-leaders by rotation. They organized the sessions so that each participant could socially interact with other participants, as well as individually playing a leading role in demonstrating the use of the tools. Finally the staff thanked the participants for their instruction and the session ended.

Table 2.  Topics of activity reminiscence therapy (ART) sessions
 Topics (Japanese)Tools
1Cooking rice (suihan)Rice kettle, furnace, measure, rice, charcoal
2Traditional games (mukasi no asobi)Beanbag for juggling, cup-and-ball toy, propeller
3Making vegetables pickled in rice-bran paste (nuka zuke)Germ, earthenware pot, vegetables, salt, water
4Washing (sentaku)Tub, washing board, washing powder
5Sewing duster (saihou)Cloth, stitch, thread, pincushion
6Making rice cake (mochitsuki)Mortar, mallet, steam pot, glutinous rice
7Usage of rice-bran (komenuka noriyou)Rice-bran, rice-bran bag, washbowl
8Grinding (suribachi de suru)Grind pot, grind stick, taro potato
9Hand made snack (tezukuri oyatsu)Barley flour, rice-bowl, spoon
10Weaving (hataori)Silkworm, pirn, shuttle
11Elementary school (syougakkou no omoide)Old textbook, imperial speech on education
12Making noodles (udon-uchi)Knead pot, rolling pin, board

At the beginning of the control period (evaluation 1), the participants underwent the Mini-Mental State Examination (MMSE),6 the Kana Pick-out tests and the Wechsler Memory Scale-Revised (WMS-R),7,8 where a logical memory test was performed (immediate and delayed recall). The care staff evaluated participants using CDR and the Multidimensional Observation Scale for Elderly Subjects (MOSES).9 The care managers interviewed participants' families, and evaluated the Dementia Behavior Disturbance scale (DBD).10 The families of the day-service users (n = 11) completed the Zarit Caregiver Burden Interview (ZBI)11 and an interview list.12 After these various tests, there was no therapy performed for a period of 3 months. Then the participants underwent a second identical evaluation just before the intervention (evaluation 2). During the subsequent 3 months the participants received ART. A third identical evaluation was administered at the end of this second 3-month intervention period (evaluation 3). We prepared two questions: did the participants change after intervention with ART and, if so, how?; and did your daily interaction with the participants change in terms of both physical care and social contact? The first question was asked of the nine staff members and the participant families individually. The second was asked of the staff members only.

Statistical analysis

Statistical analysis was performed using SPSS 14.0. One-way analysis of variance (anova) was used to compare the means among evaluations. If anova was significant, Dunnett test was used as a post hoc test.



There were 12 ART sessions (Table 2). Each participant took part in at least nine sessions, and the average attendance (mean ± SD) was 93.5 ± 9.3%.

Baseline changes in control period (between evaluations 1 and 2)

We performed three cognitive tests (MMSE, the Kana Pick-out test and WMS-R) at the beginning (evaluation 1) and end (evaluation 2) of the 3-month control period. These scores did not show significant differences between evaluation 1 and evaluation 2. Of the next four scales (CDR, MOSES, DBD and ZBI), which were performed by staff and family members, only the ‘self-care functioning’ area of MOSES worsened moderately (increased score; P = 0.033) (Table 3).

Table 3.  Results of the tests scores
ScaleEv. 1Ev. 2Ev. 3Statistical significance
Ev. 1 vs 2Ev. 2 vs 3
  • Results are expressed as: mean ± standard deviation.

  • Dunnett test. NS, not significant; Ev, evaluation.

MMSE20.3 ± 4.619.5 ± 3.820.5 ± 5.2NSNS
Kana Pick-out test8.9 ± 5.38.3 ± 4.86.6 ± 4.4NSNS
 Immediate recall4.2 ± 3.43.8 ± 4.05.5 ± 4.5NSP = 0.010
 Delayed recall0.5 ± 1.51.4 ± 2.83.5 ± 5.0NSP = 0.022
CDR0.9 ± 0.60.9 ± 0.51.1 ± 0.7NSNS
 Total63.2 ± 11.967.5 ± 18.770.3 ± 17.2NSNS
 Self-care functioning11.4 ± 2.413.1 ± 3.813.3 ± 4.0P = 0.033NS
 Disoriented behavior15.4 ± 5.115.3 ± 5.715.9 ± 4.8NSNS
 Depressed/anxious mood10.8 ± 2.810.8 ± 3.412.2 ± 4.8NSNS
 Irritable behavior10.8 ± 3.910.6 ± 3.511.3 ± 3.6NSNS
 Withdrawn behavior14.8 ± 5.916.8 ± 6.917.6 ± 7.3NSNS
DBD19.4 ± 13.719.4 ± 13.319.7 ± 13.6NSNS
ZBI (n = 11)17.2 ± 13.719.4 ± 20.520.5 ± 15.5NSNS

Changes in the intervention period with ART (between evaluations 2 and 3)

To evaluate the effect of ART, we performed the same three cognitive tests and four scales at the beginning (evaluation 2) and end (evaluation 3) of ART. In the cognitive tests, immediate and delayed recall of WMS-R showed significant improvement (P = 0.010 and P = 0.022, respectively). Immediate recall of WMS-R was preserved or improved in all participants. Delayed recall was improved in participants with a CDR score of 0 or 0.5, but not in those with a CDR score of 1 or 2, who were unable to perform delayed recall due to dementia (Fig. 1). MMSE and the Kana Pick-out test showed no significant changes. None of the four scales (including the ‘self-care functioning’ area of MOSES) showed significant changes after intervention (Table 3).

Figure 1.

We divided the participants into two groups according to clinical dementia rating (CDR) scores. One group consisted of six participants with CDR 0 and 0.5 (—) who showed improvement of both immediate (a) and delayed (b) recall by intervention between evaluation (Ev.) 2 and 3. The other group consisted of 12 participants with CDR 1 and 2 (- - - -) who showed unremarkable change.

Changes based on subjective observation

As a part of evaluation 3, we interviewed care staff and family members individually to assess whether the participants (n = 18) had changed after intervention with ART and, if so, how. All care staff (n = 9) and six of 12 family members answered as follows:

  • 1A more cheerful and positive outlook: According to family members, three of the six participants showed an increased level of positive behavior (more frequent smiling, more vital personality, general well-being, geniality and increased mellowness) and social interaction with their peers after ART. The care staff also reported positive behavior in nine of the 18 participants.
  • 2Positive anticipation of ART: According to family members and care staff, four of the six participants with a CDR score of 0 or 0.5 exhibited eagerness to participate in ART, and three of the four voluntarily prepared clothing that was required when they learned about the next week's ART topic. For example, when they heard the next ART topic would be ‘handmade noodles’, they prepared their own apron and head scarf to wear. In addition, they showed more interest in their hairstyle and clothing in daily life.
  • 3Improved peer relationships: According to the care staff, two participants showed improved peer relationships. One participant, who resides at the Ooido Group Home, has mild dementia. She did not know how to interact with other severely demented residents, so she had avoided contact. After ART, she initiated conversations with other residents and tried to develop friendships with them. Another day-service user, who had always remained with one specific friend, began to interact with many other users after ART.
  • 4Co-operative attitude toward the care staff: According to the care staff, one participant showed improved co-operation with the care staff.

Significant cases

Case 1

One participant, whose CDR score is 0.5, dramatically improved. She was depressed at the beginning of ART, because she could not perform the various activities well in contrast with the dexterity exhibited in her youth. With comfort and encouragement from the care staff, eventually she was able to readily participate in each ART session. She continued to participate, and her confidence gradually improved until finally she saw ART as a positive and joyful experience. She talked about ART topics a great deal with her family. Her tests and scales scores improved as follows: MMSE 21–25, immediate and delayed recall of WMS-R 4–7 and 1–8, respectively, ‘withdrawalbehavior’ component of MOSES 13–9. According to her family and care staff opinions, her confidence improved greatly, and she became more positive and joyful. According to the interview list answered by her family members, the topics ‘repeating same conversation and questions’ and ‘scanty ideas and standardization of thinking’ showed improvement.

Case 2

One participant, whose CDR score was 1, showed some confusion during ART. She was 93 years old and had failing vision and auditory disturbance. She usually remained gentle and sedate, and did not talk voluntarily in daily life. In ART sessions, when the leader talked to her, she talked about memories of her youth in an animated and lyrical fashion, so we expected good results. However, she showed some confusion and was medicated with an antipsychotic drug. Finally her tests and scales scores decreased as follows: MMSE 20–12, ‘depressed/anxious mood’ and ‘withdrawal behavior’ component of MOSES 11–18 and 9–27, respectively. We wonder if ART contributed to this confused state. According to her family, this participant had experienced many hardships in her early life and tapping into these memories may have played a role in the development of mental confusion after ART. Such response was limited to this case.

The effect of ART on the staff

We interviewed each member of the care staff (n = 9) individually to assess whether the relationship between participants and staff had changed after ART and, if so, how. Of the nine staff members, five noticed a change. Staff opinions were as follows: ‘I began to listen to the participants’ recollections of their younger days'; ‘I learned how to communicate with the participants in a positive manner’; ‘I learned the importance of role-reversal by being taught new things by the participants’; ‘I learned the importance of expressing appreciation when tasks are completed well’; ‘saying thank you is pleasant for both the speaker and listener’ and ‘I noticed for the first time the abilities of the participants’.


To establish evidence regarding ART, we needed to perform a randomized controlled trial. However, we could not create a control group due to ethical problems. So we set a control period of 3 months before the intervention period of 3 months. Thus, we could compare the effect of ART with that of daily care. However, this long period of study caused some participants to drop out. Finally, our study relied on data from 18 participants.

In our study, three participants started medication during the intervention period (donepezil, one person; antipsychotics, two). In one participant (CDR 1, MMSE 12) taking donepezil, WMS-R scores were generally 1 or 0 and did not improve. However, she became cooperative to the care staff after the intervention period, so donepezil might have an influence on the effect of ART. In two participants who were taking antipsychotics, one (CDR 1, MMSE 20) showed noticeable side-effects during ART, including confusion, and therefore, treated with antipsychotic drug. However, her test scores steadily declined. The second participant (CDR 1, MMSE 23) showed confusion and, therefore, was treated with antipsychotic drugs. However, her test scores declined steadily. The third participant (CDR 1, MMSE 23) started antipsychotics during the intervention period due to delusion and behavior disturbance, which were present before ART and continued even after the medication. Therefore, we considered that medications minimally affected our results. Actually, statistically significant difference of the WMS-R scores did not change even after excluding these three participants.

In ART sessions, every participant actively guided the staff in the use of old-style tools rather than verbally describing their memories. The participants became obviously enthusiastic when instructing staff members how to use these tools. We noticed their eagerness, heightened concentration and improved mood. We felt the participants were very cheerful. In fact, average attendance was high. We thought these circumstances arose for two reasons. One reason may be the use of old-style tools. The nostalgia induced through using these familiar tools led to effective recall of experiences.13 Another reason may be ‘role-reversal’. The participants taught the staff how to use tools that were unfamiliar to the staff. Through role-reversal, participants gained a sense of self-worth, and a desire to live.

Of the three cognitive tests used in our study (MMSE, the Kana Pick-out test, WMS-R), only the WMS-R score was significantly improved after ART. In the Cochrane Review of reminiscence therapy (RT) for dementia, the meta-analysis of four randomized controlled trials showed significantly improved cognition.14 Moreover, other researchers showed the improvement of memory,15 attention span and MMSE.16,17 These findings support the idea that RT is associated with improvements in or maintenance of cognition. Kobayashi supposed that reminiscence, which intervenes in episodic memory, stimulates the synapses to activate the brain and has positive effects on memory, cognition and emotion.18

During ART, participants teach the staff how to use tools, with which the staff are unfamiliar. In this way, ART involves not just episodic but procedural memory and therefore ART stimulates various memory systems (semantic, procedural and episodic).13 In addition, some participants with a CDR score of 0 or 0.5 remembered the ART topics. When they came home, they smiled while they talked to their family members about the topics. Furthermore, they voluntarily prepared the clothing that was required for the following week's ART. In doing so, we assumed that participants actively used delayed memory, thus improving delayed recall. Because our study did not have enough participants or a control group, we feel that the relationship between ART and improved memory should be investigated further.

We expected that ART improved depression, anxiety, irritable behaviors and withdrawn behaviors. However, in our study, scores on MOSES, DBD and ZBI did not show significant improvement, although interviews showed improvements in subjective aspects of the participants' communication, interaction and behavior. For example, they smiled more, displayed a more vital personality and expressed good feelings. They also showed interest in their physical appearance and developed open and friendly peer relationships. These outcomes were influenced by recovering confidence through teaching and discussing the topics with staff members and their family. Preceding studies have also showed improvements in social interaction and interest.16,19,20 The Cochrane Review of RT for dementia showed significant improvement in behavioral functioning and depression.14 RT was effective especially for residents who have not adjusted their values and standards to those of the institution.21 For these residents, RT might serve as compensation, because by talking about the past, attention is given to their values and standards. These findings support our results.

As another possibility, our results may have been influenced by promoting understanding between the participants and the care staff. When the staff participating in ART sessions took care of participants, they became more receptive and provided care that encouraged demonstration of the participants' abilities. The care staff interviews also supported these observed changes. Because high quality care helped participants recover their confidence, they demonstrated expressions of joy and a more positive outlook. In addition, it might maintain the ‘self-care functioning’ area of MOSES during intervention. Braines et al. reported significant improvements in staff knowledge regarding residents on the Personal Information Questionnaire,22 supporting our results.

Based on the above, ART is very useful in improving morale, with all participants enjoying it, and everyone should be able to participate, although careful observation is necessary to prevent adverse effects, such as confusion. ART also promotes communication with peers and care givers. ART is effective for maintaining and improving emotional function and activities of daily living, as well as memory. Participants may regain their confidence and reduce the burden on their caregivers. ART contributes to providing an environment in which demented persons can laugh and be themselves. This indicates that ART fulfills brain-activating rehabilitation guidelines: enjoyable stimulation, communication and role-reversal. Thus ART, a kind of brain-activating rehabilitation, has the potential to delay the functional decline and progression of dementia.


This study was supported by the Ooido clinic, the Ooido Group Home and the Day-Service Center Ooido. The authors would like to thanks the staff, participants and their families. Also, ART would not have been possible without the donation of several instruments by friends and colleagues.