Article ID: 2022-0030-OA
The efficacy of cognitive stimulation therapy (CST) in patients with vascular cognitive impairment has not been explored, and no studies investigating CST in the convalescent rehabilitation phase have been reported. This study examined the effect of CST on the cognitive function of patients with vascular cognitive impairment. A randomized controlled, assessor-blinded, single-centered trial with two parallel groups was conducted in a convalescent rehabilitation hospital. Twenty participants were randomly allocated to CST (n=10) and control (n=10) groups. Participants in the CST group underwent two CST sessions a day, five times a week for 8 weeks, in addition to conventional rehabilitation. Participants in the control group underwent conventional rehabilitation only. The primary outcome was the Mini-Mental State Examination (MMSE) score, and the outcome between the groups was compared using a generalized linear mixed model (GLMM). The mean (standard deviation) scores of MMSE increased by 3.50 (3.08) points and 4.50 (1.61) points from baseline to the end of the study (week 8) in the CST and control groups, respectively. The GLMM showed a significant effect of TIME on MMSE (F=21.121, P<0.001), whereas no significant effect on MMSE was observed for GROUP (intervention vs. control, P=0.817) or the interaction term (TIME×GROUP, P=0.649). Although a significant improvement in cognitive function was observed in each group, no significant effect of CST was evident. This result indicates that the effect may have been masked by improvements caused by natural history or rehabilitation. Future studies with a sufficient sample size are required to confirm the findings.
Dementia is a rapidly growing global public health problem, affecting around 50 million people, with approximately 60% living in low- or middle-income countries. Every year, there are nearly 10 million new cases. The total number of people with dementia is projected to reach 82 million by 2030 and 152 million by 2050.1 The World Health Organization considers dementia to be a public health priority that must be addressed.2
Dementia has various causes. Vascular cognitive impairment (VCI) is the second most common neuropathology in dementia. It often occurs simultaneously with other age-related diseases, such as Alzheimer’s disease.3
A meta-analysis of quantitative changes in cognitive function after stroke in 28,222 patients showed greater improvement in the intervention cohort (g=0.47; confidence interval [CI], 0.37–0.58) than in the observation cohort (g=0.28; CI, 0.20–0.36), suggesting that treatment improves cognitive function.4 In particular, moderate recovery (i.e., g≥0.5) was noted in patients with first-ever stroke in the subacute phase (61–180 days after onset), suggesting that intervention in the subacute phase may have been highly significant.4
Pharmacological and non-pharmacological interventions are recommended for the treatment of dementia.5 Non-pharmacological therapies include various cognitive approaches, but the evidence of efficacy is generally weak. However, evidence suggests that cognitive stimulation therapy (CST) and physical activity positively affect patients with dementia.6
CST is a group program that involves multiple sessions of themed activities and has been found to be effective for cognitive function in patients with mild-to-moderate dementia.7 CST is also recommended as an evidence-based treatment in the UK Government NICE clinical guidelines.8 However, no randomized controlled trials (RCTs) have focused on VCI as the only target disease. Hence, the efficacy of CST for patients with VCI has not been established. Furthermore, no studies have been conducted during the subacute phase of rehabilitation, when cognitive improvement may be more likely to be achieved. Therefore, this study aimed to investigate the efficacy of CST on cognitive function in patients with VCI admitted to a convalescent rehabilitation ward.
The study protocol was developed in accordance with the Consolidated Standards of Reporting Trials (CONSORT) 2010 guidelines and checklists.9 The study was designed as a randomized controlled, observer-blinded, single-centered trial with two parallel groups.
Ethics approval and informed consentInformed consent could not be obtained from individuals who were deemed to be vulnerable. In such cases, the investigators obtained informed consent for participation in the study from a proxy. Explanations were provided to the participants and their legal representatives using the consent form. All procedures were in accordance with the ethical standards of the responsible committee at which the studies were conducted and were in accordance with the Declaration of Helsinki 1964 and later versions. The study protocol was approved by the Ethics Committee of the Tokyo Bay Rehabilitation Hospital (approval no. 124–2) and was registered prior to the study (UMIN000019234). The recruitment and follow-up period for participants was from 9 October 2015 to 30 September 2017. The recruitment ended when the target number of participants was reached.
Study settingThe study was conducted at Tokyo Bay Rehabilitation Hospital, located in the urban area of Chiba prefecture, Japan, which has convalescent Kaifukuki Rehabilitation Wards (KRWs). The KRW is a system for intensive inpatient rehabilitation for patients during their subacute phase, which is covered by governmental medical insurance, and was introduced in Japan in 2000.10 Patients with stroke are eligible for admission to the KRW within 2 months (early subacute phase)11 and can stay for up to 180 days when they have accompanying severe disabilities and cognitive disorders.
ParticipantsConsecutively screened patients with first-ever stroke who were admitted to the rehabilitation wards in the hospital from October 2015 were enrolled in the study after meeting the following criteria: 1) International Classification of Diseases, 10th edition criteria for vascular dementia,12 2) a Mini-Mental State Examination (MMSE)13 score in the range of 10 to 23, 3) underwent speech-language-hearing therapy, and 4) provided written informed consent as described above. Patients were excluded if they met any of the following criteria: 1) severe aphasia, 2) disturbance of consciousness, 3) severe visual impairment, 4) severe hearing impairment, 5) history of stroke, or 6) history of dementia.
A flowchart for participant involvement is presented in Figure 1. Twenty out of 335 consecutive stroke survivors admitted to the rehabilitation ward met the inclusion criteria. They were randomly assigned to one of the two groups: the intervention group (ten patients) and the control group (ten patients).
Flowchart of participant recruitment.
MMSE, Mini-Mental State Examination.
CST consisting of two sessions per day was conducted for the CST group five times a week for 8 weeks in addition to the conventional rehabilitation program. Each session lasted for 40 min. CST was conducted by physical therapists, occupational therapists, nurses, and care workers. The program consisted of gymnastics practiced to commands and music (gymnastic exercise), reality orientation, music therapy, recreation therapy, cognitive stimulation, reminiscence therapy, and footbath (Table 1).
| Monday | Tuesday | Wednesday | Thursday | Friday | |
|---|---|---|---|---|---|
| Morning | Gymnastics Reality orientation Cognitive stimulation |
Gymnastics Reality orientation Music therapy Reminiscence therapy |
Gymnastics Reality orientation Cognitive stimulation |
Gymnastics Reality orientation Music therapy Reminiscence therapy |
Gymnastics Reality orientation Cognitive stimulation |
| Afternoon | Reality orientation Recreation |
Reality orientation Cognitive stimulation Music therapy |
Reality orientation Footbath |
Reality orientation Recreation |
Reality orientation Reminiscence therapy Music therapy |
One facilitator and two or three assistants were needed to run each CST session. Each session was conducted with approximately ten patients, including those who were ineligible for recruitment in the present study (such as patients with hip fracture). The sessions were conducted in the plaza of the hospital cafeteria. The participants were transported by the patients’ care staff and nurses.
Gymnastics activities were performed according to “Radio Taiso” (see https://www.youtube.com/watch?v=0xfDmrcI7OI). Reality orientation was also conducted to reduce disorientation by asking participants questions about dates, seasonal events, and their hometowns. Cognitive training included a Stroop task, rock–paper–scissors, and spot the difference. For the footbath, each participant dipped their feet into a bath containing fragrant bath salts for relaxation. Recreational activities included making calendars using origami and other crafts, painting, bowling, ball games, and ring toss. Singing was performed in a choral style with lyrics and videos (from YouTube) projected on a large television. At times, handbells were used for ensembles. During the CST sessions, the staff made efforts to make the interaction engaging. Assistants aided patients as needed to encourage participation, taking into consideration the degree of dementia and disabilities, such as hemiplegia.
Participants in both the CST and control groups also participated in the inpatient rehabilitation program. In the KRW, physical, occupational, and speech-language-hearing therapies lasting for a maximum of 3 h per day were provided seven times per week. Rehabilitation consisted mainly of repetitive task-specific training aimed at reacquiring proficiency in activities of daily living. Speech-language-hearing therapy included feeding and swallowing training and cognitive function training for cognitive impairment. Cognitive training was aimed at improving each isolated cognitive domain and was intended to be generalized to non-training tasks. Task-oriented cognitive training was also provided for cognitive problems in daily living situations. The transdisciplinary team provided patients and their families with a comprehensive monthly rehabilitation plan, including information about achieved goals, planned goals, rehabilitative approaches to the planned goals, discharge planning, and social resources necessary for discharge.
RandomizationThe block randomization method was used. A randomization sequence was created using computer-generated random numbers with a 1:1 allocation using random block sizes of 10. When a participant provided consent to participate in the study, the research assistant drew a card at random on which the allocation was written. Then, the research assistant sent the participant’s name and group allocation to the nurse in charge of the CST.
OutcomesThe primary outcome was the MMSE score,13 a cognitive screening tool with high interrater reliability,14 which is recognized as a valid evaluation method for patients with stroke15,16 and has been used in many previous clinical studies on CST.7 It is designed as a 30-item questionnaire with a score range of 0–30, wherein a higher score indicates better cognitive function.
Secondary outcomes included scores for the Functional Independence Measure (FIM),17,18 Dementia Behavior Disturbance Scale (DBD),19 Neuropsychiatric Inventory Questionnaire (NPI-Q),20 Vitality Index (VI),21 and the short version of the Quality-of-Life Questionnaire for Dementia (short QOL-D).22
The FIM is a validated and reliable tool used to assess the activities of daily living in stroke survivors.23 It consists of 18 items grouped into two domains: motor and cognitive. The motor domain includes 13 items with scores ranging from 13 to 91, whereas the cognitive domain includes 5 items with scores ranging from 5 to 35. The total score (sum of the motor and cognitive scores) may range from 18 to 126, with higher scores indicating higher degrees of independence.
The DBD and NPI-Q were selected to evaluate behavioral and psychological symptoms of dementia (BPSD). The DBD is a scale consisting of 28 questions about BPSD for caregivers where each item is rated on a 5-point scale from “never” (0 points) to “always” (4 points) to obtain a total score (maximum, 112 points). The NPI-Q is a 12-item questionnaire-based neuropsychological test that quantifies the frequency and severity of BPSD and caregiver burden in patients with dementia. The burden scale is rated by the caregivers and has six levels ranging from “none at all” (0) to “extremely burdensome and cannot be handled” (5). The total score ranges from 0 to 60, with higher scores reflecting a higher burden of care. The reliability and validity of the DBD and NPI-Q for the assessment of BPSD for patients with dementia have been described elsewhere.19,20
VI was chosen to evaluate vitality because previous studies have reported that VCI is more likely to reduce vitality than Alzheimer’s disease. The VI consists of five items: waking pattern, communication, feeding, toilet visits, and rehabilitation and other activities. Each item is scored based on a 3-point scale. The total score is 0–10 points, with a higher score indicating more vitality. The validity and reliability of the VI have been established for patients with dementia.21
The short QOL-D is a scale for evaluating the QOL of patients with dementia. It consists of nine items: six positive symptoms and three negative symptoms. Each item is rated on a 4-point scale from 1 (not observed) to 4 (often observed), and the total score is calculated. A higher total score reflects a higher QOL. The reliability and validity of the short QOL-D as an assessment of the QOL for patients with dementia have been reported elsewhere.22
All outcomes were measured at baseline, week 4, and week 8. The MMSE score was evaluated by a single physician who was not aware of the study allocation. FIM, DBD, NPI-Q, short QOL-D, and VI were rated by the nurses in charge. Nurses observed participants’ daily lives and rated the scores. The Alzheimer’s Disease Assessment Scale-cognitive subscale, which was selected in the preliminary protocol, was excluded from this evaluation because this subscale was not feasible for use in several patients who had paralyzed hands.
BlindingAssessments of the primary outcome were conducted by an assessor blinded to the treatment allocation. Because of the nature of the intervention, neither participants nor staff could be blinded to the allocation. Assessments regarding secondary outcomes were not blinded, because these evaluations needed to be observed directly in participants’ daily lives.
Sample sizeThe sample size and power calculations were estimated for the change in the MMSE score. The sample size calculation was performed using the G*Power 3 program.24 Using an effect size calculated by the t-test, a moderate effect size (d=0.5)25 based on a meta-analysis,7 an alpha of 5%, and a power of 80%, we obtained a required sample size of 34 patients in each group. Given that some patients were likely to be lost to follow-up, the total target sample size was 70 patients.
However, the number of recruited participants did not meet our requirements. Therefore, we revised the number of participants to the highest possible number that could be recruited by October 2017, 2 years after the start of the study. There were two main reasons that the number of participants recruited was lower than that initially proposed. First, one of the three wards in which we had planned to recruit participants could not be used because of staffing issues, and, second, continuation of the study became challenging because CST was a burden on managing operations in the wards.
Statistical methodsAll data were analyzed on an intention-to-treat basis. The characteristics of the participants in the intervention group were compared with those of the control group. Comparisons between groups were performed using the chi-squared test, t-test, and Mann–Whitney U test, depending on the variable type. We used generalized linear mixed models (GLMM) for primary and secondary outcomes and TIME and GROUP (CST group vs. control group) and their interaction (TIME×GROUP) as covariates. In addition, the effect size (Cohen’s d) of the primary outcome was calculated. Effect size was interpreted as small (d=0.2), medium (0.5), or large (0.8).25 All statistical analyses were performed using SPSS version 27.0 (IBM, Tokyo, Japan). A P value of ≤0.05 was considered statistically significant.
Two of the ten patients in the intervention group were discharged before week 8; therefore, the final evaluation was not conducted for these patients. In addition, a record of MMSE T1 in one patient in the control group was accidentally lost (Fig. 1). There was no statistically significant difference in patient characteristics at baseline between the control and CST groups (Table 2).
| Characteristic | Total (n=20) |
Intervention group (n=10) |
Control group (n=10) |
P value | |
| Age, years | 75.8 (9.2) | 73.5 (9.2) | 78.0 (9.1) | 0.929 | |
| Sex (male/female) | 6/14 | 3/7 | 3/7 | 0.999 | |
| Type of stroke (infarction/hemorrhage/subarachnoid hemorrhage) | 13/5/2 | 7/3/0 | 6/2/2 | 0.699 | |
| Education history, years | 12.4 (1.9) | 12.4 (1.7) | 12.4 (2.1) | 0.761 | |
| Time from stroke onset to admission, days | 39.25 (10.6) | 38.2 (11.7) | 41.1 (11.1) | 0.603 | |
| MMSE | 19.5 [17.0–22.0] | 21.5 [17.0–22.8] | 19 [18.3–20.0] | 0.631 | |
| DBD | 5 [1.8–9.5] | 2.0 [0–7.3] | 7.0 [4.3–13.3] | 0.075 | |
| NPI-Q care burden | 2 [0–4.5] | 0.5 [0–5.8] | 2.5 [0.5–4.0] | 0.631 | |
| NPI-Q severity | 2 [0–3.3] | 0.5 [0–3.5] | 2 [0.5–3.0] | 0.436 | |
| Vitality index | 7.5 [6.0–9.0] | 9 [8.0–9.8] | 7 [6.0–7.0] | 0.035 | |
| Short QOL-D positive | 15 [10.3–17.3] | 15.5 [13.3–18.5] | 14.5 [8.0–15.0] | 0.190 | |
| Short QOL-D negative | 12 [11.8–12.0] | 12 [11.3–12.0] | 12 [12.0–12.0] | 0.684 | |
| Short QOL-D total | 25.5 [22.3–29.3] | 26 [24.0–30.5] | 25.5 [20.0–26.8] | 0.280 | |
| FIM | Motor items | 32.5 [18.8–48.3] | 36 [29.5–43.5] | 28 [18.3–52.5] | 0.971 |
| Cognitive items | 22 [17.0–26.0] | 22.5 [18.0–25.0] | 21.5 [17.5–27.5] | 0.853 | |
| Total | 58.5 [41.0–67.5] | 57.5 [51.0–65.3] | 58.5 [36.5–74.5] | 0.999 | |
Data are presented as number, mean (standard deviation), or median [interquartile range].
MMSE, Mini-Mental State Examination; DBD, Dementia Behavior Disturbance scale; NPI-Q, Neuropsychiatric Inventory Questionnaire; Short QOL-D, Short version of Quality-of-Life Questionnaire for Dementia; FIM, Functional Independence Measure.
The results of the GLMM analysis for the primary outcome showed a significant effect of TIME on the MMSE score (F=21.121, P<0.001; Fig. 2). The effect of GROUP (intervention vs. control) on the MMSE score was not significant (F=0.055, P=0.817). The effect size of the intervention (Cohen’s d), judged by the improvement in the MMSE score, was −0.74 (CI, −1.00 to 0.86). Furthermore, the interaction term (TIME ×GROUP) was not significant (F=0.438, P=0.649). The mean (standard deviation) improvement in MMSE scores from baseline to the end of the study (week 8) was 3.50 (3.08) points and 4.50 (1.61) points in the CST and control groups, respectively. The effect of TIME was d=0.888 (CI, −0.08 to 1.785) in the CST group and d=1.516 (CI, 0.499 to 2.532) in the control group.
Change in participant MMSE score during the study.
MMSE, Mini-Mental State Examination; CST, cognitive stimulation therapy.
For the secondary outcomes, the GLMM revealed a significant TIME effect for FIM motor items (F=36.441, P<0.001), FIM cognitive items (F=5.029, P=0.012), FIM total score (F=33.690, P<0.001), and VI (F=3.981, P=0.028). However, there was no significant effect for GROUP or TIME×GROUP. There was no significant effect of TIME, GROUP, or TIME×GROUP for the other secondary outcomes (Table 3). No study-related adverse events were reported throughout the study period.
| Item | Intervention | Time | Intervention×Time | |||||
|---|---|---|---|---|---|---|---|---|
| F | P value | F | P value | F | P value | |||
| MMSE | 0.055 | 0.817 | 21.21 | <0.001 | 0.438 | 0.649 | ||
| FIM motor items | 0.027 | 0.870 | 36.441 | <0.001 | 0.253 | 0.778 | ||
| FIM cognitive items | 0.000 | 0.983 | 5.029 | 0.012 | 0.332 | 0.720 | ||
| FIM total | 0.023 | 0.879 | 33.690 | <0.001 | 0.388 | 0.681 | ||
| Vitality index | 0.988 | 0.325 | 3.981 | 0.028 | 1.969 | 0.156 | ||
| DBD | 0.245 | 0.623 | 0.256 | 0.776 | 0.599 | 0.555 | ||
| NPI-Q care burden | 0.001 | 0.971 | 0.166 | 0.848 | 0.392 | 0.679 | ||
| NPI severity | 0.123 | 0.727 | 0.458 | 0.637 | 0.460 | 0.635 | ||
| Short QOL-D positive | 0.543 | 0.464 | 2.781 | 0.076 | 0.111 | 0.895 | ||
| Short QOL-D negative | 0.126 | 0.724 | 0.099 | 0.906 | 0.203 | 0.817 | ||
| Short QOL-D total | 0.457 | 0.502 | 2.091 | 0.139 | 0.245 | 0.784 | ||
MMSE, Mini-Mental State Examination; DBD, Dementia Behavior Disturbance scale; NPI-Q, Neuropsychiatric Inventory Questionnaire; Short QOL-D, short version of Quality-of-Life Questionnaire for Dementia; FIM, Functional Independence Measure.
We investigated whether the addition of CST to conventional rehabilitation in the subacute phase improves cognitive function and BPSD in patients with VCI. Although there was no significant difference in primary or secondary outcomes between the two groups, both showed statistically significant improvements in cognitive function compared to the baseline and at the end of the study when examined using the MMSE score.
In a previous study conducted in a convalescent rehabilitation hospital, patients with subacute stroke who showed cognitive impairment, defined as an MMSE score of 23 points or lower at admission, eventually showed improved cognitive function by 3.4 points in the mean MMSE score during a stay of approximately 3 months in a rehabilitation ward.26 Although the intervention period of the present study was 2 months, which was shorter than that of the previous study, we observed the same level of improvement in the MMSE score.
The cognitive improvement observed in both groups could be attributed to spontaneous recovery and rehabilitative intervention. Cognitive impairment that appears in the acute phase of stroke often improves within a few months to a few years after stroke onset,27,28 and this improvement may, at least to some extent, be explained by spontaneous recovery.
Previous studies have also investigated the effect of rehabilitation on cognitive function. In the present study, the participants in both the CST and control groups underwent 3 h of rehabilitation per day. Therefore, rehabilitation may have had a positive effect on their cognitive function. A systematic review suggested that physical activity has a positive effect on improving global cognitive ability and potential benefits for memory, attention, and the visuospatial domain of cognition in stroke survivors.29 Some mechanisms have been suggested to explain the effects of exercise on cognition after stroke, including an increase in cerebral blood volume, increased expression of growth factors, such as brain-derived neurotrophic factor, and a positive effect on depressive symptoms.29
In the present study, participants in both groups showed a significant effect of time in improving MMSE scores. Compared with the results of a meta-analysis of CST,7 which reported a moderate effect size (standardized mean difference of 0.41; 95% CI, 0.25–0.57), our study showed large effect sizes. The change in cognitive function over time might be sufficiently evident to mask the effects of CST.
Regarding feasibility in terms of cost effectiveness, each session of CST required two or three assistants and one facilitator. A maximum number of ten people could be treated in each session. Therefore, treatment was considered reasonable in terms of cost compared to one-on-one rehabilitation. However, the need to gather several participants in one place at a specific time was sometimes challenging for the staff. In addition, having multiple patients and staff gather together to perform the treatment could increase the risk of infection.
Study limitationsThis study had limitations. First, the study was conducted in a single facility consisting of a few participants, and patients with aphasia were excluded. Second, a longer follow-up conducted across multiple facilities is required to confirm our findings. Finally, participants in the control group underwent rehabilitation as a basic treatment, making it impossible to consider the control group as a non-intervention group.
The generalizability of this study is limited by the characteristics of the participants. This study included older patients with VCI between ages of 65 and 85 years who were in the subacute phase of their disease with moderate-to-mild cognitive impairment. Therefore, caution should be taken when applying these results to younger patients with VCI, patients with other stages of the disease, or patients with VCI with other types of dementia or severe cognitive impairment. Although the sample size was smaller than initially estimated, the study was a well-designed RCT. In the absence of prior research on this issue, this preliminary study addresses an area of unmet need in this field. In conclusion, we found no significant effect of CST intervention in patients with VCI admitted to rehabilitation wards. However, significant improvement in cognitive function was observed in the control and CST groups. This indicates that the effect, if any, was negligible and may have been masked by the overall improvement. Future studies with a sufficient sample size are required to confirm the results of this study.
The authors gratefully acknowledge the contributions of the nurses and caregivers of the Tokyo Bay Rehabilitation Hospital and those of the therapists involved in this study.
The authors declare no conflict of interest.
