Abstract
Objective: The purpose of this study is to investigate the effect of an intervention combining exercise and cognitive activity on cognitive function in healthy older adults. Methods: This pilot randomized controlled trial recruited 33 eligible, healthy communitydwelling older adults (mean age, 77.1 years old; women, 51.5%), who were divided into intervention and waitlist control groups. The intervention group was engaged weekly in a group activity comprising exercise and discussions of homework, which included reading aloud, simple arithmetic, and simple activities, like spotting differences, for cognitive stimulation. They were also required to complete cognitive activity homework twice a week. The waitlist control group received no intervention. The main outcomes were cognitive function assessed using the Mini-Mental State Examination, delayed recall score on the Logical Memory IIA of the Wechsler Memory Scale Revised, Trail Making Test, and digit symbol substitution test. Results: According to the results, Mini-Mental State Examination scores were maintained in the intervention group but declined in the control group [Mean change in outcomes in control group (95% confidence interval): -1.68 (-2.89 to -0.48)]. Additional mean change in outcomes in intervention group were found [1.68 (0.02 to 3.35)]. Conclusions: Interventions combining exercise and cognitive activity can be helpful for preserving cognitive function in healthy older adults.
Keywords: Cognitive activity, Cognitive functon, Exercise, Older adults
The prevalence of Alzheimer's disease has significantly increased with the expansion of the older adult population, and as per estimates, the current 33.9 million cases worldwide will triple by 20501,2). In particular, the already high prevalence of dementia in Asian countries will only increase further2). Further, the incidence of dementia has been reported to be decreasing in Western countries over time; meanwhile, it has been increasing in Japan3-5). This implies that the prevention of dementia is essential in Japan. Prevention of dementia is still a critical unmet medical need, and many researches failed to develop disease-modifying therapies for Alzheimer's disease6).
Exercise has received much attention as a possible preventive strategy for dementia. This is unsurprising, given its positive effects on cardiovascular health, diabetes, and mental health, which are reported to be modifiable risk factors for dementia7). Sufficient amounts of exercise are reported to be equivalent to polypills and have a low risk of adverse effects8). A meta-analysis published in 2019 revealed a small effect of exercise on cognitive function in healthy older adults including aerobic and resistance training (Hedge's g = 0.31); however, this study entailed a publication bias9). Recently, it is pointed out that a combination of exercise and cognitive activity may be more efficacious for the improvement of cognitive function in healthy older adults. A meta-analysis including non-randomized control studies suggested that the combination training improves cognitive function more than physical exercise alone and control in healthy older adults10). However, there is a need for more randomized controlled trials investigating this combination effect, especially in Asian countries, where there have so far been only two studies of this type in healthy older adults11,12).
The purpose of this study is to investigate the effect of an intervention combining exercise and cognitive activity on cognitive function in community-dwelling older adults.
Methods
Design, setting, and participants
This randomized controlled trial was conducted between June 2017 and March 2018. Community-dwelling older adults were recruited from Tomogaoka, Northern Suma area, in Kobe city. Posters describing the study were displayed at community centers and invitation letters were sent to recruit participants. Thirty-four older adults agreed to participate in this trial. The inclusion criteria were older adults who: 1) were aged 70 years or above and 2) could walk independently 3) had no cognitive decline (Mini-Mental State Examination (MMSE) score > 24). The exclusion criteria were chronic diseases that hindered physical activity. After excluding one respondent owing to cognitive decline, 33 participants were included in the study (mean age, 77.1 years old; women, 51.5%). Respondents received no incentive, monetary or otherwise, for participation in this trial. They were divided into two groups: intervention group (n=17) and wait-list control (n=16). Baseline assessment except for cognitive function was conducted on June 6 and 7, 2017, and baseline cognitive assessment was conducted between June 8 and 14. The intervention duration was from June 28 to September 20. Immediate follow-up assessment except for cognitive function was conducted on September 21 and 22, and follow-up cognitive function assessment was conducted between September 25 and 29. Additional follow-up assessment was conducted between February 26 and March 23, 2018, which was determined randomly. No intervention was conducted on the intervention group between immediate follow-up (September 2017) and additional refollow-up (February and March 2018).
The Research Ethics Committee of Kobe University Graduate School of Health Sciences approved the study (authorization number 592), and written informed consent was obtained from all participants. This randomized controlled trial was conducted in accordance with CONSORT criteria (Table S1) (registration number: UMIN000027740; https://upload.umin.ac.jp/cgi-open-bin/ctr_e/ctr_view.cgi?recptno=R000031784).
Intervention group
The intervention comprised two components: group activity and homework. Every group activity was conducted after a physical and vitals check.
Group activity, which included exercise and discussing homework, was conducted once weekly for three months. During the abovementioned intervention period, we skipped one week because of a traditional Japanese holiday. In total, group activity was conducted 12 times. Trained physiotherapists supervised this program. Each class contained 5-15 participants. The first group activity session included an orientation, where the supervisor provided an explanation of the importance of exercise and cognitive activity for preserving cognitive function. Here, participants were informed about the homework and physical activity requirements. In the second and third sessions, cognitive activity was conducted to teach how to do homework more detail after checking the answer of homework (35 min), stretching and muscle strength training (30 min), and dual-task aerobic exercise (20 min). Sessions 4 to 12, which were 90 minutes long, involved discussing homework (25 min), stretching and muscle strength training (30 min), and dual-task aerobic exercise (30 min). Participants were given a break every 30 minutes. In the dual-task aerobic exercise, participants performed physical activity (a step exercise using a platform) while making calculations, taking verbal fluency tests (categories and letters), or playing a Japanese word chain game called “Shiritori,” wherein the person in question has to name a word beginning with the final letter of the previous word. The cognitive activity contents were based on previous studies13,14). The participants received two sets of homework after group activity: reading aloud and simple arithmetic13). Additionally, the homework sheets included tasks that involved spotting differences, searching task, and riddles14). A previous study used a computer-based cognitive activity tool, “Atama-no-dojo,” and we used its paper-and-pencil version14). The supervisor asked the participants to do the homework twice a week and bring it to the following week's group activity.
Waitlist control group
The waitlist control group received no intervention in the trial period. After three months, this group received the identical intervention.
Randomization and blinding
Simple randomization was conducted. Participants were randomly divided into two groups with 1:1 allocation as per a computer-generated randomization schedule. The first author created two categories (0 or 1) based on random numbers. To ensure concealed allocation, the researcher conducting the randomization was blinded to which number indicated the intervention group and which one indicated the waitlist control group. No other blinding, such as of subjects, therapists, and assessor, was conducted.
Outcome
Outcome was measured thrice in the intervention group―at baseline, follow-up three months after baseline assessment, and five months after the conclusion of the intervention― and twice in the waitlist control group―at baseline and follow-up.
Cognitive function was evaluated using the MMSE15), delayed recall score on the Logical Memory IIA (LM IIA) of the Wechsler Memory Scale Revised, Trail Making Test16), and digit symbol substitution test from the Wechsler Adult Intelligence Scale-Revised17).
Other measurements
The other data, collected using a self-reported questionnaire, included age, sex, smoking status (never smoker, ex-smoker, current smoker), alcohol consumption (never, rarely, sometimes, every day), number of medications, medication type (tranquilizer, analgesic, depressor, antidepressant), education (< 10 years, 10-12 years, > 12 years), comorbidities (hypertension, stroke, diabetes, Parkinson's disease, hyperlipidemia, knee osteoarthritis, disc herniation, compression fracture, spinal canal stenosis, rheumatoid arthritis, cancer, respiratory illness, heart disease), household income, and depressive symptoms. Depressive symptoms were assessed using the Geriatric Depression Scale-15 (GDS)18), a 15-item inventory in a yes/no format. Depressive symptoms were defined as a GDS score ≥ 619). Body mass index was calculated as weight (kilograms) divided by height (meters) squared. Systolic and diastolic blood pressure, measured once, were assessed following standard methods. Total cholesterol, high-density lipoprotein, triglycerides, blood sugar, and hemoglobin A 1c were assessed using blood samples. Gait speed was assessed using a stopwatch; the participants were asked to walk 6.4 m (divided into two 2.0 m zones at each end and a 2.4 m middle zone) at their usual pace, and we measured the time required (in seconds) to complete the 2.4 m middle zone over two trials to calculate the mean gait speed (m/s). Mobility was assessed with a Timed Up and Go Test using a stopwatch. The participants were asked to stand up from a standard armchair, walk 3.0 m, then turn around, walk back to the chair, and sit down again. We measured the time required to complete the test. Handgrip strength was measured as muscle strength using a handgrip dynamometer (T.K.K. 5401; Takei Scientific Instruments, Niigata, Japan). One trial for each hand was performed, and the result from the stronger hand was used in the analysis.
Statistical analysis
The groups' baseline demographic and clinical characteristics were compared using the Fisher's exact test for categorical variables and unpaired t-test for continuous variables, as appropriate.
Intention to treat analyses were conducted using multivariate imputation by chained equations. Mixed effects models were developed to investigate the effect of the intervention program on cognitive function. The independent variables were allocation (categorical variable: intervention group and control group), time (categorical variable: baseline and follow-up), and the time by allocation interaction. Individual participants' intercepts for cognitive function were allowed to vary as random effects. Fifty imputed data sets were generated, and Rubin's rules were used to combine the results. Comparing with multivariate imputation analysis, complete case analysis was also applied.
Another mixed effects model was developed solely to investigate changes in cognitive function in the intervention group. The independent variable was time (categorical variable: baseline, follow-up, and five months after intervention). The random effects of the intercept were also entered. Both multivariate imputation analysis and complete case analysis were applied.
All linear mixed effects models were calculated using the lmer function of lme4 package20), and multivariate imputation by chained equations were conducted using mice package21). Statistical significance was set at P < 0.05, and all analyses were conducted using R (3.6.0).
Results
Demographic data
The baseline demographic and clinical characteristics of the two groups are depicted in Table 1. At baseline, there was no difference in cognitive function between the groups. The median (interquartile range) of the number of group activity participation is 10 (9-12). No adverse events were observed. Figure 1 depicts the flow of this trial. Two participants dropped out of this trial―one each in the intervention and control groups. The follow-up rate was 94%.
Table 1.
Comparison of Baseline Demographic and Clinical Characteristics Between the Groups
| Control (n = 16) | Intervention (n = 17) | p | |
|---|---|---|---|
| Abbreviations: SD, standard deviation; NA, not available; BMI, body mass index; HDL, high-density lipoprotein; HbA1c, hemoglobin A1c; MMSE, Mini-Mental State Examination; TMT, Trail Making Test; DSST, digit symbol substitution test; WMS-R, Logical Memory subtest of the Wechsler Memory Scale Revised | |||
| Age, years, mean (SD) | 76.81 (3.60) | 77.35 (3.94) | 0.684 |
| Women, n (%) | 10 (62.5) | 7 (41.2) | 0.303 |
| BMI, kg/m2, mean (SD) | 23.37 (2.06) | 23.78 (3.95) | 0.725 |
| Missing, n (%) | 1 (6.3) | 0 (0.0) | |
| Systolic blood pressure, mmHg, mean (SD) | 139.47 (11.27) | 141.47 (23.18) | 0.763 |
| Missing, n (%) | 1 (6.3) | 0 (0.0) | |
| Diastolic blood pressure, mmHg, mean (SD) | 80.13 (7.13) | 81.71 (13.76) | 0.694 |
| Missing | 1 (6.3) | 0 (0.0) | |
| Smoking status, n (%) | 0.141 | ||
| Never smoker | 12 (75.0) | 9 (52.9) | |
| Ex-smoker | 3 (18.8) | 8 (47.1) | |
| Current smoker | 1 (6.2) | 0 (0.0) | |
| Alcohol consumption, n (%) | 1 | ||
| No | 4 (25.0) | 3 (17.6) | |
| Rarely | 3 (18.8) | 4 (23.5) | |
| Sometimes | 2 (12.5) | 3 (17.6) | |
| Every day | 7 (43.8) | 7 (41.2) | |
| Education, years, n (%) | 0.534 | ||
| < 10 years | 3 (18.8) | 4 (23.5) | |
| 10-12 years | 7 (43.8) | 10 (58.8) | |
| > 12 years | 6 (37.5) | 3 (17.6) | |
| Household income, n (%) | 0.518 | ||
| < 2 million | 3 (18.8) | 4 (23.5) | |
| 2-4 million | 9 (56.2) | 10 (58.8) | |
| 4-6 million | 1 (6.2) | 3 (17.6) | |
| 6-8 million | 2 (12.5) | 0 (0.0) | |
| ≥ 8 million | 0 (0.0) | 0 (0.0) | |
| Missing | 1 (6.2) | 0 (0.0) | |
| Comorbidities | |||
| Hypertension, n (%) | 6 (37.5) | 8 (47.1) | 0.728 |
| Stroke, n (%) | 0 (0.0) | 0 (0.0) | NA |
| Diabetes, n (%) | 1 (6.2) | 2 (11.8) | 1 |
| Parkinson's disease, n (%) | 0 (0.0) | 0 (0.0) | NA |
| Hyperlipidemia, n (%) | 3 (18.8) | 2 (11.8) | 0.656 |
| Knee osteoarthritis, n (%) | 3 (18.8) | 3 (17.6) | 1 |
| Cancer, n (%) | 2 (12.5) | 2 (11.8) | 1 |
| Respiratory illness, n (%) | 0 (0.0) | 2 (11.8) | 0.485 |
| Heart disease, n (%) | 2 (12.5) | 3 (17.6) | 1 |
| Medication type | |||
| Tranquilizer, n (%) | 0 (0.0) | 0 (0.0) | 1 |
| Analgesic, n (%) | 1 (6.2) | 4 (23.5) | 0.335 |
| Depressor, n (%) | 7 (43.8) | 11 (64.7) | 0.303 |
| Antidepressant, n (%) | 0 (0.0) | 0 (0.0) | NA |
| Blood samples | |||
| Total cholesterol, mg/dl, mean (SD) | 219.62 (29.75) | 198.24 (33.03) | 0.06 |
| HDL, mg/dl, mean (SD) | 71.75 (20.81) | 63.59 (18.23) | 0.239 |
| Triglyceride mg/dl, mean (SD) | 168.50 (111.87) | 131.06 (61.87) | 0.239 |
| Blood sugar mg/dl, mean (SD) | 106.94 (32.91) | 94.35 (10.41) | 0.144 |
| HbA1c, %, mean (SD) | 5.81 (0.55) | 5.73 (0.41) | 0.622 |
| Physical measurements | |||
| Gait speed, s/m, mean (SD) | 1.23 (0.19) | 1.19 (0.19) | 0.613 |
| Grip strength, kg, mean (SD) | 26.50 (9.64) | 27.12 (6.06) | 0.825 |
| Timed Up-and-Go Test, sec, mean (SD) | 8.97 (1.88) | 8.99 (1.44) | 0.969 |
| Depressive symptoms, n (%) | 0 (0.0) | 4 (23.5) | 0.103 |
| Missing | 1 (6.2) | 0 (0.0) | |
| Cognitive assessments | |||
| MMSE, point, mean (SD) | 28.50 (1.67) | 27.47 (1.91) | 0.11 |
| TMT-A, sec, mean (SD) | 44.49 (11.72) | 45.43 (9.75) | 0.803 |
| TMT-B, sec, mean (SD) | 117.45 (46.52) | 117.90 (59.41) | 0.981 |
| Missing | 0 (0.0) | 1 (5.9) | |
| DSST, count, mean (SD) | 45.19 (10.37) | 43.94 (10.76) | 0.737 |
| WMS-R short-term, count, mean (SD) | 10.31 (4.56) | 7.88 (3.87) | 0.108 |
| WMS-R long-term, count, mean (SD) | 7.69 (4.05) | 7.35 (4.29) | 0.819 |
Fig. 1.
Study Flow.
Effect of intervention on cognitive function
Table 2 presents the results of the mixed effects model with multivariate imputation by chained equations. The control group's MMSE score fell by 1.68 points, while in the intervention group, the score was maintained (0.00 points). No significant changes were observed in the other outcome models. Table S2 also showed similar results using complete case analysis. Figure S1 depicts changes in cognitive function.
Table 2.
Effect of Exercise and Cognitive Activity on Cognitive Function: Mixed Effects Models
| Mean (Standard deviation) | Results of mixed effects model | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Control group | Intervention group | Difference baseline value for intervention group (allocation) | Mean decline in outcomes in control group (time) | Additional decline in outcomes in intervention group (interaction) | ||||||||||
| Baseline | Follow-up | Baseline | Follow-up | Estimate (95% CI) | p | Estimate (95% CI) | p | Estimate (95% CI) | p | |||||
| Abbreviations: 95% CI, 95% confidence interval; MMSE, Mini-Mental State Examination, TMT, Trail Making Test; DSST, Digit Symbol Substitution Test; WMS-R, Logical Memory Subtest of the Wechsler Memory Scale Revised. The mixed effects models were built using multiple imputation and means (standard deviations) were calculated using complete case data. There were three missing value in TMT-B at the follow-up assessment (Intervention Group: 2, Control Goup: 1), other outcome variables at follow-up had two missing value (Intervention Group: 1, Control Group: 1). | ||||||||||||||
| MMSE | 28.47 (1.73) | 26.73 (2.15) | 27.38 (1.93) | 27.44 (2.00) | -1.03 (-2.34 to 0.28) | 0.122 | -1.68 (-2.89 to -0.48) | 0.007 | 1.68 (0.02 to 3.35) | 0.048 | ||||
| TMT-A | 44 (11.96) | 41.06 (12.36) | 45.35 (10.06) | 42.28 (9.22) | 0.94 (-6.5 to 8.38) | 0.801 | -2.94 (-7.65 to 1.77) | 0.216 | -0.12 (-6.68 to 6.44) | 0.971 | ||||
| TMT-B | 118.97 (47.74) | 106.71 (44.03) | 110.68 (66.18) | 103.65 (32.72) | 1.14 (-30.71 to 32.98) | 0.943 | -12.35 (-32.63 to 7.92) | 0.228 | -2.62 (-30.87 to 25.63) | 0.854 | ||||
| DSST | 44.73 (10.57) | 44.67 (13.24) | 44.94 (10.27) | 47.69 (8.14) | -1.25 (-8.81 to 6.32) | 0.743 | 0.02 (-3.14 to 3.19) | 0.988 | 2.65 (-1.77 to 7.06) | 0.235 | ||||
| WMS-R long-term | 7.87 (4.12) | 9.47 (5.32) | 6.81 (3.78) | 8.31 (3.53) | -0.34 (-3.42 to 2.75) | 0.829 | 1.6 (-0.26 to 3.45) | 0.09 | -0.09 (-2.68 to 2.49) | 0.942 | ||||
| WMS-R short-term | 10.6 (4.56) | 11.6 (5.03) | 7.62 (3.84) | 10.81 (3.82) | -2.43 (-5.43 to 0.57) | 0.11 | 1.07 (-0.65 to 2.79) | 0.218 | 2.06 (-0.34 to 4.45) | 0.092 | ||||
Change in cognitive function in intervention group
Table 3 presents the results of the mixed effects model with multivariate imputation by chained equations to investigate changes in cognitive function in the intervention group. A significant improvement in long-term memory was observed in the intervention group at five or six months after the conclusion of the intervention. A significant improvement in short-term memory was observed in the intervention group at follow-up and five or six months after the conclusion of the intervention. Table S3 presents the results of the mixed effects model with complete case to investigate changes in cognitive function in the intervention group. A significant improvement in long-term and short-term memory was observed in the intervention group at follow-up and five or six months after the conclusion of the intervention.
Table 3.
Change in Cognitive Function in Intervention Group: Mixed Effects Models
| Mean (standard deviation) | Results of mixed effects model | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Difference in predicted value between baseline and follow-up | Difference in predicted value between baseline and 5 or 6 months after intervention | ||||||||
| Baseline | Follow- up | 5 or 6 months after intervention | Estimate (95% CI) | p | Estimate (95% CI) | p | |||
| Abbreviations: 95% CI, 95% confidence interval; MMSE, Mini-Mental State Examination; TMT, Trail Making Test; DSST, digit symbol substitution test; WMS-R, logical memory of the Wechsler Memory Scale Revised. The mixed effects models were built using multiple imputation and means (standard deviations) were calculated using complete case data. There were two missing value in the TMT-B at follow-up assessment, another outcome variable at follow-up has one missing value. | |||||||||
| MMSE | 27.38 (1.93) | 27.44 (2.00) | 28.27 (1.53) | 0.03 (-1.15 to 1.21) | 0.96 | 0.84 (-0.36 to 2.04) | 0.163 | ||
| TMT-A | 45.35 (10.06) | 42.28 (9.22) | 39.85 (12.81) | -2.79 (-8.39 to 2.80) | 0.319 | -4.49 (-10.11 to 1.13) | 0.114 | ||
| TMT-B | 110.68 (66.18) | 103.65 (32.72) | 103.25 (21.71) | -12.96 (-35.03 to 9.12) | 0.243 | -15.12 (-37.1 to 6.86) | 0.172 | ||
| DSST | 44.94 (10.27) | 47.69 (8.14) | 45.47 (7.74) | 2.98 (-0.03 to 5.99) | 0.053 | 0.73 (-2.38 to 3.85) | 0.637 | ||
| WMS-R long-term | 6.81 (3.78) | 8.31 (3.53) | 9.60 (3.40) | 1.20 (-0.39 to 2.78) | 0.134 | 2.47 (0.85 to 4.08) | 0.004 | ||
| WMS-R short-term | 7.62 (3.84) | 10.81 (3.82) | 10.73 (3.37) | 2.94 (1.03 to 4.86) | 0.003 | 3.03 (1.08 to 4.98) | 0.003 | ||
Discussion
Our study investigated the effect of a combination of exercise and cognitive activity on cognitive function in community-dwelling older adults. As per the results, the decline in MMSE scores in the intervention group was lower than in the control group. Improvements in long-term and short-term memory were observed in the intervention group at follow-up and five months after the intervention as compared to the baseline.
Some randomized controlled trials have demonstrated the effect of the combination of exercise and cognitive activity on cognitive function10). Two randomized controlled trials in Japan reported higher cognitive function improvement in dual-task exercise groups than control groups11,12). The intervention contents in the present trial included a combination of cognitive activity and physical exercise, which helped in the maintenance of cognitive function as assessed using the MMSE. In previous combination studies of exercise and cognitive activity, cognitive activity was conducted either with or without exercise. In the present study, cognitive activity was conducted both with and without exercise, since the dual-task approach may be more efficacious than these activities conducted separately and cognitive activity homework is easy to implement10). The results of our study are in line with those of previous randomized controlled trials, aiding in the accumulation of insight into the combination effect on cognitive function.
Our result also suggested improvements in long-term and short-term memory in the intervention group at follow-up or five months after the intervention as compared to the baseline. Our intervention included memory training in dual-task exercises and cognitive activity homework, which might lead to the improvement of memory ability. However, this improvement was only observed in the pre-post comparison. Therefore, our study could not conclude that this intervention caused memory improvement. Further studies should be conducted to clarify the long-term effects of the combination intervention by comparison to the control group.
In this trial, group activity was conducted once a week. In addition, participants had to do homework related to cognitive activity twice a week; thus, participants received intervention thrice a week. This increased the efficacy of our intervention without any additional effort on the part of the researchers. A previous meta-analysis reported that interventions with a frequency of five times or more per week have a greater effect on cognitive function than interventions with a frequency of four times or under per week10). Therefore, there is a need for studies that conduct more frequent interventions using homework.
Despite its contributions, this study has several limitations. First, the small sample size led to low statistical power. Estimated values in this sample may also be far from the true values of population. Our results suggested that MMSE decline in the control group was about 1.7, which is relatively larger than that of previous studies11). Attention should be paid to this point. However, it is noteworthy that through the insight it has accumulated, our study can be included in meta-analyses on this topic. Second, although it was not possible in this scenario, the lack of blinding of participants, supervisors, and assessors may have led to some bias. Third, the adherence to homework completion could not be measured as we could not observe it.
Conclusion
In conclusion, our results revealed that the decline in MMSE scores in the intervention group was lower than in the control group. Combining exercise and cognitive activity can be helpful for preserving cognitive function in healthy older adults.
Conflict of Interest
The author reports no conflicts of interest in this work.
Appendix
Supplementary material (Appendix):
Acknowledgments
We would like to thank Mr. Yamato Tsuboi for supervising the group activity and providing meaningful help. The work was funded by Total Brain Care Co., Ltd and Grant-in-Aid for JSPS Fellows (17J05200).
References
- 1.Barnes DE and Yaffe K: The projected effect of risk factor reduction on Alzheimer's disease prevalence. Lancet Neurol. 2011; 10: 819-828. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Brookmeyer R, Johnson E, et al.: Forecasting the global burden of Alzheimer's disease. Alzheimers Dement. 2007; 3: 186-191. [DOI] [PubMed] [Google Scholar]
- 3.Ohara T, Hata J, et al.: Trends in dementia prevalence, incidence, and survival rate in a Japanese community. Neurology. 2017; 88: 1925-1932. [DOI] [PubMed] [Google Scholar]
- 4.Roehr S, Pabst A, et al.: Is dementia incidence declining in high-income countries? A systematic review and meta-analysis. Clin Epidemiol. 2018; 10: 1233-1247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Satizabal CL, Beiser AS, et al.: Incidence of dementia over three decades in the Framingham Heart Study. N Engl J Med. 2016; 374: 523-532. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Cummings J, Lee G, et al.: Alzheimer's disease drug development pipeline: 2019. Alzheimers Dement (NY). 2019; 5: 272-293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Livingston G, Sommerlad A, et al.: Dementia prevention, intervention, and care. Lancet. 2017; 390: 2673-2734. [DOI] [PubMed] [Google Scholar]
- 8.Fiuza-Luces C, Garatachea N, et al.: Exercise is the real polypill. Physiology (Bethesda). 2013; 28: 330-358. [DOI] [PubMed] [Google Scholar]
- 9.Falck RS, Davis JC, et al.: Impact of exercise training on physical and cognitive function among older adults: a systematic review and meta-analysis. Neurobiol Aging. 2019; 79: 119-130. [DOI] [PubMed] [Google Scholar]
- 10.Zhu X, Yin S, et al.: The more the better? A meta-analysis on effects of combined cognitive and physical intervention on cognition in healthy older adults. Ageing Res Rev. 2016; 31: 67-79. [DOI] [PubMed] [Google Scholar]
- 11.Nishiguchi S, Yamada M, et al.: A 12-Week Physical and Cognitive Exercise Program Can Improve Cognitive Function and Neural Efficiency in Community-Dwelling Older Adults: A Randomized Controlled Trial. J Am Geriatr Soc. 2015; 63: 1355-1363. [DOI] [PubMed] [Google Scholar]
- 12.Hiyamizu M, Morioka S, et al.: Effects of dual task balance training on dual task performance in elderly people: a randomized controlled trial. Clin Rehabil. 2012; 26: 58-67. [DOI] [PubMed] [Google Scholar]
- 13.Nouchi R, Taki Y, et al.: Reading aloud and solving simple arithmetic calculation intervention (learning therapy) improves inhibition, verbal episodic memory, focus attention and processing speed in healthy elderly people: evidence from a randomized controlled trial. Front Hum Neurosci. 2016; 10: 217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Otsuka T, Tanemura R, et al.: Development of computer-aided cognitive training program for elderly and its effectiveness through a 6 months group intervention study. Curr Alzheimer Res. 2015; 12: 553-562. [DOI] [PubMed] [Google Scholar]
- 15.Folstein MF, Folstein SE, et al.: “Mini-mental state”: A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975; 12: 189-198. [DOI] [PubMed] [Google Scholar]
- 16.Sánchez-Cubillo I, Periáñez JA, et al.: Construct validity of the Trail Making Test: role of task-switching, working memory, inhibition/interference control, and visuomotor abilities. J Int Neuropsychol Soc. 2009; 15: 438-450. [DOI] [PubMed] [Google Scholar]
- 17.Iwatsubo T, Iwata A, et al.: Japanese and North American Alzheimer's disease neuroimaging initiative studies: Harmonization for international trials. Alzheimers Dement. 2018; 14: 1077-1087. [DOI] [PubMed] [Google Scholar]
- 18.Yesavage JA: Geriatric Depression Scale. Psychopharmacol Bull. 1988; 24: 709-711. [PubMed] [Google Scholar]
- 19.Friedman B, Heisel MJ, et al.: Psychometric properties of the 15-item Geriatric Depression Scale in functionally impaired, cognitively intact, community-dwelling elderly primary care patients. J Am Geriatr Soc. 2005; 53: 1570-1576. 10.1111/j.1532-5415.2005.53461.x. [DOI] [PubMed] [Google Scholar]
- 20.Bates D, Mächler M, et al.: Fitting linear mixed-effects models using lme4. J Stat Softw. 2015; 67: 1-48. [Google Scholar]
- 21.Buuren Sv and Groothuis-Oudshoorn K: mice: Multivariate imputation by chained equations in R. J Stat Softw. 2010; 1-68. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.