Parkinson’s disease dementia (PDD) is highly prevalent [
Meta-analyses of controlled and randomized controlled trials (RCTs) have shown the efficacy of cognitive training on cognitive and noncognitive symptoms in nondemented PD patients [
Thus, we conducted a pilot randomized crossover trial of CS in PDD patients to examine the feasibility and potential effects. We hypothesized that CS is feasible in this context and that the PDD patients participating in a CS program would show benefits in cognition, neuropsychiatric symptoms, and QoL.
The reporting of this pilot RCT follows the CONSORT guidelines [
The data assessment occurred between January 2017 and June 2017 in a long-term care facility in Kerkrade, the Netherlands. This institution is unique as, to the best knowledge of the authors, it is the only facility with a specific care unit for residents with PD and cognitive dysfunctions. At the beginning of the study, 21 PD patients lived in the unit. The aim was to include as many patients as possible from the care unit. Therefore, the inclusion criteria were defined rather broadly as (i) being a resident in the PD care unit, (ii) idiopathic PD diagnosed by a neurologist or psychiatrist, (iii) cognitive dysfunction operationalized by the Mini-Mental State Examination (MMSE; 10 to 25 points) [
The exclusion criteria were (i) depressive symptoms, which were not compatible with the participation in a CS program, (ii) bedridden residents, as this is an obstacle for participating in the group intervention, (iii) life-threatening illness, (iv) history of alcohol or drug abuse in the last three years, and (v) acute suicidal tendencies or acute psychotic symptoms.
The study was designed as a randomized crossover trial with two groups (Figure
Study design.
The allocation to group A or B was conducted by an independent member of the research group who was not involved in the study. Each group consisted of six patients. After inclusion, participants were matched according to their MMSE total score. The pairs of matched individuals were randomly allocated to either group A or B by picking a note with the patient’s identification code, which was composed of two random letters and two random numbers, from an urn. The person who was picked first was assigned to group A and the second one to group B. Each group was split into groups of three to meet the group size requirement for the CS program.
The participants and facilitators for the CS program were not blinded. Both outcome assessors were blinded at the beginning of the data collection, but at the subsequent time points of assessment, only one assessor was blinded to the group allocation for organizational reasons. Nursing staff involved in external assessments were not blind to the group allocations.
The intervention comprises a modification of the structured CS program “NEUROvitalis senseful” [
The CG received the usual PD unit care including a variety of nonpharmacological interventions such as sports, music, and arts. These activities were open and voluntary for all residents throughout the study.
Cognitive functioning was assessed with the Dutch version of the CERAD test battery [
Neuropsychiatric symptoms were assessed with the Neuropsychiatric Inventory (NPI) [
The neuropsychological examinations were administered in a fixed order by two Dutch psychologists with a Master’s degree trained in neuropsychological testing. The external assessments were conducted as interviews with the nursing staff most involved in the patients’ care.
Further data, including demographics and clinical details, were collected from the residents’ files. The levodopa equivalent daily dose (LEDD) was calculated with the Levodopa Equivalent Dose Calculator (
SPSS 25 statistics software (IBM) was used for data analyses. Only patients who completed at least 11 sessions of the CS program were included in analyses. No further techniques for dealing with missing data (e.g., imputations) were used. As the number of participants in the special setting was naturally limited and this was a pilot study, no a priori sample size calculation was performed. A post hoc power analysis was calculated using G∗power 3.1 [
The alpha level was set at 0.05 for all analyses. An adjustment of
Due to the specific crossover design, the total number of participants in the CS group increases (group A + group B,
Figure
Flowchart of the participants recruited for this study.
At baseline, all patients were classified as having PDD according to the MDS criteria [
Baseline sociodemographic and clinical characteristics of the study sample.
Group A ( |
Group B ( |
| |
---|---|---|---|
Age (years) | 76.67 ± 5.58 | 76.50 ± 8.94 | 0.970 |
Gender (male/female) | 5/1 | 5/1 | 1.000 |
Years of education | 10.17 ± 1.60 | 9.50 ± 0.55 | 0.445 |
Family status | 0.392 | ||
|
0 | 1 | |
|
3 | 1 | |
|
0 | 1 | |
|
3 | 3 | |
Inpatient (months) | 16.50 ± 14.95 | 10.33 ± 5.50 | 0.394 |
Hoehn & Yahr stage | 0.255 | ||
|
0 | 0 | |
|
2 | 1 | |
|
0 | 2 | |
|
2 | 2 | |
|
2 | 1 | |
Months since PD diagnosis | 72.00 ± 37.18 | 74.00 ± 48.84 | 0.938 |
Months since dementia diagnosis | 26.17 ± 28.20 | 27.67 ± 12.66 | 0.908 |
LEDD | 290.17 ± 236.60 | 239.67 ± 228.06 | 0.714 |
CCI | 3.50 ± 3.78 | 2.33 ± 1.03 | 0.937 |
Number of antidementiva | 0.83 ± 0.98 | 1.00 ± 0.63 | 0.699 |
Number of antidepressiva | 0.83 ± 1.17 | 0.42 ± 0.66 | 0.699 |
Number of other medications | 11.11 ± 3.27 | 10.42 ± 2.33 | 0.699 |
MMSE (max. 30 points) | 17.50 ± 5.75 | 18.17 ± 5.35 | 0.839 |
CERAD total score (max. 111 points) | 39.17 ± 9.37 | 38.50 ± 13.78 | 0.924 |
GDS (max. 15 points) | 6.00 ± 2.00 | 8.67 ± 4.46 | 0.394 |
Values are presented as the mean ± standard deviation or frequency. Abbreviations: LEDD, levodopa equivalent daily dose; CCI, Charlson Comorbidity Index; MMSE, Mini-Mental State Examination; CERAD, Consortium to Establish a Registry for Alzheimer's Disease; GDS, Geriatric Depression Scale.
No adverse effects occurred in relation to CS participation. All outcomes for both groups with between- and within-group statistics are presented in Table
Medians and ranges of all main outcomes for all time points of measurement differentiated for intervention and control group and related within- and between-comparisons.
Cognitive stimulation ( |
Usual care ( |
Cognitive stimulation vs. usual care | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Preintervention | Postintervention | Within-comparisons | Follow-up | Within-comparisons | Pre-usual care | Post-usual care | Within-comparisons | Between-comparisons | |||||
Median (range) | Median (range) |
|
|
Median (range) |
|
|
Median (range) | Median (range) |
|
|
|
|
|
|
|||||||||||||
CERAD total scorea | 35.50 (20–61) | 43.50 (23–73) |
|
0.70 | 49.50 (15.00–69.00)2 | 0.161 | 0.40 | 40.50 (19–58) | 28.50 (20–57) | 0.345 | 0.27 |
|
0.43 |
|
|||||||||||||
|
|||||||||||||
GDSb | 8.50 (3–14) | 6.00 (3–12)5 | 0.950 | 0.21 | 3.00 (1–11)3 |
|
0.61 | 8.50 (4–14) | 10.50 (3–14) | 0.832 | 0.33 | 0.808 | 0.22 |
CSDDb | 3.00 (0–18) | 2.00 (0–15) | 1.000 | 0.04 | 1.00 (0–14)4 | 0.320 | 0.41 | 5.00 (4–8) | 4.00 (2–13) | 1.000 | 0.04 | 0.500 | 0.28 |
|
|||||||||||||
|
|||||||||||||
NPI total score (items A–L)b | 10.00 (1–35)5 | 8.00 (0–51)5 | 0.439 | 0.22 | 7.50 (0–41)4 | 0.575 | 0.16 | 10.00 (1–21)1 | 13.00 (5–35) | 0.414 | 0.33 |
|
0.42 |
|
|||||||||||||
Barthel Indexa | 14.50 (2–20) | 15.00 (4–20)5 | 0.572 | 0.16 | 13.50 (0–18)4 |
|
0.71 | 15.50 (9–20) | 15.50 (9–20) | 0.414 | 0.33 | 0.961 | 0.02 |
|
|||||||||||||
|
|||||||||||||
EQ-5D-5L index valuea | 0.68 (0.09–1.00) | 0.71 (0.09–1.00)5 | 1.000 | 0.06 | 0.74 (0.05–0.83)4 | 1.000 | 0.34 | 0.71 (0.11–0.99) | 0.71 (0.16–1.00) | 0.545 | 0.65 | 1.000 | 0.08 |
EQ-5D-5L-VASa | 55.00 (5–85) | 65.00 (45–80)5 | 0.700 | 0.43 | 55.00 (35–90)3 | 1.000 | 0.28 | 50.00 (40–80) | 55.00 (5–75) | 1.000 | 0.04 | 1.000 | 0.24 |
EQ-5D-5L index value (proxy)a | 0.73 (0.02–1.00) | 0.73 (0.14–0.93) | 1.000 | 0.15 | 0.76 (0.09–1.00)3 | 1.000 | 0.20 | 0.72 (0.51–0.91) | 0.73 (0.54–1.00) | 1.000 | 0.17 | 1.000 | 0.12 |
EQ-5D-5L-VAS (proxy)a | 57.50 (10–90) | 50.00 (25–80) | 1.000 | 0.19 | 60.00 (40–80)3 | 1.000 | 0.27 | 50.00 (35–60) | 52.50 (10–70) | 1.000 | 0.17 | 1.000 | 0.06 |
QUALIDEM total scorea | 16.77 (8.53–22.25)4 | 17.24 (13.36–24.83)5 | 0.430 | 0.50 | 17.43 (7.70–23.50)3 | 0.455 | 0.49 | 17.96 (13.79–22.52)1 | 16.32 (13.45–17.78)1 | 1.000 | 0.30 | 0.995 | 0.33 |
Statistical trends are in italics. Abbreviations: CERAD, Consortium to Establish a Registry for Alzheimer's Disease; GDS, Geriatric Depression Scale; CSDD, Cornell Scale for Depression in Dementia; NPI, Neuropsychiatric Inventory; VAS, visual analog scale. aHigher scores indicate a better performance. bLower scores indicate a better performance. 1
The between-group analysis revealed a group difference favoring CS which just failed to reach statistical significance for the CERAD total score (
The within-group analysis showed a positive trend, which just failed to reach statistical significance but showed a strong effect size for the CERAD total score (
The post hoc power analysis for the within-group CS effects demonstrated a statistical power of 9.5% for small effects, 48.9% for moderate effects, and 94.2% for large effects.
The comparison of the pre- and post-usual care assessment of the CG revealed no significant results. Figures
Median change of the CERAD total score (max. 111 points; higher scores indicating better performance) of the experimental (EG) and control group (CG) from preintervention/pre-usual care to postintervention/post-usual care to follow-up assessment.
Median change of the Geriatric Depression Scale (GDS; max. 15 points; lower scores indicating less depressive symptoms) of the experimental (EG) and control group (CG) from preintervention/pre-usual care to postintervention/post-usual care to follow-up assessment.
This randomized crossover pilot study examined short- and long-term effects of an eight-week CS program for long-term care residents with PDD.
Although our data must be regarded as preliminary due to the small sample size, it shows that CS for PDD patients might be an effective therapy option. First of all, our findings indicate that CS is a safe therapy option for patients with PDD and has possible positive short-term effects on cognition and neuropsychiatric symptoms for residents with PDD in long-term care. These benefits were demonstrated with between- and within-group analyses, with moderate to strong effect sizes for all effects. Furthermore, preliminary long-term CS benefits were found at the six-week FU for depressive symptoms, again with a strong effect size. Finally, a significant deterioration was observed from baseline to FU in ADL performance of the CS group.
Despite the small sample size, this pilot trial was able to identify potential beneficial effects on cognitive and noncognitive parameters, which are consistent with CS meta-analyses in non-PD dementia patients showing positive effects on global cognition [
Notably, our data show a significant deterioration of ADL in the CS group from preintervention to FU. As PDD is a progressive neurodegenerative disease, and the impact on daily functioning is its key characteristic, this decline can most probably be ascribed to the progress of the disease. However, in the light of this decline, the benefits of CS appear to be even more remarkable. Unfortunately, as an FU assessment for the CG is lacking, no between-group analysis was possible. Therefore, potentially different rates of decline in ADL will need to be the subject of future research.
The intervention integrated well into the daily routine of this specialized PD care unit, despite the fact that the residents’ weekly schedule included further activities. This is verified by the high participation score ≥ 94%. We, therefore, conclude that the CS program is feasible for PDD patients. Furthermore, participants reported that they enjoyed the CS sessions. The nursing team was interested in the study, helped with organizing the assessments, provided information about each patient within the external assessments, and arranged the CS sessions. The CS materials remained in the institution after the final assessment, and the program is still provided to the residents.
Some limitations need to be considered when interpreting our findings. First, the power of our study is limited due to the small sample size. As we used the unique possibility to study CS effects in a specialized care unit for PDD patients, no a priori sample size estimation was possible. However, it should be noted that our post hoc power analyses indicated that the power for the within-group effects of the CS group was satisfactory at 94% for large effects. Furthermore, it seems remarkable that despite the small sample size, significant clear trends with moderate to large effect sizes were found.
The crossover design, which was used to enlarge our sample size, may be associated with some critical aspects, especially in a study of the effects of a nonpharmacological intervention, where no or limited washout effects could be expected. For example, no CG was available for analyzing between-group differences of the long-term effects. Furthermore, the patients had different numbers of measurement time points (three vs. four measurements). As PDD is a progressive disease, a possible deterioration in group B from pre-usual care to post-usual care assessment needs to be considered. Finally, one of the two assessors was not blinded to group allocation. Full blinding should be achieved in future studies.
As indicated, participants reported that they enjoyed participation in the CS program. However, a limitation is that, we did not systematically collect data on the important patient-related outcome. Therefore, future studies should include measures of the individuals’ motivation and fun while participating in a CS program. Possible approaches may include the use of training diaries and specific questionnaires as well as short qualitative interviews at the end of the CS program involving both participants and relatives/nursing staff for self- and external ratings.
A particular strength of the study is that it is the first to show preliminary data for the beneficial effects of CS in PDD patients. The consideration of the CONSORT guidelines for study reporting is a further strength. Finally, we provide FU data to investigate possible long-term effects, which has been lacking in many CS studies [
In conclusion, CS seems to be a promising and safe nonpharmacological intervention to enhance cognitive and noncognitive symptoms in PDD patients in long-term care. Further research with larger sample sizes and longer FU periods are of high importance to confirm and specify the effects of CS for out- and inpatients with PDD. The best-suited frequency and duration of CS sessions should be identified. Predictor analyses could help to identify factors influencing CS benefits. Finally, health economic evaluations will be necessary.
The data used to support the findings of this study are available from the corresponding author upon request.
EK received grants from the German Ministry of Education and Research, the German Parkinson Fonds, and the German Parkinson Society as well as honoraria from Oticon GmbH, Hamburg, Germany; Lilly Pharma GmbH, Bad Homburg, Germany; Bernafon AG, Bern, Switzerland; Desitin GmbH, Hamburg, Germany. AKF received a grant from the German Parkinson Society and honoraria from ProLog Wissen GmbH, Cologne, Germany. MED, MR, MM, JK, OT, MA, ATS, DS, and LH have no conflicts of interest to report.
Ann-Kristin Folkerts and Miriam E. Dorn have contributed equally to this work.
We thank all the residents and their relatives from the Verpleeghuis Lückerheide, Kerkrade, the Netherlands, for their participation in our study. We also thank the nursing staff for their support with the study organization and their help with the external assessments. We gratefully acknowledge Alissa Paas and Sarina Kers for conducting the CS program with the residents, as well as Julia Pauquet and Fabian Schunk for data management.
Supplementary Table 1: structure and elements of the cognitive stimulation program NEUROvitalis senseful; Supplementary Table 2: overview of the neuropsychological test battery; Supplementary Table 3: activities of the residents: a typical week as reported by the nursing staff; Supplementary Table 4: medians and ranges of all cognitive subtests and QUALIDEM subscores for all time points of measurement differentiated for CS group and CG.