Depression, quality of life, activities of daily living, and subjective memory after deep brain stimulation in Parkinson disease—A reliable change index analysis

Objectives In the field of Parkinson disease (PD) research, many studies have shown that deep brain stimulation (DBS) can soften side effects, which arise during long‐term medical therapy. This study focuses on the changes in depressive symptoms, quality of life (with the subdivisions physical and mental health), activities of daily living, and subjective memory functioning in PD patients testing the baseline and the outcome 1 year after DBS. Methods For the first time, the reliable change index (RCI) methodology was applied to compare PD‐DBS patients (n = 22) with best medically treated PD patients (PD‐BMT; n = 28), subjects with mild cognitive impairment (MCI, n = 43) and healthy controls (n = 25) in the above‐mentioned domains. The used questionnaires included the revised Beck Depression Inventory (BDI‐II), the Short Form (36) Health Survey (SF‐36), the Bayer Activities of Daily Living Scale (B‐ADL), and the Forgetfulness Assessment Inventory (FAI). Results The reliable change indices show high constant or improved results of the PD‐DBS patients in the domains subjective memory (85.7%‐100.0%), activities of daily living (60.0%‐90.0%), physical health summary (77.8%), depressive symptoms (61.9%), and mental health summary (50.0%) in comparison with the PD‐BMT, MCI, and control group. Conclusions DBS is an established alternative to best medical treatment of PD. The comparisons between the PD‐DBS and PD‐BMT groups do suggest that the domains mental health, depressive symptoms, and physical health benefit most, while the domains activities of daily living and subjective memory functioning are rather constant. Nevertheless, further research is needed to identify mechanisms and predictors that lead to improvement in individual cases.


| INTRODUCTION
Deep brain stimulation (DBS) of the subthalamic nucleus is an established treatment for severe motor complications in Parkinson disease (PD), and since it is usually a lifelong therapy, it is essential to carefully evaluate beneficial and inadvertent effects in the long term.
Studies demonstrate a remarkable improvement of motor symptoms in PD patients, whereas psychosocial impacts of DBS surgery including social adjustment, coping strategies, and mental health-related quality of life may be variable. [1][2][3][4] It is particularly difficult to determine whether the postoperative nonmotor deficiencies are related to the progression of the disease itself, to surgery or permanent stimulation.
The gold standard to evaluate the nonmotor symptoms in PD would be a randomized controlled trial, but this is not suitable for assessing single patients in the clinical setting-therefore, the reliable change index (RCI) analysis has been designed. 5 The RCI analysis is capable of examining the influence of the disease progression over time and measures the real change in an individual case. Additionally, the standardized effect size according to the classification of Cohen has been calculated to assess the practical relevance of the changes.
Recent studies have used this RCI methodology to assess the cognitive changes of deep brain stimulated Parkinson patients, [6][7][8][9][10][11] but so far, no study has used it on measuring the psychosocial outcome.

| Data collection
The study collective (n = 118) was divided into four groups. Two groups consisted of PD patients-one group received best medical treatment (PD-BMT; n = 28) and the other received DBS and best medical treatment (PD-DBS; n = 22). PD patients with motor symptom duration of at least 5 years, good response to levodopa and/or apomorphine, and drug resistant motor complications or drug resistant tremor were included. Patients suffering from secondary parkinsonian syndromes, atypical Parkinsonism, and patients with severe cognitive impairments, such as dementia and uncontrolled mental disorders, were excluded. In addition, subjects with comorbidities and structural brain lesions that precluded DBS surgery had to be excluded from the study. The evaluation was performed during the "on-state." One group included patients with MCI (n = 43) defined according to the Petersen criteria. These patients complain about a defective cognition and show an abnormal cognitive function for their age but do not fulfill criteria for dementia. Activities of daily living are unimpaired. 12 The healthy control group (n = 25) consisted of individuals without PD and without cognitive impairment.
Test-retest interval was 12 months. The Mini-Mental-State Examination (MMSE)-to assess the cognitive state 13 -and the "Wortschatz-Test" (WST-IQ)-to asses verbal intelligence levels and speech comprehension 14 -were used as comparability parameters between the different groups. Furthermore, all study participants had to answer the following study relevant questionnaires: Beck Depression Inventory (BDI-II) is a 21-item instrument to quantify the level of depression by asking how often the subject felt certain ways within the past 2 weeks rated on a 4-point scale ranging from 0 (no symptoms) to 3 (very intense symptoms) with a maximum score of 63. The results categorize the severity of depressive symptoms in minimal (0-13), mild (14)(15)(16)(17)(18)(19), moderate (20)(21)(22)(23)(24)(25)(26)(27)(28), and severe • We compare PD-DBS patients with best medically treated PD patients, subjects with mild cognitive impairments and healthy controls.
• For comparison, we used the reliable change index (RCI) methodology.
• The comparisons between the PD-DBS and PD-BMT groups do suggest that the domains mental health, depressive symptoms, and physical health benefit most, while the domains activities of daily living and subjective memory functioning are rather constant.
physical and mental health summary, the latter two being used in the current study. The score ranges from 0 to 100-lower scores reflect worse subjective quality of life. 17 Bayer Activities of Daily Living Scale (B-ADL) is a 25-item questionnaire used to assess instrumental (eg, shopping and food preparation) and noninstrumental (eg, managing everyday activities and finding the way) activities of daily living. Ratings on a 10-point scale between "never" and "always" sum up to a total score between 0 and 10-higher scores reflect increased impairment of everyday activities. 18 Forgetfulness Assessment Inventory (FAI) is a 16-items questionnaire used to measure subjective memory complaints scored on the basis of a 5-point scale between "never" and "very often." For statistical analysis, the average score across all items was used-higher scores reflect poorer subjective memory functioning. 19 All groups were comparable for age, education, WST-IQ, and BDI-II and did not show statistically significant differences for age, education, WST-IQ, and BDI-II. The MMSE score showed a significant group difference P = .001 (η 2 = .085) with a small to medium effect. The follow-up analysis via linear contrasts revealed that the healthy control group had higher MMSE values compared with the other groups (Ps ≤ .019). The difference for the patient groups was negligible. See Table 1 for details.

| Statistical analyses
As PD is a progressive disorder, neuropsychological follow-up measurements are particularly important in individual case diagnosis. A distinction has to be made between general and differential changes in cognition. Thus, cognitive decline is also possible with constant fine motor skills or unchanged affectivity. Even therapeutic measures can influence cognitive, affective, or psychosocial variables. Temporary changes due to medication, impaired motor functioning, depressive symptoms, short-term fluctuations, or freezing need to be taken into consideration. In order to meet those demands, the retest reliability (r tt ), which was, in this study, retrieved via correlation analysis, has to be known. As a broadly examined phenomenon in serial assessments in neuropsychological research, practice effects-due to natural recovery, intervention, or prior exposure to the questionnaires used-have to be acknowledged and minimized by using the adjusted RCI formula that controls it. 5,20,21 The difference (X 2 − X 1 ) describes the individual change in performance of a test person. (M 2 − M 1 ) reflects the practice effect on retesting of the respective group-PD-BMT, MCI, or control. The standard error of difference (SED) is useful for the individual case diagnosis for elderly people, since it allows distinguishing between cognitive deterioration due to illness and general age-related degradation of cognitive capacity. 5,20-22 Ringendahl assumes that a change is significant for a z value of ±1.64 (P = .05, one-tailed). 20 So these values are used to calculate the confidence intervals for determining the upper and lower limits for significant changes. The results of the individual DBS test persons RCIs are then compared with the limits of the PD-BMT, MCI, and control group.
To assess the practical relevance of changes found by the RCI comparisons, the standardized effect size has been calculated using  23 The absolute values of Cohen d can be interpreted as follows:|d| ≥ 0.20 is a small,|d|≥ 0.50 is a medium, and|d|≥ 0.80 is a large effect. 24 Based on the 95% confidence intervals for deterioration and improve-  Table 5.

| PD-DBS versus PD-BMT
The RCI analysis showed an improvement of 42.86% and a deterioration of 38.10% of the PD-DBS patients for depressive symptoms (BDI-II) with no effect (|d ppc2 | = .000

| PD-DBS versus MCI
Again, the depressive symptoms (BDI-II) depicted an improvement of 42.86% and a deterioration of 38.10% of the PD-DBS patients with

| PD-DBS versus healthy control
The depressive symptoms (BDI-II) showed an improvement of 42.86% and a deterioration of 38.10% of the PD-DBS patients with no effect     however, verify the improvements also found in the current study. 25,26 The majority of the studies investigating quality of life demonstrated an improvement of the physical health-which is no surprise since DBS is known for its fast effects on the physical symptomsand no change or even deterioration of the mental health summary due to the high expectations of the patients, the progression of PD values, which had to be taken out of consideration (see Table 2).
Due to the small sample size, only medium to large effects reached statistical significance. Natural regression effect is a limitation that In order to be able to estimate the psychosocial outcome after the deep brain stimulation individually for each patient as precisely as possible-despite the still existing uncertainties about the mechanisms and emerging impairments-further studies are necessary in order to find predictors for improvement or worsening and to develop optimal individual therapy options.

CONFLICT OF INTEREST
None declared.

AUTHOR CONTRIBUTIONS
Dr Foki collected the data and assisted discussing and writing the article. Dr Wiesbauer performed the statistical analysis and assisted discussing and writing the article. Dr Pirker collected the data and assisted with writing the article. Dr Novak collected the data and assisted with writing the article. Dr Pusswald collected the data and assisted with writing the article. Dr Lehrner designed the study, supervised the data collection, and wrote the paper.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.