J Neurol (2015) 262:881–889 DOI 10.1007/s00415-015-7647-1
ORIGINAL COMMUNICATION
Preoperative brain metabolism and quality of life after subthalamic nucleus stimulation in Parkinson’s disease Sophie Langner-Lemercier • Sophie Drapier • Florian Naudet • Nathalie Le Clanche • Jean-Franc¸ois Houvenaghel • Paul Sauleau • Pierre Jannin • Claire Haegelen • Florence Le Jeune • Marc Ve´rin
Received: 20 October 2014 / Revised: 8 January 2015 / Accepted: 13 January 2015 / Published online: 30 January 2015 Springer-Verlag Berlin Heidelberg 2015
Abstract Subthalamic nucleus deep brain stimulation (STN-DBS) has been proven to improve health-related quality of life (HRQoL) in patients with Parkinson’s disease (PD) presenting medically refractory motor complications and dyskinesia. However, some patients fail to benefit from STN-DBS despite rigorous preoperative selection. We postulated that they have a particular, clinically ineloquent, brain metabolism before surgery. We divided 40 stimulated PD patients into two groups (responders vs. nonresponders) depending on whether they reported or not a clinically significant improvement in their quality of life 1 year after surgery. We retrospectively compared their preoperative brain metabolism on the basis of 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) scans. We also analyzed their neuropsychological and psychiatric
profiles before and after surgery. All 40 patients met the STN-DBS selection criteria, but only 50 % of them had significantly improved 1 year after surgery. Preoperative PET scans showed that metabolism was higher in the left insula, both inferior frontal gyri and left precentral gyrus in nonresponders than in responders. Clinically, postoperative motor scores were similar in both groups, but a worsening of the depression score was observed among nonresponders. PET imaging revealed that nonresponders were characterized by distinctive brain functioning pre-surgery, in regions involved in associative and limbic circuits, as a result of PDrelated degeneration. STN-DBS may have interfered with this already abnormal circuitry, leading to the occurrence of complex nonmotor symptoms reducing quality of life. Preoperative brain metabolism could be a useful biomarker for anticipating STN-DBS efficacy in terms of HRQoL in the context of advanced PD.
Electronic supplementary material The online version of this article (doi:10.1007/s00415-015-7647-1) contains supplementary material, which is available to authorized users. S. Langner-Lemercier (&) S. Drapier J.-F. Houvenaghel M. Ve´rin Department of Neurology, Rennes University Hospital, 2 rue Henri Le Guilloux, 35033 Rennes Cedex 9, France e-mail: [email protected] S. Drapier F. Naudet N. Le Clanche J.-F. Houvenaghel P. Sauleau F. Le Jeune M. Ve´rin Behavior and Basal Ganglia Research Unit 4712, University of Rennes 1, Rennes, France F. Naudet N. Le Clanche Department of Psychiatry, Guillaume Re´gnier Hospital Centre, Rennes, France
P. Sauleau Department of Neurophysiology, Rennes University Hospital, Rennes, France P. Jannin C. Haegelen INSERM UMR 1099, LTSI, University of Rennes 1, Rennes, France C. Haegelen Department of Neurosurgery, Rennes University Hospital, Rennes, France F. Le Jeune Department of Nuclear Medicine, Eugene Marquis Hospital Centre, Rennes, France
F. Naudet Clinical Investigation Center, INSERM 0203, Rennes University Hospital, University of Rennes 1, Rennes, France
123
882
Keywords Parkinson’s disease PET Quality of life Subthalamic nucleus stimulation
Introduction Subthalamic nucleus deep brain stimulation (STN-DBS) is currently recommended for patients with Parkinson’s disease (PD) presenting medically refractory motor symptoms [1]. The efficacy of STN-DBS was initially measured in terms of improved PD motor symptoms and reduced medication after surgery [2]. However, to consider the nonmotor symptoms of PD [3], and nonmotor effects of STN-DBS [4], health-related quality of life (HRQoL) has become the outcome measure of choice [5]. HRQoL reflects patient’s subjective view of the impact of disease and treatment on their physical, social, and mental well-being [6]. The most popular HRQoL instrument for PD is the PDQ-39 scale, which assesses eight motor and nonmotor subdomains [7]. STN-DBS brings a mean improvement of 34.5 ± 15.3 % in PDQ-39 total scores [8], mainly owing to a dramatic improvement in motor symptoms. Scores on mental and social functioning, both important determinants of quality of life, change little [9]. To increase the likelihood of a beneficial outcome, STNDBS candidates are carefully selected, and those with poor levodopa responsiveness, axial symptoms, psychiatric disturbance or cognitive impairment are excluded [1]. Although the patients who do receive STN-DBS, therefore, seem preoperatively homogeneous, we have noted considerable disparities in the impact of STN-DBS on their quality of life, even within the first year, pointing to the existence of infraclinical preoperative differences between future responders and nonresponders, arising from distinctive brain functioning. We, therefore, retrospectively compared the preoperative brain metabolism of responders and nonresponders using 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) and analyzed their clinical data, looking for factors that might determine changes in HRQoL.
Materials and methods Participants Participants had consecutively undergone bilateral STNDBS after a preoperative 18F-FDG PET brain scan at Rennes University Hospital (France) between September 2005 and December 2012. They all met the criteria of the Parkinson’s Disease UK Brain Bank for idiopathic
123
J Neurol (2015) 262:881–889
Parkinson’s disease [10]. PD patients were selected for bilateral STN-DBS if they had disabling levodopa-induced symptoms refractory to medical treatment, disease lasting more than 5 years, and if they were under 70 years old. They were free from axial motor signs, cognitive decline (Mattis Dementia Rating Scale [130 and absence of major impaired executive functions), and psychiatric disturbance. A preoperative magnetic resonance imaging (MRI) brain scan excluded morphological contraindications for neurosurgery, such as atrophy or vascular lesions. Surgical procedure and electrode location Quadripolar deep brain stimulation electrodes (3389; Medtronic, Minneapolis, MN, USA) were stereotactically implanted in the left and right STN. The overall methodology was similar to that previously described by Deuschl et al. [11], including local anesthesia, MR imaging, targeting by means of a stereotactic frame, and microelectrode recording to determine the upper boundary of the STN. The exact locations of the selected electrode contacts were determined using stereotactic coordinates derived from a postoperative 3D computer tomography (CT) scan. The contact’s coordinates were expressed as millimeters along three axes originating from the middle of the bicommissural line: the first axis was parallel to the bicommissural line, the second was perpendicular to the anterior commissure-posterior commissure (AC-PC) line, and the third was perpendicular to the midsagittal plane. The pulse generator (Kinetra or Soletra, Medtronic, Minneapolis, MN, USA) was placed in a subcutaneous infraclavicular pocket either on the same day or a few days later. Clinical follow-up and treatment optimization Follow-up was carried out in accordance with the Core Assessment Program for Intracerebral Transplantation (CAPIT) [12], 3 months before and 12 months after STNDBS. Patients underwent Part III of the Unified Parkinson’s Disease Rating Scale (UPDRS) in an OFF-drug condition, where medication was withdrawn at least 12 h beforehand, and an ON-drug condition, defined as the best motor state following the usual morning dose of levodopa enhanced by 50 mg of levodopa, as well as in OFF-stimulation and ONstimulation conditions after surgery. In addition, patients underwent the UPDRS-I, UPDRS-II, and UPDRS-IV, and were rated on the Hoehn and Yahr (HY) and Schwab and England (SE) scales. In the ON-drug condition, they underwent a neuropsychological battery that included the Mattis Dementia Rating Scale (MDRS) [13], Stroop test [14], Trail Making Test (TMT) [15], Wisconsin Card Sorting Test (WCST) [16], and phonemic and semantic
J Neurol (2015) 262:881–889
verbal fluency tasks [17], as well as a psychiatric exami˚ sberg Depression Ratnation featuring the Montgomery-A ing Scale (MADRS) [18] and the Apathy Evaluation Scale (AES) [19]. The levodopa equivalent daily dose (LEDD) was calculated for each patient using Deuschl’s method [11]. During follow-up, experienced neurologists regularly reassessed the patients and adjusted the stimulation parameters and medication to provide the most suitable antiparkinsonian therapy. Quality of life assessment Patients were asked to fill in the PDQ-39 quality-of-life scale [20]. Its eight dimensions (mobility, activities of daily living, emotional wellbeing, stigma, social support, cognition, communication, and bodily discomfort) are each scored between 0 and 100, and their average produces a total score. The lower this score, the better the perceived health status. We considered a clinically significant improvement in quality of life to correspond to an effect size of 0.4 on the PDQ-39 total score, and used this threshold to dichotomize our data. In line with previous studies [21], it was coherent with the consensual notion that 0.4 is a medium effect size [22]. Applying this definition to our data meant that a reduction C5 points in the PDQ-39 score represented an improvement. Beyond its statistical properties, this cut-off point was consistent with our clinical impressions. PET data acquisition Each patient underwent an 18F-FDG PET brain scan three months before surgery. All PET measurements were performed in the same center and in the same conditions, using a dedicated Discovery ST PET/CT scanner in 2D mode (General Electric Medical Systems, Milwaukee, WI, USA). X-ray CT-based attenuation correction was performed prior to the emission scan. Patients were then studied using intravenous 18F-FDG in a resting state with their eyes open. They had fasted for at least 6 h (normal glycemia levels 80–120 mg/dL), but had taken their usual medication. Following scatter, deadtime, and random corrections, the PET images were reconstructed by means of 2D filtered backprojection, yielding 47 contiguous transaxial 3.75-mm thick slices. PET-data analysis All the participant’s data were normalized to the Montreal Neurological Institute (MNI) template. An affine transformation was performed to determine the 12 optimum parameters for registering the brain to the template. The
883
differences between the transformed image and the template were removed by applying a nonlinear registration method. The spatially normalized images were then smoothed, using an isotropic 12-mm full-width at halfmaximum Gaussian kernel to compensate for interindividual anatomical variability and render the imaging data more normally distributed. Standardized data were analyzed voxel-by-voxel using statistical parametric mapping software (SPM; Wellcome Department of Cognitive Neurology, London, United Kingdom) implemented in Matlab, Version 7 (MathWorks, Sherborn, MA, USA). SPM combines the general linear model (to create the statistical map) with the theory of Gaussian fields to make statistical inferences about regional effects. To identify the metabolic patterns of nonresponders versus responders, we ran a two-sample t test at each voxel on SPM. Clusters of at least 20 contiguous voxels with a threshold of p \ 0.005 were considered to be significant. Statistics Descriptive statistics were obtained for each variable. They are expressed as means (± standard deviation, SD) for the quantitative data and as numbers (%) for the qualitative outcomes. The relative magnitude of the effects was standardized to a percentage by applying the following formula: magnitude of the effect (%) = (final value-baseline value) 9 100/baseline value. We compared the two groups on the distribution of the clinical parameters using nonparametric tests (Wilcoxon rank-sum test and Fisher’s exact test) with a significance threshold of p \ 0.05.
Results We studied 40 consecutive patients. All patients were free from unstable comorbidities both at the time of surgery and during follow-up. There were no major postoperative complications. Comparison of responder’s and nonresponder’s quality of life data At one year post STN-DBS, there was an 18 % decrease in the PDQ-39 total score. When we applied the 5-point threshold for the change in the PDQ-39 total score, 50 % (n = 20) of our patients were classified as responders. PDQ-39 total scores at baseline did not differ significantly between the two groups although responders tended to have a worse perceived HRQoL. The mean decrease in the responders’ PDQ-39 total score was -45.3 % (± 18.0 %),
123
884
J Neurol (2015) 262:881–889
Table 1 PDQ-39 total score and domain scores Preoperative
Postoperative
Responders
Non-responders
p
$
Responders
Non-responders
p$
Total
35.9 ± 11.8
29.0 ± 12.4
0.09
20.2 ± 11.2
31.1 ± 12.7
0.07
Mobility
47.5 ± 18.4
35.1 ± 20.8
0.07
27.4 ± 18.1
31.5 ± 19.0
0.50
Activities of daily living
43.6 ± 17.9
33.8 ± 19.4
0.06
20.6 ± 15.6
29.8 ± 18.4
0.11
Emotional well-being
38.6 ± 18.0
30.0 ± 14.0
0.09
19.8 – 19.4
40 – 21.5
0.01*
Stigma
38.6 ± 23.5
30.9 ± 20.4
0.29
13.8 ± 13.8
25.6 ± 25.6
0.18
Social support
15.6 ± 19.5
13.8 ± 20.5
0.48
11.7 ± 18.6
20.4 ± 16.1
0.05
Cognition Communication
29.3 ± 17.4 35.7 – 19.6
26.0 ± 17.5 19.6 – 14.4
0.55
ORIGINAL COMMUNICATION
Preoperative brain metabolism and quality of life after subthalamic nucleus stimulation in Parkinson’s disease Sophie Langner-Lemercier • Sophie Drapier • Florian Naudet • Nathalie Le Clanche • Jean-Franc¸ois Houvenaghel • Paul Sauleau • Pierre Jannin • Claire Haegelen • Florence Le Jeune • Marc Ve´rin
Received: 20 October 2014 / Revised: 8 January 2015 / Accepted: 13 January 2015 / Published online: 30 January 2015 Springer-Verlag Berlin Heidelberg 2015
Abstract Subthalamic nucleus deep brain stimulation (STN-DBS) has been proven to improve health-related quality of life (HRQoL) in patients with Parkinson’s disease (PD) presenting medically refractory motor complications and dyskinesia. However, some patients fail to benefit from STN-DBS despite rigorous preoperative selection. We postulated that they have a particular, clinically ineloquent, brain metabolism before surgery. We divided 40 stimulated PD patients into two groups (responders vs. nonresponders) depending on whether they reported or not a clinically significant improvement in their quality of life 1 year after surgery. We retrospectively compared their preoperative brain metabolism on the basis of 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) scans. We also analyzed their neuropsychological and psychiatric
profiles before and after surgery. All 40 patients met the STN-DBS selection criteria, but only 50 % of them had significantly improved 1 year after surgery. Preoperative PET scans showed that metabolism was higher in the left insula, both inferior frontal gyri and left precentral gyrus in nonresponders than in responders. Clinically, postoperative motor scores were similar in both groups, but a worsening of the depression score was observed among nonresponders. PET imaging revealed that nonresponders were characterized by distinctive brain functioning pre-surgery, in regions involved in associative and limbic circuits, as a result of PDrelated degeneration. STN-DBS may have interfered with this already abnormal circuitry, leading to the occurrence of complex nonmotor symptoms reducing quality of life. Preoperative brain metabolism could be a useful biomarker for anticipating STN-DBS efficacy in terms of HRQoL in the context of advanced PD.
Electronic supplementary material The online version of this article (doi:10.1007/s00415-015-7647-1) contains supplementary material, which is available to authorized users. S. Langner-Lemercier (&) S. Drapier J.-F. Houvenaghel M. Ve´rin Department of Neurology, Rennes University Hospital, 2 rue Henri Le Guilloux, 35033 Rennes Cedex 9, France e-mail: [email protected] S. Drapier F. Naudet N. Le Clanche J.-F. Houvenaghel P. Sauleau F. Le Jeune M. Ve´rin Behavior and Basal Ganglia Research Unit 4712, University of Rennes 1, Rennes, France F. Naudet N. Le Clanche Department of Psychiatry, Guillaume Re´gnier Hospital Centre, Rennes, France
P. Sauleau Department of Neurophysiology, Rennes University Hospital, Rennes, France P. Jannin C. Haegelen INSERM UMR 1099, LTSI, University of Rennes 1, Rennes, France C. Haegelen Department of Neurosurgery, Rennes University Hospital, Rennes, France F. Le Jeune Department of Nuclear Medicine, Eugene Marquis Hospital Centre, Rennes, France
F. Naudet Clinical Investigation Center, INSERM 0203, Rennes University Hospital, University of Rennes 1, Rennes, France
123
882
Keywords Parkinson’s disease PET Quality of life Subthalamic nucleus stimulation
Introduction Subthalamic nucleus deep brain stimulation (STN-DBS) is currently recommended for patients with Parkinson’s disease (PD) presenting medically refractory motor symptoms [1]. The efficacy of STN-DBS was initially measured in terms of improved PD motor symptoms and reduced medication after surgery [2]. However, to consider the nonmotor symptoms of PD [3], and nonmotor effects of STN-DBS [4], health-related quality of life (HRQoL) has become the outcome measure of choice [5]. HRQoL reflects patient’s subjective view of the impact of disease and treatment on their physical, social, and mental well-being [6]. The most popular HRQoL instrument for PD is the PDQ-39 scale, which assesses eight motor and nonmotor subdomains [7]. STN-DBS brings a mean improvement of 34.5 ± 15.3 % in PDQ-39 total scores [8], mainly owing to a dramatic improvement in motor symptoms. Scores on mental and social functioning, both important determinants of quality of life, change little [9]. To increase the likelihood of a beneficial outcome, STNDBS candidates are carefully selected, and those with poor levodopa responsiveness, axial symptoms, psychiatric disturbance or cognitive impairment are excluded [1]. Although the patients who do receive STN-DBS, therefore, seem preoperatively homogeneous, we have noted considerable disparities in the impact of STN-DBS on their quality of life, even within the first year, pointing to the existence of infraclinical preoperative differences between future responders and nonresponders, arising from distinctive brain functioning. We, therefore, retrospectively compared the preoperative brain metabolism of responders and nonresponders using 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) and analyzed their clinical data, looking for factors that might determine changes in HRQoL.
Materials and methods Participants Participants had consecutively undergone bilateral STNDBS after a preoperative 18F-FDG PET brain scan at Rennes University Hospital (France) between September 2005 and December 2012. They all met the criteria of the Parkinson’s Disease UK Brain Bank for idiopathic
123
J Neurol (2015) 262:881–889
Parkinson’s disease [10]. PD patients were selected for bilateral STN-DBS if they had disabling levodopa-induced symptoms refractory to medical treatment, disease lasting more than 5 years, and if they were under 70 years old. They were free from axial motor signs, cognitive decline (Mattis Dementia Rating Scale [130 and absence of major impaired executive functions), and psychiatric disturbance. A preoperative magnetic resonance imaging (MRI) brain scan excluded morphological contraindications for neurosurgery, such as atrophy or vascular lesions. Surgical procedure and electrode location Quadripolar deep brain stimulation electrodes (3389; Medtronic, Minneapolis, MN, USA) were stereotactically implanted in the left and right STN. The overall methodology was similar to that previously described by Deuschl et al. [11], including local anesthesia, MR imaging, targeting by means of a stereotactic frame, and microelectrode recording to determine the upper boundary of the STN. The exact locations of the selected electrode contacts were determined using stereotactic coordinates derived from a postoperative 3D computer tomography (CT) scan. The contact’s coordinates were expressed as millimeters along three axes originating from the middle of the bicommissural line: the first axis was parallel to the bicommissural line, the second was perpendicular to the anterior commissure-posterior commissure (AC-PC) line, and the third was perpendicular to the midsagittal plane. The pulse generator (Kinetra or Soletra, Medtronic, Minneapolis, MN, USA) was placed in a subcutaneous infraclavicular pocket either on the same day or a few days later. Clinical follow-up and treatment optimization Follow-up was carried out in accordance with the Core Assessment Program for Intracerebral Transplantation (CAPIT) [12], 3 months before and 12 months after STNDBS. Patients underwent Part III of the Unified Parkinson’s Disease Rating Scale (UPDRS) in an OFF-drug condition, where medication was withdrawn at least 12 h beforehand, and an ON-drug condition, defined as the best motor state following the usual morning dose of levodopa enhanced by 50 mg of levodopa, as well as in OFF-stimulation and ONstimulation conditions after surgery. In addition, patients underwent the UPDRS-I, UPDRS-II, and UPDRS-IV, and were rated on the Hoehn and Yahr (HY) and Schwab and England (SE) scales. In the ON-drug condition, they underwent a neuropsychological battery that included the Mattis Dementia Rating Scale (MDRS) [13], Stroop test [14], Trail Making Test (TMT) [15], Wisconsin Card Sorting Test (WCST) [16], and phonemic and semantic
J Neurol (2015) 262:881–889
verbal fluency tasks [17], as well as a psychiatric exami˚ sberg Depression Ratnation featuring the Montgomery-A ing Scale (MADRS) [18] and the Apathy Evaluation Scale (AES) [19]. The levodopa equivalent daily dose (LEDD) was calculated for each patient using Deuschl’s method [11]. During follow-up, experienced neurologists regularly reassessed the patients and adjusted the stimulation parameters and medication to provide the most suitable antiparkinsonian therapy. Quality of life assessment Patients were asked to fill in the PDQ-39 quality-of-life scale [20]. Its eight dimensions (mobility, activities of daily living, emotional wellbeing, stigma, social support, cognition, communication, and bodily discomfort) are each scored between 0 and 100, and their average produces a total score. The lower this score, the better the perceived health status. We considered a clinically significant improvement in quality of life to correspond to an effect size of 0.4 on the PDQ-39 total score, and used this threshold to dichotomize our data. In line with previous studies [21], it was coherent with the consensual notion that 0.4 is a medium effect size [22]. Applying this definition to our data meant that a reduction C5 points in the PDQ-39 score represented an improvement. Beyond its statistical properties, this cut-off point was consistent with our clinical impressions. PET data acquisition Each patient underwent an 18F-FDG PET brain scan three months before surgery. All PET measurements were performed in the same center and in the same conditions, using a dedicated Discovery ST PET/CT scanner in 2D mode (General Electric Medical Systems, Milwaukee, WI, USA). X-ray CT-based attenuation correction was performed prior to the emission scan. Patients were then studied using intravenous 18F-FDG in a resting state with their eyes open. They had fasted for at least 6 h (normal glycemia levels 80–120 mg/dL), but had taken their usual medication. Following scatter, deadtime, and random corrections, the PET images were reconstructed by means of 2D filtered backprojection, yielding 47 contiguous transaxial 3.75-mm thick slices. PET-data analysis All the participant’s data were normalized to the Montreal Neurological Institute (MNI) template. An affine transformation was performed to determine the 12 optimum parameters for registering the brain to the template. The
883
differences between the transformed image and the template were removed by applying a nonlinear registration method. The spatially normalized images were then smoothed, using an isotropic 12-mm full-width at halfmaximum Gaussian kernel to compensate for interindividual anatomical variability and render the imaging data more normally distributed. Standardized data were analyzed voxel-by-voxel using statistical parametric mapping software (SPM; Wellcome Department of Cognitive Neurology, London, United Kingdom) implemented in Matlab, Version 7 (MathWorks, Sherborn, MA, USA). SPM combines the general linear model (to create the statistical map) with the theory of Gaussian fields to make statistical inferences about regional effects. To identify the metabolic patterns of nonresponders versus responders, we ran a two-sample t test at each voxel on SPM. Clusters of at least 20 contiguous voxels with a threshold of p \ 0.005 were considered to be significant. Statistics Descriptive statistics were obtained for each variable. They are expressed as means (± standard deviation, SD) for the quantitative data and as numbers (%) for the qualitative outcomes. The relative magnitude of the effects was standardized to a percentage by applying the following formula: magnitude of the effect (%) = (final value-baseline value) 9 100/baseline value. We compared the two groups on the distribution of the clinical parameters using nonparametric tests (Wilcoxon rank-sum test and Fisher’s exact test) with a significance threshold of p \ 0.05.
Results We studied 40 consecutive patients. All patients were free from unstable comorbidities both at the time of surgery and during follow-up. There were no major postoperative complications. Comparison of responder’s and nonresponder’s quality of life data At one year post STN-DBS, there was an 18 % decrease in the PDQ-39 total score. When we applied the 5-point threshold for the change in the PDQ-39 total score, 50 % (n = 20) of our patients were classified as responders. PDQ-39 total scores at baseline did not differ significantly between the two groups although responders tended to have a worse perceived HRQoL. The mean decrease in the responders’ PDQ-39 total score was -45.3 % (± 18.0 %),
123
884
J Neurol (2015) 262:881–889
Table 1 PDQ-39 total score and domain scores Preoperative
Postoperative
Responders
Non-responders
p
$
Responders
Non-responders
p$
Total
35.9 ± 11.8
29.0 ± 12.4
0.09
20.2 ± 11.2
31.1 ± 12.7
0.07
Mobility
47.5 ± 18.4
35.1 ± 20.8
0.07
27.4 ± 18.1
31.5 ± 19.0
0.50
Activities of daily living
43.6 ± 17.9
33.8 ± 19.4
0.06
20.6 ± 15.6
29.8 ± 18.4
0.11
Emotional well-being
38.6 ± 18.0
30.0 ± 14.0
0.09
19.8 – 19.4
40 – 21.5
0.01*
Stigma
38.6 ± 23.5
30.9 ± 20.4
0.29
13.8 ± 13.8
25.6 ± 25.6
0.18
Social support
15.6 ± 19.5
13.8 ± 20.5
0.48
11.7 ± 18.6
20.4 ± 16.1
0.05
Cognition Communication
29.3 ± 17.4 35.7 – 19.6
26.0 ± 17.5 19.6 – 14.4
0.55
