European Journal of Neurology 2014, 21: 922–928
doi:10.1111/ene.12412
The burden of white matter hyperintensities is a predictor of progressive mild cognitive impairment in patients with Parkinson’s disease M. K. Sunwooa, S. Jeonb, J. H. Hama, J. Y. Hongc, J. E. Leea, J.-M. Leeb, Y. H. Sohna and P. H. Leea,d a
Department of Neurology, Yonsei University College of Medicine, Seoul; bDepartment of Biomedical Engineering, Hanyang University,
Seoul; cDepartment of Neurology, Yonsei University Wonju College of Medicine, Wonju; and dSeverance Biomedical Science Institute, Seoul, Korea
Keywords:
dementia converters, Parkinson’s disease, white matter hyperintensities Received 11 November 2013 Accepted 7 February 2014
Background and purpose: To evaluate whether white matter hyperintensities (WMHs) may act as an independent predictor for progression of cognitive status, the authors analyzed the longitudinal effects of WMHs on cognitive dysfunction in non-demented patients with Parkinson’s disease (PD). Methods: A total of 111 patients with PD were enrolled, including subjects with mild cognitive impairment (MCI, n = 65) and cognitively normal subjects (CN, n = 46). These individuals were classified as MCI converters (n = 22) or MCI non-converters (n = 43) and CN converters (n = 18) or CN non-converters (n = 28) based on whether they were subsequently diagnosed with PD dementia or PD-MCI during a minimum 24-month follow-up. The WMH burden and the Cholinergic Pathway Hyperintensities Scale (CHIPS) and their relationships to longitudinal changes in cognitive performance were examined. Results: PD-MCI converters had larger WMH volume (14421.0 vs. 5180.4, P < 0.001) and higher CHIPS score (22.6 vs. 11.2, P = 0.001) compared with PDMCI non-converters. Logistic regression analysis revealed in patients with PD-MCI that WMH volume (odds ratio 1.616, P = 0.009) and CHIPS score (odds ratio 1.084, P = 0.007) were independently associated with PD dementia conversion. However, WMH volume and CHIPS score did not differ between PD-CN converters and PD-CN non-converters. In patients with PD-MCI, both WMH volume and CHIPS score were closely associated with longitudinal decline in general cognition, semantic fluency and Stroop test scores. Conclusions: The present study demonstrates that WMH burden is a significant predictor of conversion from PD-MCI to PD dementia and is related to ongoing decline in frontal-lobe-based cognitive performance.
Introduction Cognitive impairment is one of most disabling nonmotor symptoms of Parkinson’s disease (PD). According to a population-based cohort study, nearly 80% of patients with PD develop cognitive dysfunction [1]. Risk factors for the development of dementia in PD include the severity of motor symptoms, axial motor symptoms, advanced age and cognitive dysfunction at diagnosis [2]. Mild cognitive impairment (MCI) in Correspondence: P. H. Lee, Department of Neurology, Yonsei University College of Medicine, 250 Seongsanno, Seodaemun-gu, Seoul 120-752, South Korea (tel.: +82 2 2228 1608; fax: +82 2 393 0705; e-mail: [email protected]).
922
patients with PD has been accepted as a transition state between normal aging and dementia [3,4]. PDMCI has been suggested as a risk factor for dementia in PD; about 50% of PD-MCI patients converted to PD dementia (PDD) compared with only 20% of patients who had normal cognition [5,6]. In terms of contributing factors for ongoing cognitive decline, longitudinal studies have demonstrated that cognitive subdomains encompassing posterior cortical functions, such as semantic fluency and visuospatial function, are closely associated with the development of PDD [7,8]. Atrophy in the frontostriatal area and cholinergic structures associated with frontal executive dysfunction are also considered a predictor of PDD © 2014 The Author(s) European Journal of Neurology © 2014 EAN
WMHs in Parkinson’s disease
[9–11]. Additionally, changes in cerebral glucose metabolism in the posterior visual association cortex and posterior cingulate area were associated with a decline in cognitive performance [12]. Silent vascular pathology, known as white matter hyperintensities (WMHs), is associated with cognitive decline in normal aging [13]. Likewise, WMHs have been demonstrated as a risk factor for a transition to Alzheimer disease (AD) in patients with MCI [14]. In patients with PD, however, the relationship between cognitive dysfunction and WMHs is controversial. Dalaker et al. showed that the total volume and spatial distribution of WMHs in PD patients (with or without MCI) did not differ from those in controls [15]. Meanwhile, growing evidence suggests that the burden of WMHs may have a negative impact on cognitive performance and may therefore be a major risk factor for ongoing cognitive dysfunction in PDMCI and PDD [16,17]. Similarly, the magnetic resonance imaging (MRI) based Cholinergic Pathways Hyperintensities Scale (CHIPS), which is the visual rating scale developed based on immunohistochemical tracings of the cholinergic pathway, is also closely associated with cognitive status in PD [16]. However, those studies were cross-sectional in design and therefore had limited ability to discern whether the baseline burden of WMHs influenced subsequent cognitive performance. In the present study, the longitudinal effects of WMHs on cognitive dysfunction in PD patients who are cognitively normal (PD-CN) and those with PD-MCI were examined to evaluate whether the burden of WMHs acts as an independent predictor of cognitive status progression.
Patients and methods Subjects
This prospective cohort study enrolled 111 patients with PD who visited a university hospital from January 2008 to January 2013. The study was approved by the Yonsei University Severance Ethical Standards Committee on Human Experimentation. The basic demographic data for age, gender, disease variables and history of vascular risk factors were collected by interview or chart review. Vascular risk factors included history of hypertension, diabetes mellitus, hyperlipidemia, coronary artery obstructive disease, cerebrovascular accident and smoking history. PD was diagnosed according to the clinical diagnostic criteria of the UK PD Society Brain Bank [18]. Subjects who underwent MRI and neuropsychological tests to assess baseline cognitive status were included in the study, and all testing was completed within a 3-week © 2014 The Author(s) European Journal of Neurology © 2014 EAN
923
period. To estimate the change in cognition, all subjects underwent the same follow-up neuropsychological test about 24 months after the time of baseline cognitive evaluation and MRI. The Korean version of the Mini-Mental State Examination (K-MMSE) to screen for general cognitive status and subtests of the Seoul Neuropsychological Screening Battery (SNSB) to determine cognitive status were used. The SNSB covers attention, language, visuoconstructive function, verbal and visual memory, and frontal/executive function [19,20]. For these, the quantifiable tests comprised the digit span task (forward and backward), the Korean version of the Boston Naming Test (KBNT), the Rey Complex Figure Test (RCFT; copying, immediate and 20 min delayed recall, and recognition), the Seoul Verbal Learning Test (SVLT; three free recall trials of 12 words, a 20 min delayed recall trial for the 12 items and a recognition test), phonemic and semantic Controlled Oral Word Association Test (COWAT), a go-no-go test and contrasting programming, and the Stroop test (word and color reading of 112 items over a 2 min period). Age, sex and education specific norms for each test based on 447 normal subjects are available. The scores of these quantifiable cognitive tests were classified as abnormal when they were below the 16th percentiles of the norms for age, sex and education matched normal subjects. PD-CN was defined as no objective cognitive dysfunction. To diagnose PD-MCI, two neuropsychological tests were designated to represent each of the four domains except the language domain. Attention was tested with a digit span task and the Stroop color-word test, executive function was tested using the COWAT and a clock drawing test, memory was tested using SVLT and RCFT, visuospatial function was tested using the RCFT copy and a pentagon drawing test, and the language domain was tested using only K-BNT. Following the Movement Disorder Society Task Force guidelines [21], a diagnosis of MCI in patients with PD was made if at least two tests for each of the four domains except the language domain (level 2) or five domains (level 1) was abnormal. All PD-MCI patients had a subjective cognitive complaint, scores of K-MMSE above the 16th percentile for age and educational appropriate norm and also showed no evidence of abnormal activities of daily living. Furthermore, MCI was categorized into four subtypes depending on the presence of memory impairment and the number of cognitive domains impaired: amnestic MCI single domain, amnestic MCI multiple domain, non-amnestic MCI single domain or non-amnestic MCI multiple domain [22]. PDD was diagnosed according to the clinical diagnostic criteria for probable PDD [23]. An [18F]FP-CIT positron
924
M. K. Sunwoo et al.
emission tomography scan was performed on all subjects, all of whom exhibited decreased dopamine transporter uptake in the posterior putamen. Exclusion criteria included vascular parkinsonism [24], drug-induced parkinsonism, evidence of focal brain lesion on MRI or presence of other neurodegenerative disease that might account for dementia. Possible medical comorbidities were also excluded by laboratory tests including thyroid function tests, vitamin B12 and folic acid levels, and a screening test for syphilis. Imaging acquisition
All scans of patients were acquired using a Philips 3.0 T scanner (Philips Intera; Philips Medical System, Best, The Netherlands) with a SENSE head coil (SENSE factor = 2). T1-weighted and fluid-attenuated inversion recovery sequence (FLAIR) images were acquired with the following parameters : axial acquisition with a 224 9 256 matrix; 256 9 256 reconstructed matrix with 182 slices; 220 mm field of view; 0.98 9 0.98 9 1.2 mm3 voxels; TE 4.6 ms; TR 9.6 ms; flip angle 8˚; and slice gap 0 mm. Imaging analysis
White matter hyperintensity volume was quantified on FLAIR images using an automated method as previously described [25]. The WMH candidate region mask was first extracted using T1-weighted images. Two modalities were aligned using rigid body co-registration. Then segmentation of the WMH region was performed on the FLAIR images after applying the candidate mask. If the results contained false positive or negative, the threshold value was reselected through visual inspection by two raters. Finally, the total WMH volume was calculated in the native FLAIR space with consideration of voxel dimension. Axial sections from FLAIR images were also used to evaluate WMHs in the cholinergic pathway using the recently developed visual rating scale, CHIPS [16,26]. On four selected axial images, the severity of WMHs in the cholinergic pathways was rated on a three-point scale for 10 regions, identified with major anatomical landmarks: low and high external capsule, corona radiata, centrum semiovale. The CHIPS score was scored blindly (by SMK and HJH), and the intra- and interrater reliabilities expressed as correlation coefficients were 0.87 and 0.84 for CHIPS, respectively. Statistical analysis
Data are expressed as mean standard deviation (SD). Numerical and non-numerical demographic
variables of the patient groups were analyzed by chisquared and independent t tests, respectively. Forward logistic regression analysis was performed to assess the factors contributing to PD-MCI conversion and PD-CN conversion. Prior to this analysis, correlations between WMH volume and CHIPS score were evaluated using Pearson’s correlation; these two variables were considered separately for logistic regression analysis to avoid multicollinearity. Therefore, the study was constructed using two models. Model 1 included age, gender, parkinsonism duration, number of risk factors and WMHs. Model 2 included the CHIPS score instead of the WMH volume as the independent variable. Additionally, to evaluate the relationship between scores for WMHs and CHIPS and changes in cognitive performance (score on the first task – score on the second task), Pearson’s correlation coefficients were calculated after adjusting for age, MMSE and education duration. Statistical analyses were performed using the Statistical Package for the Social Sciences version 20.0 (SPSS Inc., Chicago, IL, USA).
Results Of 111 patients, 65 were diagnosed as PD-MCI and the other 46 patients as PD-CN based on the initial cognitive status. Mean disease duration of all subjects was 17.78 months. The 65 patients with PD-MCI were reclassified as PD-MCI converters (n = 22) or PD-MCI non-converters (n = 43) based on whether they were subsequently diagnosed with PDD during a mean follow-up period of 29.8 months. The demographic characteristics of the PD-MCI subjects are shown in Table 1. No significant differences in age, gender, education duration, parkinsonism duration, general cognitive status as measured by the K-MMSE or number of vascular risk factors were observed between PD-MCI converters and non-converters. PDMCI converters showed a greater change in K-MMSE scores during the follow-up period than did PD-MCI non-converters (6.60 vs. 0.40, P = 0.001). WMH volume in PD-MCI converters was significantly larger than in PD-MCI non-converters (14421.0 vs. 5180.44, P < 0.001). Similarly, CHIPS score was higher in PDMCI converters than in PD-MCI non-converters (22.59 vs. 11.23, P = 0.001). There was no difference in WMH burden with respect to PD-MCI subtypes (Table S1). Logistic regression analysis in patients with PD-MCI revealed that the WMH volume in model 1 (odds ratio 1.616; 95% confidence interval 1.126–2.317, P = 0.009) and the CHIPS score in model 2 (odds ratio 1.084; 95% confidence interval 1.022–1.150, P = 0.007) were independently associated with PDD conversion (Table 2). © 2014 The Author(s) European Journal of Neurology © 2014 EAN
925
WMHs in Parkinson’s disease
Table 1 Demographic characteristics between PD-MCI converters versus nonconverters
PD-MCI converters (n = 22) Age (years) Gender (number of men, %) Education duration (years) Parkinsonism duration (months) Interval neuropsychological test (months) K-MMSE K-MMSE difference Number of risk factors Hypertension, n (%) Diabetes mellitus, n (%) Hyperlipidemia, n (%) Cardiovascular disease, n (%) Smoking, n (%) WMH volume (mm3) CHIPS score
74.32 11 9.36 20.32 28.00 25.27 6.60 0.64 8 2 5 8 3 14421.00 22.59
PD-MCI non-converters (n = 43)
(6.99) (50.0) (4.92) (13.14) (13.14) (2.29) (4.07) (0.73) (36.4) (9.1) (22.7) (36.4) (13.6) (13228.20) (15.68)
71.51 23 8.86 16.45 30.72 26.30 0.40 0.86 21 8 11 13 7 5180.44 11.23
(6.10) (53.4) (5.20) (16.31) (9.55) (2.86) (1.95) (0.74) (48.8) (18.6) (25.6) (30.2) (16.3) (6483.31) (9.83)
P value NS NS NS NS NS NS 0.001 NS NS NS NS NS NS
doi:10.1111/ene.12412
The burden of white matter hyperintensities is a predictor of progressive mild cognitive impairment in patients with Parkinson’s disease M. K. Sunwooa, S. Jeonb, J. H. Hama, J. Y. Hongc, J. E. Leea, J.-M. Leeb, Y. H. Sohna and P. H. Leea,d a
Department of Neurology, Yonsei University College of Medicine, Seoul; bDepartment of Biomedical Engineering, Hanyang University,
Seoul; cDepartment of Neurology, Yonsei University Wonju College of Medicine, Wonju; and dSeverance Biomedical Science Institute, Seoul, Korea
Keywords:
dementia converters, Parkinson’s disease, white matter hyperintensities Received 11 November 2013 Accepted 7 February 2014
Background and purpose: To evaluate whether white matter hyperintensities (WMHs) may act as an independent predictor for progression of cognitive status, the authors analyzed the longitudinal effects of WMHs on cognitive dysfunction in non-demented patients with Parkinson’s disease (PD). Methods: A total of 111 patients with PD were enrolled, including subjects with mild cognitive impairment (MCI, n = 65) and cognitively normal subjects (CN, n = 46). These individuals were classified as MCI converters (n = 22) or MCI non-converters (n = 43) and CN converters (n = 18) or CN non-converters (n = 28) based on whether they were subsequently diagnosed with PD dementia or PD-MCI during a minimum 24-month follow-up. The WMH burden and the Cholinergic Pathway Hyperintensities Scale (CHIPS) and their relationships to longitudinal changes in cognitive performance were examined. Results: PD-MCI converters had larger WMH volume (14421.0 vs. 5180.4, P < 0.001) and higher CHIPS score (22.6 vs. 11.2, P = 0.001) compared with PDMCI non-converters. Logistic regression analysis revealed in patients with PD-MCI that WMH volume (odds ratio 1.616, P = 0.009) and CHIPS score (odds ratio 1.084, P = 0.007) were independently associated with PD dementia conversion. However, WMH volume and CHIPS score did not differ between PD-CN converters and PD-CN non-converters. In patients with PD-MCI, both WMH volume and CHIPS score were closely associated with longitudinal decline in general cognition, semantic fluency and Stroop test scores. Conclusions: The present study demonstrates that WMH burden is a significant predictor of conversion from PD-MCI to PD dementia and is related to ongoing decline in frontal-lobe-based cognitive performance.
Introduction Cognitive impairment is one of most disabling nonmotor symptoms of Parkinson’s disease (PD). According to a population-based cohort study, nearly 80% of patients with PD develop cognitive dysfunction [1]. Risk factors for the development of dementia in PD include the severity of motor symptoms, axial motor symptoms, advanced age and cognitive dysfunction at diagnosis [2]. Mild cognitive impairment (MCI) in Correspondence: P. H. Lee, Department of Neurology, Yonsei University College of Medicine, 250 Seongsanno, Seodaemun-gu, Seoul 120-752, South Korea (tel.: +82 2 2228 1608; fax: +82 2 393 0705; e-mail: [email protected]).
922
patients with PD has been accepted as a transition state between normal aging and dementia [3,4]. PDMCI has been suggested as a risk factor for dementia in PD; about 50% of PD-MCI patients converted to PD dementia (PDD) compared with only 20% of patients who had normal cognition [5,6]. In terms of contributing factors for ongoing cognitive decline, longitudinal studies have demonstrated that cognitive subdomains encompassing posterior cortical functions, such as semantic fluency and visuospatial function, are closely associated with the development of PDD [7,8]. Atrophy in the frontostriatal area and cholinergic structures associated with frontal executive dysfunction are also considered a predictor of PDD © 2014 The Author(s) European Journal of Neurology © 2014 EAN
WMHs in Parkinson’s disease
[9–11]. Additionally, changes in cerebral glucose metabolism in the posterior visual association cortex and posterior cingulate area were associated with a decline in cognitive performance [12]. Silent vascular pathology, known as white matter hyperintensities (WMHs), is associated with cognitive decline in normal aging [13]. Likewise, WMHs have been demonstrated as a risk factor for a transition to Alzheimer disease (AD) in patients with MCI [14]. In patients with PD, however, the relationship between cognitive dysfunction and WMHs is controversial. Dalaker et al. showed that the total volume and spatial distribution of WMHs in PD patients (with or without MCI) did not differ from those in controls [15]. Meanwhile, growing evidence suggests that the burden of WMHs may have a negative impact on cognitive performance and may therefore be a major risk factor for ongoing cognitive dysfunction in PDMCI and PDD [16,17]. Similarly, the magnetic resonance imaging (MRI) based Cholinergic Pathways Hyperintensities Scale (CHIPS), which is the visual rating scale developed based on immunohistochemical tracings of the cholinergic pathway, is also closely associated with cognitive status in PD [16]. However, those studies were cross-sectional in design and therefore had limited ability to discern whether the baseline burden of WMHs influenced subsequent cognitive performance. In the present study, the longitudinal effects of WMHs on cognitive dysfunction in PD patients who are cognitively normal (PD-CN) and those with PD-MCI were examined to evaluate whether the burden of WMHs acts as an independent predictor of cognitive status progression.
Patients and methods Subjects
This prospective cohort study enrolled 111 patients with PD who visited a university hospital from January 2008 to January 2013. The study was approved by the Yonsei University Severance Ethical Standards Committee on Human Experimentation. The basic demographic data for age, gender, disease variables and history of vascular risk factors were collected by interview or chart review. Vascular risk factors included history of hypertension, diabetes mellitus, hyperlipidemia, coronary artery obstructive disease, cerebrovascular accident and smoking history. PD was diagnosed according to the clinical diagnostic criteria of the UK PD Society Brain Bank [18]. Subjects who underwent MRI and neuropsychological tests to assess baseline cognitive status were included in the study, and all testing was completed within a 3-week © 2014 The Author(s) European Journal of Neurology © 2014 EAN
923
period. To estimate the change in cognition, all subjects underwent the same follow-up neuropsychological test about 24 months after the time of baseline cognitive evaluation and MRI. The Korean version of the Mini-Mental State Examination (K-MMSE) to screen for general cognitive status and subtests of the Seoul Neuropsychological Screening Battery (SNSB) to determine cognitive status were used. The SNSB covers attention, language, visuoconstructive function, verbal and visual memory, and frontal/executive function [19,20]. For these, the quantifiable tests comprised the digit span task (forward and backward), the Korean version of the Boston Naming Test (KBNT), the Rey Complex Figure Test (RCFT; copying, immediate and 20 min delayed recall, and recognition), the Seoul Verbal Learning Test (SVLT; three free recall trials of 12 words, a 20 min delayed recall trial for the 12 items and a recognition test), phonemic and semantic Controlled Oral Word Association Test (COWAT), a go-no-go test and contrasting programming, and the Stroop test (word and color reading of 112 items over a 2 min period). Age, sex and education specific norms for each test based on 447 normal subjects are available. The scores of these quantifiable cognitive tests were classified as abnormal when they were below the 16th percentiles of the norms for age, sex and education matched normal subjects. PD-CN was defined as no objective cognitive dysfunction. To diagnose PD-MCI, two neuropsychological tests were designated to represent each of the four domains except the language domain. Attention was tested with a digit span task and the Stroop color-word test, executive function was tested using the COWAT and a clock drawing test, memory was tested using SVLT and RCFT, visuospatial function was tested using the RCFT copy and a pentagon drawing test, and the language domain was tested using only K-BNT. Following the Movement Disorder Society Task Force guidelines [21], a diagnosis of MCI in patients with PD was made if at least two tests for each of the four domains except the language domain (level 2) or five domains (level 1) was abnormal. All PD-MCI patients had a subjective cognitive complaint, scores of K-MMSE above the 16th percentile for age and educational appropriate norm and also showed no evidence of abnormal activities of daily living. Furthermore, MCI was categorized into four subtypes depending on the presence of memory impairment and the number of cognitive domains impaired: amnestic MCI single domain, amnestic MCI multiple domain, non-amnestic MCI single domain or non-amnestic MCI multiple domain [22]. PDD was diagnosed according to the clinical diagnostic criteria for probable PDD [23]. An [18F]FP-CIT positron
924
M. K. Sunwoo et al.
emission tomography scan was performed on all subjects, all of whom exhibited decreased dopamine transporter uptake in the posterior putamen. Exclusion criteria included vascular parkinsonism [24], drug-induced parkinsonism, evidence of focal brain lesion on MRI or presence of other neurodegenerative disease that might account for dementia. Possible medical comorbidities were also excluded by laboratory tests including thyroid function tests, vitamin B12 and folic acid levels, and a screening test for syphilis. Imaging acquisition
All scans of patients were acquired using a Philips 3.0 T scanner (Philips Intera; Philips Medical System, Best, The Netherlands) with a SENSE head coil (SENSE factor = 2). T1-weighted and fluid-attenuated inversion recovery sequence (FLAIR) images were acquired with the following parameters : axial acquisition with a 224 9 256 matrix; 256 9 256 reconstructed matrix with 182 slices; 220 mm field of view; 0.98 9 0.98 9 1.2 mm3 voxels; TE 4.6 ms; TR 9.6 ms; flip angle 8˚; and slice gap 0 mm. Imaging analysis
White matter hyperintensity volume was quantified on FLAIR images using an automated method as previously described [25]. The WMH candidate region mask was first extracted using T1-weighted images. Two modalities were aligned using rigid body co-registration. Then segmentation of the WMH region was performed on the FLAIR images after applying the candidate mask. If the results contained false positive or negative, the threshold value was reselected through visual inspection by two raters. Finally, the total WMH volume was calculated in the native FLAIR space with consideration of voxel dimension. Axial sections from FLAIR images were also used to evaluate WMHs in the cholinergic pathway using the recently developed visual rating scale, CHIPS [16,26]. On four selected axial images, the severity of WMHs in the cholinergic pathways was rated on a three-point scale for 10 regions, identified with major anatomical landmarks: low and high external capsule, corona radiata, centrum semiovale. The CHIPS score was scored blindly (by SMK and HJH), and the intra- and interrater reliabilities expressed as correlation coefficients were 0.87 and 0.84 for CHIPS, respectively. Statistical analysis
Data are expressed as mean standard deviation (SD). Numerical and non-numerical demographic
variables of the patient groups were analyzed by chisquared and independent t tests, respectively. Forward logistic regression analysis was performed to assess the factors contributing to PD-MCI conversion and PD-CN conversion. Prior to this analysis, correlations between WMH volume and CHIPS score were evaluated using Pearson’s correlation; these two variables were considered separately for logistic regression analysis to avoid multicollinearity. Therefore, the study was constructed using two models. Model 1 included age, gender, parkinsonism duration, number of risk factors and WMHs. Model 2 included the CHIPS score instead of the WMH volume as the independent variable. Additionally, to evaluate the relationship between scores for WMHs and CHIPS and changes in cognitive performance (score on the first task – score on the second task), Pearson’s correlation coefficients were calculated after adjusting for age, MMSE and education duration. Statistical analyses were performed using the Statistical Package for the Social Sciences version 20.0 (SPSS Inc., Chicago, IL, USA).
Results Of 111 patients, 65 were diagnosed as PD-MCI and the other 46 patients as PD-CN based on the initial cognitive status. Mean disease duration of all subjects was 17.78 months. The 65 patients with PD-MCI were reclassified as PD-MCI converters (n = 22) or PD-MCI non-converters (n = 43) based on whether they were subsequently diagnosed with PDD during a mean follow-up period of 29.8 months. The demographic characteristics of the PD-MCI subjects are shown in Table 1. No significant differences in age, gender, education duration, parkinsonism duration, general cognitive status as measured by the K-MMSE or number of vascular risk factors were observed between PD-MCI converters and non-converters. PDMCI converters showed a greater change in K-MMSE scores during the follow-up period than did PD-MCI non-converters (6.60 vs. 0.40, P = 0.001). WMH volume in PD-MCI converters was significantly larger than in PD-MCI non-converters (14421.0 vs. 5180.44, P < 0.001). Similarly, CHIPS score was higher in PDMCI converters than in PD-MCI non-converters (22.59 vs. 11.23, P = 0.001). There was no difference in WMH burden with respect to PD-MCI subtypes (Table S1). Logistic regression analysis in patients with PD-MCI revealed that the WMH volume in model 1 (odds ratio 1.616; 95% confidence interval 1.126–2.317, P = 0.009) and the CHIPS score in model 2 (odds ratio 1.084; 95% confidence interval 1.022–1.150, P = 0.007) were independently associated with PDD conversion (Table 2). © 2014 The Author(s) European Journal of Neurology © 2014 EAN
925
WMHs in Parkinson’s disease
Table 1 Demographic characteristics between PD-MCI converters versus nonconverters
PD-MCI converters (n = 22) Age (years) Gender (number of men, %) Education duration (years) Parkinsonism duration (months) Interval neuropsychological test (months) K-MMSE K-MMSE difference Number of risk factors Hypertension, n (%) Diabetes mellitus, n (%) Hyperlipidemia, n (%) Cardiovascular disease, n (%) Smoking, n (%) WMH volume (mm3) CHIPS score
74.32 11 9.36 20.32 28.00 25.27 6.60 0.64 8 2 5 8 3 14421.00 22.59
PD-MCI non-converters (n = 43)
(6.99) (50.0) (4.92) (13.14) (13.14) (2.29) (4.07) (0.73) (36.4) (9.1) (22.7) (36.4) (13.6) (13228.20) (15.68)
71.51 23 8.86 16.45 30.72 26.30 0.40 0.86 21 8 11 13 7 5180.44 11.23
(6.10) (53.4) (5.20) (16.31) (9.55) (2.86) (1.95) (0.74) (48.8) (18.6) (25.6) (30.2) (16.3) (6483.31) (9.83)
P value NS NS NS NS NS NS 0.001 NS NS NS NS NS NS
