Neurobiology of Aging 36 (2015) 1743e1750
Contents lists available at ScienceDirect
Neurobiology of Aging journal homepage: www.elsevier.com/locate/neuaging
Progressive cortical thinning and subcortical atrophy in dementia with Lewy bodies and Alzheimer’s disease Elijah Mak a, Li Su a, Guy B. Williams b, Rosie Watson c, d, Michael J. Firbank d, Andrew M. Blamire e, John T. O’Brien a, * a
Department of Psychiatry, School of Clinical Medicine, University of Cambridge, Cambridge, UK Department of Clinical Neurosciences, Wolfson Brain Imaging Centre, University of Cambridge, UK Department of Aged Care, The Royal Melbourne Hospital, Melbourne, Australia d Institute of Neuroscience, Newcastle University, Campus for Ageing and Vitality, UK e Newcastle Magnetic Resonance Centre, Institute of Cellular Medicine, Newcastle University, Newcastle, UK b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 13 October 2014 Received in revised form 16 December 2014 Accepted 30 December 2014 Available online 8 January 2015
Patterns of progressive cortical thinning in dementia with Lewy bodies (DLB) remain poorly understood. We examined spatiotemporal patterns of cortical thinning and subcortical atrophy over 12 months in DLB (n ¼ 13), compared with Alzheimer’s disease (AD) (n ¼ 23) and healthy control subjects (HC) (n ¼ 33). Rates of temporal thinning in DLB were relatively preserved compared with AD. Volumetric analyses subcortical changes revealed that the AD group demonstrated significantly increased hippocampal atrophy (5.8%) relative to the HC (1.7%; p < 0.001) and DLB groups (2.5%, p ¼ 0.006). Significant lateral ventricular expansion was also observed in AD (8.9%) compared with HC (4.3%; p < 0.001) and DLB (4.7%; p ¼ 0.008) at trend level. There was no significant difference in subcortical atrophy and ventricular expansion between DLB and HC. In the DLB group, increased rates of cortical thinning in the frontal and parietal regions were significantly correlated with decline in global cognition (Mini-Mental State Examination) and motor deterioration (Unified Parkinson’s Disease Rating Scale 3), respectively. Overall, AD and DLB are characterized by different spatiotemporal patterns of cortical thinning over time. Our findings warrant further consideration of longitudinal cortical thinning as a potential imaging marker to differentiate DLB from AD. Crown Copyright Ó 2015 Published by Elsevier Inc. All rights reserved.
Keywords: Dementia Alzheimer’s disease Lewy bodies MRI Neuroimaging Atrophy
1. Introduction Dementia with Lewy bodies (DLB) is the second leading cause of degenerative dementia in older people after Alzheimer’s disease (AD), accounting for up to 15% of cases confirmed at autopsy (Geser et al., 2005; McKeith et al., 1996; Vann Jones and O’Brien, 2014). Because low sensitivity for the diagnosis for DLB remains a problem (Nelson et al., 2010), there is a need for the development of reliable imaging markers to help distinguish DLB from other subtypes of dementia. Cortical thickness is increasingly recognized as a more precise parameter of age-associated decline in gray matter compared with the voxel-based morphometry technique (Cardinale et al., 2014; Hutton et al., 2009). In a previous study, we found a greater extent of cortical thinning in the AD group affecting predominantly temporoparietal areas, whereas DLB was characterized with cortical thinning in posterior structures (Watson et al., 2014). This finding is consistent with a growing literature of reduced global * Corresponding author at: Department of Psychiatry, School of Clinical Medicine, University of Cambridge, Box 189, Level E4 Cambridge Biomedical Campus, Cambridge, CB2 0SP, UK. Tel.: þ44 01223 760682; fax: þ44 01223 336968. E-mail address: [email protected] (J.T. O’Brien).
atrophy in DLB compared with AD (Mak et al., 2014), whereas the preservation of the medial temporal lobe in DLB has been incorporated as a supportive feature in the revised criteria for the diagnosis of DLB (McKeith et al., 2005). As with our previous investigation, most of the imaging studies in DLB have been cross-sectional, and no study to date has investigated the longitudinal progression of cortical thinning in DLB. To address this gap in the present literature, our aim in this study was to compare the progression of cortical thickness over a 12-month period in AD and DLB and similarly aged healthy control subjects (HC). Based on earlier cross-sectional findings (Mak et al., 2014), we hypothesized that DLB would have significantly lower rates of cortical thinning compared with AD, particularly in the temporal lobe. 2. Methods 2.1. Subjects, assessment, and diagnosis Thirty-six subjects with probable AD (McKhann et al., 1984) and 35 with probable DLB (McKeith et al., 2005) were recruited from a community dwelling population of patients referred to local old age
0197-4580/$ e see front matter Crown Copyright Ó 2015 Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.neurobiolaging.2014.12.038
1744
E. Mak et al. / Neurobiology of Aging 36 (2015) 1743e1750
psychiatry, geriatric medicine, or neurology services as previously described (Watson et al., 2012). Subjects underwent clinical and neuropsychological evaluations at baseline and follow-up at 1 year. Thirty-five similarly aged control subjects were recruited from relatives and friends of subjects with dementia or volunteered via advertisements in local community newsletters. For the purpose of the present study, we included only subjects with magnetic resonance imaging (MRI) assessments from both baseline and 1-year follow-up. Of the 36 AD subjects, 25 were included after 11 were unable to participate in the follow-up assessment. Of the 35 DLB subjects, 14 were included after 12 declined to participate as they or their caregivers felt they were too unwell, and 9 subjects had died. However, there were no significant differences in age, gender, educational level, Unified Parkinson’s Disease Rating Scale Part III, Neuropsychiatric Inventory (NPI), or cognitive scores between the DLB subjects who dropped out and the DLB subjects who were included in the present study (Table 1). Of the 35 HC subjects, 33 were included in the present analyses after 2 declined to participate because of other reasons. The research was approved by the local ethics committee. All subjects or, where appropriate, their nearest relative, provided written informed consent. At baseline and follow-up assessments, global cognitive measures included the Cambridge Cognitive Examination (CAMCOG) (Huppert et al., 1995), which incorporates the Mini-Mental State Examination (MMSE) (Folstein et al., 1975) in addition to a number of subscales assessing
domains including orientation, language, memory, attention, praxis, calculation, abstract thinking, and perception. Visuospatial memory was assessed with the Brief Visuospatial Memory Test (BVMT) (Benedict et al., 1996). Motor parkinsonism was evaluated with the Unified Parkinson’s Disease Rating Scale Part III (Movement Disorder Society Task Force on Rating Scales for Parkinson’s Disease, 2003). For subjects with dementia, neuropsychiatric features were examined with the NPI (Cummings et al., 1994), and cognitive fluctuations were assessed with the cognitive fluctuation scale (Walker et al., 2000). Functional ability was assessed with the Bristol Activities of Daily Living Scale (Bucks et al., 1996). 2.2. MRI acquisition Subjects underwent both baseline and repeat MR imaging with a 12-month interval. At each time point, subjects underwent T1-weighted MR scanning on the same 3T MRI system using an 8-channel head coil (Intera Achieva scanner, Philips Medical Systems, Eindhoven, the Netherlands). The sequence was a standard T1-weighted volumetric sequence covering the whole brain (3D magnetisation prepared rapid acquisition gradient echo, sagittal acquisition, 1 mm isotropic resolution, and matrix size of 240 [anterior-posterior] 240 [superior-inferior] 180 [right left]; repetition time ¼ 9.6 ms; echo time ¼ 4.6 ms; flip angle ¼ 8 ; SENSitivity Encoding [SENSE] factor ¼ 2).
Table 1 Demographics, clinical, and neuropsychological measures
N Gender (M:F) Age (y) Education (y) Disease duration (mo) ChEI (%) BADL UPDRS Baseline Follow-up Change NPI total Baseline Follow-up Change CogFluct Baseline Follow-up Change MMSE Baseline Follow-up Change CAMCOG Baseline Follow-up Change BVMT total Baseline Follow-up Change Interscan interval (d)
HC
DLB
AD
33 20:13 76.7 5.3 11.8 2.6
13 12:1 77.0 8.3 10.6 1.9 52.2 20.4 76.92
23 13:10 76.5 5.4 11.3 3.8 51.8 26.5 91.30
17.41 10.17
14.09 7.87
c2 ¼ 5.28, 0.07a F2,68 ¼ 0.03, p ¼ 0.97b p ¼ 0.16c p ¼ 0.96d c2 ¼ 1.44, p ¼ 0.23a p ¼ 0.23e
27.7 8.0 32.6 13.2 4.9 8.4
4.7 4.1 5.7 4.8 1.0 2.4
p < 0.01c p < 0.01c p ¼ 0.09e
21.1 16.8 24.8 14.9 3.7 17.4
19.4 12.4 19.7 15.0 0.3 11.8
p ¼ 0.81e p ¼ 0.30e p ¼ 0.50d
8.1 3.4 7.8 5.4 0.1 4.4
2.8 3.6 1.8 3.3 1.0 4.6
p < 0.01e p < 0.01e p ¼ 0.52d
29.2 0.9 29.2 0.9 0.1 1.0
21.3 6.3 19.8 5.8 2.6 2.9
20.9 4.0 18.8 4.2 2.0 3.2
p ¼ 0.80d p ¼ 0.60d p ¼ 0.63d
97.8 3.3 98.6 2.8 0.8 2.50
69.9 18.0 66.8 17.9 5.8 10.8
69.2 11.3 62.2 14.4 7.0 10.2
p ¼ 0.90d p ¼ 0.42d p ¼ 0.74d
18.9 6.7 21.9 5.8 3.0 5.3 370.9 13.3
6.23 6.7 7.8 7.7 0.1 4.5 379.1 18.8
4.2 2.7 5.2 2.6 1.0 2.7 379.6 17.8
p p p p
1.9 1.8 2.1 2.0 0.2 2.0
p-value
¼ ¼ ¼ ¼
0.51e 0.51e 0.48d 0.21c
Values expressed as mean 1 standard deviation. Key: AD, Alzheimer’s disease; BADLS, Bristol Activities of Daily Living Scale; CAMCOG, Cambridge Cognitive Examination; CogFluct, Cognitive Fluctuation Scale; DLB, dementia with Lewy bodies; HC, Healthy control; MMSE, Mini-Mental State Examination; NPI total, Neuropsychiatry Inventory Part III; UPDRS III, Unified Parkinson’s Disease Rating Scale. a c2dDLB, AD, and control subjects. b Analysis of variancedHC, DLB, and AD. c Kruskal-Wallis test. d Student t testdAD and DLB. e Wilcoxon rank-sum testdAD and DLB.
E. Mak et al. / Neurobiology of Aging 36 (2015) 1743e1750 Table 2 Demographics and clinical characteristics of DLB subjects
N Gender (M:F) Age (y) Education (y) UPDRS NPI total CogFluct MMSE CAMCOG
DLBdropped-out
DLBreturned
21 14:7 79.1 11.0 25.1 21.4 4.6 19.7 66.2
14 13:1 77.2 10.5 27.2 21.5 8.4 21.2 69.9
6.2 3.0 12.3 18.1 3.3 4.7 14.0
p-value
c2 ¼ 3.3, 0.07a 8.0 1.9 7.9 16.1 3.4 6.0 17.3
p p p p p p p
¼ ¼ ¼ ¼ < ¼ ¼
0.43b 0.69c 0.58b 0.70c 0.01b 0.42b 0.49b
Values expressed as mean 1 standard deviation. Key: CAMCOG, Cambridge Cognitive Examination; CogFluct, Cognitive Fluctuation Scale; DLB, dementia with Lewy bodies; MMSE, Mini-Mental State Examination; NPI total, Neuropsychiatry Inventory Part III; UPDRS III, Unified Parkinson’s Disease Rating Scale. a c2 test. b Wilcoxon rank-sum test. c Student t test.
2.3. Image analysis Cortical reconstruction and volumetric segmentation of MRI scans were processed on the same workstation using the FreeSurfer 5.3 image analysis suite (http://surfer.nmr.mgh.harvard.edu/). The technical details are described previously (Fischl and Dale, 2000; Fischl et al., 1999). The initial processing of T1-weighted MRI images, for each subject and each time point, includes the following steps: removal of non-brain tissue, automated Talairach transformation, segmentation of the subcortical white matter, and deep gray matter volumetric structures, intensity normalization, tessellation of the gray matter and white matter boundary, automated topology correction, and surface deformation to optimally place the gray matter and white matter and gray matter and cerebrospinal fluid boundaries. The cortical thickness is calculated as the closest distance from the gray and/or white matter boundary to the gray and cerebrospinal fluid boundary at each vertex. All surface models in our study were inspected for accuracy, and manual corrections were performed in the event of tissue misclassification and/or white matter errors. However, 3 subjects (2 AD, 1 DLB) still had extensive pial and/or white matter surface errors and were excluded. The data set for all subsequent analyses comprised 33 HC, 23 AD, and 13 DLB. Subsequently, for the longitudinal processing, an unbiased within-subject template space (Reuter and Fischl, 2011) was created using robust inverse consistent registration (Reuter et al., 2010). Several processing steps, such as skull stripping, Talairach transformations, atlas registration, as well as spherical surface maps, and parcellations were then initialized with common information from the within-subject template, significantly increasing reliability and statistical power (Reuter et al., 2012). The cortical thickness maps were smoothed using a 15-mm full width at half maximum Gaussian kernel to reduce local variations in the measurements. In addition, the following volumetric measures at both timepoints were automatically obtained using FreeSurfer: total intracranial volume, lateral ventricles, and 7 subcortical structures including the thalamus, caudate, putamen, pallidum, hippocampus, amygdala, and the nucleus accumbens. 2.4. Statistical analyses 2.4.1. Demographic and clinical measures Statistical analyses were performed with the STATA 13 (http:// www.stata.com/) software. The distribution of continuous variables was tested for normality using the Skewness-Kurtosis test and visual inspection of histograms. Parametric data were assessed
1745
using either t tests or analysis of variance for continuous variables. For nonparametric data, Kruskal-Wallis test was used. c2 tests were used to examine differences between categorical measures. For each test statistic, a 2-tailed probability value of
Contents lists available at ScienceDirect
Neurobiology of Aging journal homepage: www.elsevier.com/locate/neuaging
Progressive cortical thinning and subcortical atrophy in dementia with Lewy bodies and Alzheimer’s disease Elijah Mak a, Li Su a, Guy B. Williams b, Rosie Watson c, d, Michael J. Firbank d, Andrew M. Blamire e, John T. O’Brien a, * a
Department of Psychiatry, School of Clinical Medicine, University of Cambridge, Cambridge, UK Department of Clinical Neurosciences, Wolfson Brain Imaging Centre, University of Cambridge, UK Department of Aged Care, The Royal Melbourne Hospital, Melbourne, Australia d Institute of Neuroscience, Newcastle University, Campus for Ageing and Vitality, UK e Newcastle Magnetic Resonance Centre, Institute of Cellular Medicine, Newcastle University, Newcastle, UK b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 13 October 2014 Received in revised form 16 December 2014 Accepted 30 December 2014 Available online 8 January 2015
Patterns of progressive cortical thinning in dementia with Lewy bodies (DLB) remain poorly understood. We examined spatiotemporal patterns of cortical thinning and subcortical atrophy over 12 months in DLB (n ¼ 13), compared with Alzheimer’s disease (AD) (n ¼ 23) and healthy control subjects (HC) (n ¼ 33). Rates of temporal thinning in DLB were relatively preserved compared with AD. Volumetric analyses subcortical changes revealed that the AD group demonstrated significantly increased hippocampal atrophy (5.8%) relative to the HC (1.7%; p < 0.001) and DLB groups (2.5%, p ¼ 0.006). Significant lateral ventricular expansion was also observed in AD (8.9%) compared with HC (4.3%; p < 0.001) and DLB (4.7%; p ¼ 0.008) at trend level. There was no significant difference in subcortical atrophy and ventricular expansion between DLB and HC. In the DLB group, increased rates of cortical thinning in the frontal and parietal regions were significantly correlated with decline in global cognition (Mini-Mental State Examination) and motor deterioration (Unified Parkinson’s Disease Rating Scale 3), respectively. Overall, AD and DLB are characterized by different spatiotemporal patterns of cortical thinning over time. Our findings warrant further consideration of longitudinal cortical thinning as a potential imaging marker to differentiate DLB from AD. Crown Copyright Ó 2015 Published by Elsevier Inc. All rights reserved.
Keywords: Dementia Alzheimer’s disease Lewy bodies MRI Neuroimaging Atrophy
1. Introduction Dementia with Lewy bodies (DLB) is the second leading cause of degenerative dementia in older people after Alzheimer’s disease (AD), accounting for up to 15% of cases confirmed at autopsy (Geser et al., 2005; McKeith et al., 1996; Vann Jones and O’Brien, 2014). Because low sensitivity for the diagnosis for DLB remains a problem (Nelson et al., 2010), there is a need for the development of reliable imaging markers to help distinguish DLB from other subtypes of dementia. Cortical thickness is increasingly recognized as a more precise parameter of age-associated decline in gray matter compared with the voxel-based morphometry technique (Cardinale et al., 2014; Hutton et al., 2009). In a previous study, we found a greater extent of cortical thinning in the AD group affecting predominantly temporoparietal areas, whereas DLB was characterized with cortical thinning in posterior structures (Watson et al., 2014). This finding is consistent with a growing literature of reduced global * Corresponding author at: Department of Psychiatry, School of Clinical Medicine, University of Cambridge, Box 189, Level E4 Cambridge Biomedical Campus, Cambridge, CB2 0SP, UK. Tel.: þ44 01223 760682; fax: þ44 01223 336968. E-mail address: [email protected] (J.T. O’Brien).
atrophy in DLB compared with AD (Mak et al., 2014), whereas the preservation of the medial temporal lobe in DLB has been incorporated as a supportive feature in the revised criteria for the diagnosis of DLB (McKeith et al., 2005). As with our previous investigation, most of the imaging studies in DLB have been cross-sectional, and no study to date has investigated the longitudinal progression of cortical thinning in DLB. To address this gap in the present literature, our aim in this study was to compare the progression of cortical thickness over a 12-month period in AD and DLB and similarly aged healthy control subjects (HC). Based on earlier cross-sectional findings (Mak et al., 2014), we hypothesized that DLB would have significantly lower rates of cortical thinning compared with AD, particularly in the temporal lobe. 2. Methods 2.1. Subjects, assessment, and diagnosis Thirty-six subjects with probable AD (McKhann et al., 1984) and 35 with probable DLB (McKeith et al., 2005) were recruited from a community dwelling population of patients referred to local old age
0197-4580/$ e see front matter Crown Copyright Ó 2015 Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.neurobiolaging.2014.12.038
1744
E. Mak et al. / Neurobiology of Aging 36 (2015) 1743e1750
psychiatry, geriatric medicine, or neurology services as previously described (Watson et al., 2012). Subjects underwent clinical and neuropsychological evaluations at baseline and follow-up at 1 year. Thirty-five similarly aged control subjects were recruited from relatives and friends of subjects with dementia or volunteered via advertisements in local community newsletters. For the purpose of the present study, we included only subjects with magnetic resonance imaging (MRI) assessments from both baseline and 1-year follow-up. Of the 36 AD subjects, 25 were included after 11 were unable to participate in the follow-up assessment. Of the 35 DLB subjects, 14 were included after 12 declined to participate as they or their caregivers felt they were too unwell, and 9 subjects had died. However, there were no significant differences in age, gender, educational level, Unified Parkinson’s Disease Rating Scale Part III, Neuropsychiatric Inventory (NPI), or cognitive scores between the DLB subjects who dropped out and the DLB subjects who were included in the present study (Table 1). Of the 35 HC subjects, 33 were included in the present analyses after 2 declined to participate because of other reasons. The research was approved by the local ethics committee. All subjects or, where appropriate, their nearest relative, provided written informed consent. At baseline and follow-up assessments, global cognitive measures included the Cambridge Cognitive Examination (CAMCOG) (Huppert et al., 1995), which incorporates the Mini-Mental State Examination (MMSE) (Folstein et al., 1975) in addition to a number of subscales assessing
domains including orientation, language, memory, attention, praxis, calculation, abstract thinking, and perception. Visuospatial memory was assessed with the Brief Visuospatial Memory Test (BVMT) (Benedict et al., 1996). Motor parkinsonism was evaluated with the Unified Parkinson’s Disease Rating Scale Part III (Movement Disorder Society Task Force on Rating Scales for Parkinson’s Disease, 2003). For subjects with dementia, neuropsychiatric features were examined with the NPI (Cummings et al., 1994), and cognitive fluctuations were assessed with the cognitive fluctuation scale (Walker et al., 2000). Functional ability was assessed with the Bristol Activities of Daily Living Scale (Bucks et al., 1996). 2.2. MRI acquisition Subjects underwent both baseline and repeat MR imaging with a 12-month interval. At each time point, subjects underwent T1-weighted MR scanning on the same 3T MRI system using an 8-channel head coil (Intera Achieva scanner, Philips Medical Systems, Eindhoven, the Netherlands). The sequence was a standard T1-weighted volumetric sequence covering the whole brain (3D magnetisation prepared rapid acquisition gradient echo, sagittal acquisition, 1 mm isotropic resolution, and matrix size of 240 [anterior-posterior] 240 [superior-inferior] 180 [right left]; repetition time ¼ 9.6 ms; echo time ¼ 4.6 ms; flip angle ¼ 8 ; SENSitivity Encoding [SENSE] factor ¼ 2).
Table 1 Demographics, clinical, and neuropsychological measures
N Gender (M:F) Age (y) Education (y) Disease duration (mo) ChEI (%) BADL UPDRS Baseline Follow-up Change NPI total Baseline Follow-up Change CogFluct Baseline Follow-up Change MMSE Baseline Follow-up Change CAMCOG Baseline Follow-up Change BVMT total Baseline Follow-up Change Interscan interval (d)
HC
DLB
AD
33 20:13 76.7 5.3 11.8 2.6
13 12:1 77.0 8.3 10.6 1.9 52.2 20.4 76.92
23 13:10 76.5 5.4 11.3 3.8 51.8 26.5 91.30
17.41 10.17
14.09 7.87
c2 ¼ 5.28, 0.07a F2,68 ¼ 0.03, p ¼ 0.97b p ¼ 0.16c p ¼ 0.96d c2 ¼ 1.44, p ¼ 0.23a p ¼ 0.23e
27.7 8.0 32.6 13.2 4.9 8.4
4.7 4.1 5.7 4.8 1.0 2.4
p < 0.01c p < 0.01c p ¼ 0.09e
21.1 16.8 24.8 14.9 3.7 17.4
19.4 12.4 19.7 15.0 0.3 11.8
p ¼ 0.81e p ¼ 0.30e p ¼ 0.50d
8.1 3.4 7.8 5.4 0.1 4.4
2.8 3.6 1.8 3.3 1.0 4.6
p < 0.01e p < 0.01e p ¼ 0.52d
29.2 0.9 29.2 0.9 0.1 1.0
21.3 6.3 19.8 5.8 2.6 2.9
20.9 4.0 18.8 4.2 2.0 3.2
p ¼ 0.80d p ¼ 0.60d p ¼ 0.63d
97.8 3.3 98.6 2.8 0.8 2.50
69.9 18.0 66.8 17.9 5.8 10.8
69.2 11.3 62.2 14.4 7.0 10.2
p ¼ 0.90d p ¼ 0.42d p ¼ 0.74d
18.9 6.7 21.9 5.8 3.0 5.3 370.9 13.3
6.23 6.7 7.8 7.7 0.1 4.5 379.1 18.8
4.2 2.7 5.2 2.6 1.0 2.7 379.6 17.8
p p p p
1.9 1.8 2.1 2.0 0.2 2.0
p-value
¼ ¼ ¼ ¼
0.51e 0.51e 0.48d 0.21c
Values expressed as mean 1 standard deviation. Key: AD, Alzheimer’s disease; BADLS, Bristol Activities of Daily Living Scale; CAMCOG, Cambridge Cognitive Examination; CogFluct, Cognitive Fluctuation Scale; DLB, dementia with Lewy bodies; HC, Healthy control; MMSE, Mini-Mental State Examination; NPI total, Neuropsychiatry Inventory Part III; UPDRS III, Unified Parkinson’s Disease Rating Scale. a c2dDLB, AD, and control subjects. b Analysis of variancedHC, DLB, and AD. c Kruskal-Wallis test. d Student t testdAD and DLB. e Wilcoxon rank-sum testdAD and DLB.
E. Mak et al. / Neurobiology of Aging 36 (2015) 1743e1750 Table 2 Demographics and clinical characteristics of DLB subjects
N Gender (M:F) Age (y) Education (y) UPDRS NPI total CogFluct MMSE CAMCOG
DLBdropped-out
DLBreturned
21 14:7 79.1 11.0 25.1 21.4 4.6 19.7 66.2
14 13:1 77.2 10.5 27.2 21.5 8.4 21.2 69.9
6.2 3.0 12.3 18.1 3.3 4.7 14.0
p-value
c2 ¼ 3.3, 0.07a 8.0 1.9 7.9 16.1 3.4 6.0 17.3
p p p p p p p
¼ ¼ ¼ ¼ < ¼ ¼
0.43b 0.69c 0.58b 0.70c 0.01b 0.42b 0.49b
Values expressed as mean 1 standard deviation. Key: CAMCOG, Cambridge Cognitive Examination; CogFluct, Cognitive Fluctuation Scale; DLB, dementia with Lewy bodies; MMSE, Mini-Mental State Examination; NPI total, Neuropsychiatry Inventory Part III; UPDRS III, Unified Parkinson’s Disease Rating Scale. a c2 test. b Wilcoxon rank-sum test. c Student t test.
2.3. Image analysis Cortical reconstruction and volumetric segmentation of MRI scans were processed on the same workstation using the FreeSurfer 5.3 image analysis suite (http://surfer.nmr.mgh.harvard.edu/). The technical details are described previously (Fischl and Dale, 2000; Fischl et al., 1999). The initial processing of T1-weighted MRI images, for each subject and each time point, includes the following steps: removal of non-brain tissue, automated Talairach transformation, segmentation of the subcortical white matter, and deep gray matter volumetric structures, intensity normalization, tessellation of the gray matter and white matter boundary, automated topology correction, and surface deformation to optimally place the gray matter and white matter and gray matter and cerebrospinal fluid boundaries. The cortical thickness is calculated as the closest distance from the gray and/or white matter boundary to the gray and cerebrospinal fluid boundary at each vertex. All surface models in our study were inspected for accuracy, and manual corrections were performed in the event of tissue misclassification and/or white matter errors. However, 3 subjects (2 AD, 1 DLB) still had extensive pial and/or white matter surface errors and were excluded. The data set for all subsequent analyses comprised 33 HC, 23 AD, and 13 DLB. Subsequently, for the longitudinal processing, an unbiased within-subject template space (Reuter and Fischl, 2011) was created using robust inverse consistent registration (Reuter et al., 2010). Several processing steps, such as skull stripping, Talairach transformations, atlas registration, as well as spherical surface maps, and parcellations were then initialized with common information from the within-subject template, significantly increasing reliability and statistical power (Reuter et al., 2012). The cortical thickness maps were smoothed using a 15-mm full width at half maximum Gaussian kernel to reduce local variations in the measurements. In addition, the following volumetric measures at both timepoints were automatically obtained using FreeSurfer: total intracranial volume, lateral ventricles, and 7 subcortical structures including the thalamus, caudate, putamen, pallidum, hippocampus, amygdala, and the nucleus accumbens. 2.4. Statistical analyses 2.4.1. Demographic and clinical measures Statistical analyses were performed with the STATA 13 (http:// www.stata.com/) software. The distribution of continuous variables was tested for normality using the Skewness-Kurtosis test and visual inspection of histograms. Parametric data were assessed
1745
using either t tests or analysis of variance for continuous variables. For nonparametric data, Kruskal-Wallis test was used. c2 tests were used to examine differences between categorical measures. For each test statistic, a 2-tailed probability value of
