Comments
Description
Transcript
Original Contribution
Original Contribution Progression of White Matter Disease and Cortical Thinning Are Not Related in Older Community-Dwelling Subjects David Alexander Dickie, PhD; Sherif Karama, MD, PhD; Stuart J. Ritchie, PhD; Simon R. Cox, PhD; Eleni Sakka, MSc; Natalie A. Royle, PhD; Benjamin S. Aribisala, PhD; Maria Valdés Hernández, PhD; Susana Muñoz Maniega, PhD; Alison Pattie, BSc; Janie Corley, MA; John M. Starr, PhD/MD; Mark E. Bastin, DPhil; Alan C. Evans, PhD; Ian J. Deary, PhD/MD*; Joanna M. Wardlaw, MD* Downloaded from http://stroke.ahajournals.org/ by guest on March 28, 2017 Background and Purpose—We assessed cross-sectional and longitudinal relationships between whole brain white matter hyperintensity (WMH) volume and regional cortical thickness. Methods—We measured WMH volume and regional cortical thickness on magnetic resonance imaging at ≈73 and ≈76 years in 351 community-dwelling subjects from the Lothian Birth Cohort 1936. We used multiple linear regression to calculate cross-sectional and longitudinal associations between regional cortical thickness and WMH volume controlling for age, sex, Mini Mental State Examination, education, intelligence quotient at age 11, and vascular risk factors. Results—We found cross-sectional associations between WMH volume and cortical thickness within and surrounding the Sylvian fissure at 73 and 76 years (rho=−0.276, Q=0.004). However, we found no significant longitudinal associations between (1) baseline WMH volume and change in cortical thickness; (2) baseline cortical thickness and change in WMH volume; or (3) change in WMH volume and change in cortical thickness. Conclusions—Our results show that WMH volume and cortical thinning both worsen with age and are associated crosssectionally within and surrounding the Sylvian fissure. However, changes in WMH volume and cortical thinning from 73 to 76 years are not associated longitudinally in these relatively healthy older subjects. The underlying cause(s) of WMH growth and cortical thinning have yet to be fully determined. (Stroke. 2016;47:00-00. DOI: 10.1161/ STROKEAHA.115.011229.) Key Words: ageing ◼ brain ◼ cortex ◼ MRI ◼ white matter hyperintensities B rain white matter hyperintensity (WMH) growth and cortical thinning are commonly seen on magnetic resonance imaging (MRI) in community-dwelling older people.1–5 The incidence of these features is highly variable between individuals but those with the largest WMH volumes and/or thinnest cortices are at increased risk of stroke, dementia, and cognitive and physical impairment.6–8 Effective interventions are dependent on understanding the mechanisms of WMH growth and cortical thinning and whether one feature may be an underlying cause of the other. Cross-sectional studies have found associations between larger whole brain WMH volume and reduced gray matter (GM) volume, density, and thickness.3,5,9–12 Those with larger WMH volumes generally had relatively reduced cortical thickness3 and density5,11 in frontotemporal and inferior parietal regions. Others have found similar cross-sectional patterns of negative associations between whole brain WMH volume and regional GM volume in the default mode network (which includes medial temporal lobe structures, the inferior parietal lobe, and cuneus) using a region of interest analysis.10 Larger WMH volume in small vessel disease patients has also been associated with reduced structural connectivity in frontotemporal and inferior parietal regions.7 These studies were cross-sectional and so cannot ascertain a direction of causation or effect. Additionally, the regions of cortical thinning that were associated with WMH volume did Received August 17, 2015; final revision received November 11, 2015; accepted November 17, 2015. From the Brain Research Imaging Centre (D.A.D., E.S., N.A.R., B.S.A., M.V.H., S.M.M., M.E.B., J.M.W.), Neuroimaging Sciences, Centre for Clinical Brain Sciences (D.A.D., E.S., N.A.R., B.S.A., M.V.H., S.M.M., M.E.B., J.M.W.), Department of Psychology (S.J.R., S.R.C., A.P., J.C., I.J.D.), and Centre for Cognitive Ageing and Cognitive Epidemiology (S.J.R., S.R.C., J.M.S., M.E.B., I.J.D., J.M.W.), Alzheimer Scotland Dementia Research Centre (J.M.S.), The University of Edinburgh, Edinburgh, UK; Scottish Imaging Network, a Platform for Scientific Excellence (SINAPSE) Collaboration (D.A.D., E.S., N.A.R., B.S.A., M.V.H., S.M.M., M.E.B., J.M.W.); Department of Neurology and Neurosurgery, McConnell Brain Imaging Center, Montreal Neurological Institute, McGill University, Montreal, QC, Canada (S.K., A.C.E.); and Department of Psychiatry, Douglas Mental Health University Institute, McGill University, Verdun, QC, Canada (S.K.). *Drs Deary and Wardlaw contributed equally. Correspondence to Joanna M. Wardlaw, MD, The University of Edinburgh, Centre for Clinical Brain Sciences, Chancellor’s Bldg, Little France, Edinburgh, EH16 4TJ, United Kingdom. E-mail [email protected] © 2015 The Authors. Stroke is published on behalf of the American Heart Association, Inc., by Wolters Kluwer. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution, and reproduction in any medium, provided that the original work is properly cited. Stroke is available at http://stroke.ahajournals.org DOI: 10.1161/STROKEAHA.115.011229 1 2 Stroke February 2016 Downloaded from http://stroke.ahajournals.org/ by guest on March 28, 2017 not overlie regions with the greatest incidence of WMH,4 for example, centrifugally around the ventricles and superiorly towards the cranial vertex. Longitudinal studies are required to determine whether larger WMH volumes at baseline and larger increases in WMH volume are associated with subsequent regional cortical thinning. A longitudinal study of the association between cortical morphology and WMH volume growth has previously been conducted in Cerebral Autosomal Dominant Arteriopathy with Subcortical Infarcts and Leucoencephalopathy (CADASIL) patients.13 This study found that although lacunar lesions were strongly related to worsening cortical morphology, WMH volume was not strongly related to worsening cortical morphology in CADASIL. In the present study, we assessed longitudinal associations between change in WMH volume and change in cortical thickness to determine if the relationship between WMH and regional cortical thinning could be causal in community-dwelling subjects from ages 73 to 76.3,11 If the relationship between WMH and the regions of cortex that were thinned was potentially causal, then we hypothesized that there would be an association between (1) baseline whole WMH volume and change in cortical thickness over the next few years; (2) baseline cortical thickness and change in WMH volume; and (3) change in WMH volume and change in cortical thickness. To test these hypotheses, we measured progression of WMH volume and cortical thinning from ≈73 to ≈76 years of age in community-dwelling subjects from the Lothian Birth Cohort 1936 (LBC1936) study. Methods Study Approval and Subject Consent Approval for the LBC1936 study protocol was obtained from the Multicentre Research Ethics Committee for Scotland (MREC/01/0/56) and Lothian Research Ethics Committee (LREC/2003/2/29). All subjects gave written, informed consent. Subjects In the present study, we assessed 351 (Nmale=202) community-dwelling subjects from the LBC1936 study14,15 that had full brain MRI measures, clinical and cognitive assessments at imaging baseline (mean age 72.71±0.72 years), and follow-up (mean age 76.40±0.64 years). These subjects were not deliberately selected, rather they were simply those who agreed to participate in brain scanning and had complete data sets at baseline and follow-up (Figure 1). We recorded Mini Mental State Examination (MMSE) scores at baseline and follow-up to screen for possible dementia.14,15 We did not exclude subjects based on MMSE but included MMSE as an adjustment variable. We recorded the following vascular risk factors (VRF) during a clinical research facility visit: history of hypertension, hypercholesterolemia, diabetes mellitus, smoking, history of cardiovascular disease, and body mass index. History of cardiovascular disease includes self-reported incidences of coronary heart disease, stroke, peripheral arterial disease, and aortic disease. We also took blood samples and measured blood pressure but did not include these variables (eg, systolic blood pressure and glycated haemoglobin) in the present analysis to maximize the number of subjects with complete data sets and to be consistent with previous work.3 Further, we have previously shown that historical variables, for example, history of hypertension, have greater associations with WMH than measured variables, for example, systolic blood pressure.4 Brain MRI Acquisition Brain MRI acquisition parameters were described in detail previously.16 Briefly, all subjects had brain MRI on the same 1.5 tesla GE Figure 1. Subject recruitment flow chart. WMH indicates white matter hyperintensity. Signa Horizon HDx clinical scanner (General Electric, Milwaukee, WI), maintained on a careful quality assurance programme, at baseline and follow-up. The scanning protocol was the same at baseline and follow-up and acquired T1-, T2-, T2*-, and fluid-attenuated inversion recovery–weighted images.16 WMH Volume and Cortical Thickness Measurement We measured intracranial volume and whole WMH volume in milliliters using a validated multispectral image processing method that combines T1-, T2-, T2*-, and fluid-attenuated inversion recovery– weighted MRI sequences for segmentation.17–19 We measured cortical thickness using the fully automated Civet image processing pipeline developed at the Montreal Neurological Institute.20,21 Civet measures cortical thickness at 81 924 vertices (the perpendicular distance between GM and WM surfaces) across the cortex.20–23 For clarity, we refer to vertex as the perpendicular distance between GM and WM surfaces, not the cranial vertex. We manually verified WMH volume masks and cortical thickness maps as per procedures described previously.17–19,22,23 Specifically, we followed STRIVE (Standards for Reporting Vascular Changes on Neuroimaging) guidelines for segmenting WMH, manually removing cortical and subcortical infarcts from WMH masks.19 Finally, the reliability of cortical thickness20–22 and WMH17–19 measurements were tested and reported previously. Statistical Analysis All statistical analyses were performed in Matrix Laboratory (MATLAB) R2014a (© 1994–2014 The MathWorks, Inc.). We assessed changes in overall mean cortical thickness (mean thickness of the whole cortical mantle), whole brain WMH volume, and continuous variables used in adjustment, for example, body mass index, from 73 years to 76 years using paired t-tests. Log-transforming the positively skewed WMH distributions had little effect on our results, and therefore, we maintained their original scale (proportion of intracranial volume) to simplify interpretation. We assessed changes in binary variables used in adjustment, for example, history of hypertension, using z-tests of proportion. Cortical vertex-wise regression analyses were performed using the SurfStat MATLAB toolbox (http://www.math.mcgill.ca/ keith/surfstat). We tested 5 vertex-wise regression models where (1) cortical thickness at 73 years at each vertex was the dependent variable and WMH volume at 73 years was the independent variable; (2) cortical thickness at 76 years at each vertex was the dependent variable and WMH volume at 76 years was the independent variable; (3) change in cortical thickness at each vertex Dickie et al WMH Do Not Predict Cortical Thinning 3 was the dependent variable and WMH volume at 73 years was the independent variable; (4) cortical thickness at 73 years at each vertex was the dependant variable and change in WMH volume was the independent variable; and (5) change in cortical thickness at each vertex was the dependent variable and change in WMH volume was the independent variable. We defined change in WMH and cortical thickness as individual measurements at 76 years minus measurements at 73 years. We used false discovery rate to correct for multiple comparisons and calculated Q values, that is, false discovery rate–corrected P values,24 for all vertex-wise regressions thresholded at 0.05. As reported by others,3 all models were controlled for sex, MMSE, age in days, years of education, body mass index, and VRF. Finally, we also included childhood (age 11) intelligence quotient as a controlling variable to test whether any associations between WMH and cortical thickness were because of the influence of premorbid levels of cognitive ability. Results Downloaded from http://stroke.ahajournals.org/ by guest on March 28, 2017 Baseline Only and Follow-Up Subject Comparisons There were no significant differences at baseline (73 years) between subjects who did return for follow-up cortical thickness measurement and subjects who did not return for follow-up (at 76 years) in overall mean cortical thickness (3.11 mm versus 3.10 mm, t=0.75; P=0.46); WMH volume (0.78% intracranial volume versus 0.84% intracranial volume, t=−0.78; P=0.43); history of cardiovascular disease (27.3% versus 26.2%, z=0.29; P=0.39); current smoking (6.5% versus 8.2%, z=−0.76; P=0.22); hypercholesterolemia (40.0% versus 43.1%, z=−0.71; P=0.24); hypertension (46.5% versus 51.8%, z=−1.2; P=0.11); diabetes mellitus (10.1% versus 11.3%, z=−0.43; P=0.33); body mass index (27.8 versus 27.8, t=−0.21; P=0.84); years of education (10.85 years versus 10.89 years, t=−0.448; P= 0.65); nor age 11 intelligence quotient (101.9 versus 100.2, t=1.24; P=0.22). However, subjects who did return had higher MMSE scores at baseline than those who did not return (28.9 versus 28.6, t=2.49; P=0.01). The remaining results are only for the 351 longitudinal subjects who had full clinical, cognitive, and brain MRI data at baseline and follow-up. Baseline, Follow-Up, and Changes in Cognitive, VRF, Cortical Thickness, and WMH Measurements Baseline, follow-up, and changes in cognitive, VRF, cortical thickness, and WMH measurements are shown in Table 1. There were significant increases in the proportions of subjects with hypertension, hypercholesterolemia, and cardiovascular disease and a decrease in MMSE from baseline (73 years) to follow-up (76 years). Overall, mean cortical thickness generally decreased (Cohen’s d of change=−0.45) with age, and WMH volume generally increased (Cohen’s d of change=0.93) with age (Table 1). The Spearman correlation matrix between all brain changes and independent variables (Table 2) shows that cortex thinning was generally more pronounced in older subjects (ρ=−0.13; P=0.02). All other independent variables had limited partial effects on WMH and cortical thinning (beyond the effect of time point; Table 2). Cross-Sectional and Longitudinal Global Correlations Between Overall Mean Cortical Thickness and WMH Volume Cross-sectional global correlations between overall mean cortical thickness and WMH volume at 73 years (r=−0.06; P=0.27) and 76 years (r=−0.08; P=0.12) were not significant. Pairwise longitudinal global correlations between (1) WMH volume at 73 years and change in overall mean cortical thickness (r=−0.07; P=0.19); (2) overall mean cortical thickness at 73 years and change in WMH volume (r=0.01; P= 0.82); and (3) change in WMH volume and change in overall mean cortical thickness (r=−0.02; P=0.67) were also not significant. Cross-Sectional Vertex-Wise Regression Models of Regional Cortical Thickness and WMH Volume Cross-sectional vertex-wise regression models of cortical thickness and WMH volume at 76 years are shown in Figure 2. Cross-sectional data from 73 years are not shown because the pattern of associations between cortical thickness and WMH volume was almost identical at 73 and 76 years. All models are corrected for VRF, MMSE, education level, and Table 1. Baseline, Follow-Up, and Changes in Cognitive, VRF, Cortical Thickness, and WMH Measurements N=351 11 Years 73 Years 76 Years 3 Year Change Childhood IQ 102.21±15.75 … … … Sex (%male) … 57.39 57.39 … Education, y … 10.86±1.18 10.86±1.18 Hypertension (% +ve) … 46.31% 52.56% 6.25% (z=1.66, P=0.049)* Hypercholesterolemia (% +ve) … 39.20% 46.02% 6.82% (z=1.83, P=0.034)* History of CVD (% +ve) … 26.14% 32.67% 6.53% (z=1.90, P=0.029)* Smoking (% current) … 6.53% 5.97% −0.85% (z=−0.31, P=0.378) Diabetes (% +ve) … 9.09% 12.22% BMI … 27.57±4.25 27.56±4.46 −0.02±1.49 (t=−0.24, P=0.812) MMSE … 28.97±1.26 28.72±1.46 −0.25±1.48 (t=−3.01, P=0.003)* Mean overall cortical thickness, mm … 3.17±0.15 3.12±0.15 −0.05±0.11 (t=−8.18, P<0.001)* WMH volume (% of ICV) … 0.76±0.71 1.02±0.91 0.26±0.28 (t=17.50, P<0.001)* … 3.13% (z=1.34, P=0.090) BMI indicates body mass index; CVD, cardiovascular disease; ICV, intracranial volume; IQ, intelligence quotient; MMSE, Mini Mental State Examination; and WMH, white matter hyperintensity. *P<0.05. 4 Stroke February 2016 Downloaded from http://stroke.ahajournals.org/ by guest on March 28, 2017 sex. Inclusion of age 11 intelligence quotient as a controlling variable made little difference to the cortical t-maps (data not shown). Warm colors in Figure 2 show regions where greater WMH volume was associated with reduced cortical thickness. The significance of cross-sectional associations is shown on the left panel of Figure 3. There were consistent patterns of negative cross-sectional associations at 73 and 76 years within and surrounding the Sylvian fissure extending superiorly to the parietal lobe, posteriorly to the occipital lobe, and anteriorly to the frontal lobe. Therefore, having greater WMH volume was cross-sectionally associated with reduced cortical thickness in specific regions only, that is, within and surrounding the Sylvian fissure. Associations between greater WMH volume and greater cortical thickness in superior regions (cold colors in Figure 2) were all nonsignificant. A scatter plot of the peak cross-sectional association in the Sylvian fissure and surrounding area at 76 years (ρ=−0.276; Q=0.004) is shown in Figure 3. Longitudinal Vertex-Wise Regression Models of Regional Cortical Thickness and WMH Volume Longitudinal vertex-wise associations between (1) baseline WMH volume and change in cortical thickness (gC-wB in Figure 2); (2) baseline cortical thickness and change in WMH volume (wC-gB in Figure 2); and (3) change in WMH volume and change in cortical thickness (C-C in Figure 2) were all nonsignificant across the cortex (Q>0.05; Figure 3). Therefore, having a larger WMH volume at 73 years (or larger change in WMH volume between 73 and 76 years) did not predict greater cortical thinning between 73 and 76 years at any part of the cortex. Neither did a thinner cortex at 73 years predict greater WMH growth between 73 and 76 years. The longitudinal association between WMH change and overall mean cortical thickness change was descriptively much stronger in subjects with MMSE≤26 (r=−0.220; P=0.41 versus r=−0.003; P=0.96) but this was not statistically significant potentially because of the small number of subjects with MMSE≤26 (N=16). Discussion We have replicated cross-sectional associations between greater WMH volume and regional cortical thinning around the Sylvian fissure3,5,9–12; however, we found no longitudinal associations between (1) baseline WMH volume and change in cortical thickness; (2) baseline cortical thickness and change in WMH volume; or (3) change in WMH volume and change in cortical thickness at any part of the cortex in communitydwelling subjects from 73 to 76 years. The cross-sectional associations found here between greater WMH volume and Table 2. Spearman Correlation Matrix of Overall Mean Cortical Thickness and WMH Changes and Independent Variables rho (PValue) Cort chng WMH chg BMI Sex CVD DIAB HCHL HBP SMOK EDU IQ11 MMSE Age Cort chng WMH chg 0.029 (0.595) BMI −0.056 (0.306) −0.043 (0.441) Sex 0.058 (0.292) 0.038 (0.486) −0.002 (0.964) CVD 0.017 (0.753) −0.050 (0.370) 0.088 (0.109) −0.148 (0.007)* DIAB 0.058 (0.295) 0.025 (0.652) 0.149 (0.007)* −0.078 (0.159) 0.098 (0.076) HCHL 0.050 (0.361) 0.055 (0.317) 0.106 (0.054) 0.003 (0.956) 0.179 (0.001)* 0.220 (<0.001)* HBP 0.052 (0.350) −0.010 (0.855) 0.195 (<0.001)* −0.018 (0.747) 0.216 (<0.001)* 0.150 (0.006)* 0.283 (<0.001)* SMOK 0.025 (0.648) 0.003 (0.962) 0.045 (0.417) −0.084 (0.126) 0.115 (0.038)* 0.068 (0.218) 0.068 (0.218) 0.024 (0.658) EDU −0.051 (0.355) 0.074 (0.183) −0.171 (0.002)* 0.039 (0.482) −0.019 (0.728) −0.084 (0.126) 0.011 (0.846) −0.020 (0.717) −0.086 (0.119) IQ11 0.060 (0.273) −0.029 (0.604) −0.142 (0.010)* 0.073 (0.185) −0.006 (0.913) −0.072 (0.191) 0.019 (0.736) −0.017 (0.756) −0.121 (0.028)* 0.528 (<0.001)* MMSE 0.035 (0.529) 0.012 (0.835) −0.064 (0.250) 0.104 (0.059) −0.043 (0.432) −0.061 (0.271) −0.035 (0.521) −0.048 (0.386) −0.107 (0.052) 0.234 0.378 (<0.001)* (<0.001)* Age −0.128 (0.020)* 0.039 (0.483) −0.004 (0.941) 0.086 (0.119) −0.015 (0.783) 0.022 (0.692) 0.011 (0.849) −0.003 (0.951) −0.048 (0.387) −0.009 (0.864) −0.014 (0.802) −0.077 (0.164) BMI indicates body mass index; Cort chng, change in overall mean cortical thickness from 73 to 76 years; CVD, cardiovascular disease; DIAB, diabetes mellitus; EDU, education; HCHL, high cholesterol; HBP, high blood pressure; IQ11, intelligence quotient at age 11 years; MMSE, Mini Mental State Examination; SMOK, smoking; and WMH, white matter hyperintensity. *P<0.05. Dickie et al WMH Do Not Predict Cortical Thinning 5 Downloaded from http://stroke.ahajournals.org/ by guest on March 28, 2017 Figure 2. Cross-sectional (76y panel) and longitudinal (C-C, gC-wB, and wC-gB panels) t-maps of vertex-wise associations between cortical thickness and whole white matter hyperintensity (WMH) volume. Warm colors show where greater WMH volume is associated with reduced cortical thickness. The significance of these associations is shown in Figure 3. C-C indicates cortical thickness change and WMH volume change from 73 to 76 years; gC-wB, cortical thickness change and WMH volume at 73 years; wC-gB, WMH volume change and cortical thickness at 73 years. reduced cortical thickness in the region of the Sylvian fissure are consistent with previous GM volume,10 voxel-based morphometry,5,11 and cortical thickness3 studies. As with previous studies, the regions of cortical thinning–WMH associations that we found are not consistent with the most frequent WMH locations and areas of expansion,4 for example, centrifugally around the ventricles and superiorly towards the cranial vertex. Our results suggest that WMH volume and cortical atrophy both worsen with age and that their individual differences share some causes—thus the cross-sectional associations. However, their changes from 73 to 76 years do not appear to be associated, and such correlated change would have been one indicator of a possible causal association. This conclusion is consistent with a longitudinal study in CADASIL patients that, although finding strong associations between lacunar lesions and cortical morphological changes, found a limited association between cortical morphological changes and WMH volume.13,25 Strengths of our study include the ability to test longitudinal and cross-sectional associations between WMH volume and regional cortical thickness in a large sample of community-dwelling subjects. Other strengths include the agehomogeneous subjects with childhood intelligence quotient assessments and who are now in the eighth decade of life where the risk of dementia increases substantially.26 As well as the narrow age range, other novel features of the LBC1936 study (eg, all subjects are white Caucasian) may have minimized any potentially strong confounding effects that factors such as age, mixed ethnicity, and geography might have had in a less homogeneous sample. We measured WMH volume and cortical thickness using well-validated quantitative techniques that we manually checked and quality controlled post-pipeline for each subject at both time points.17,20,22 The raw brain MRI from which we measured WMH volume and cortical thickness were obtained using the same protocol on the same carefully maintained scanner at both time points.16 Despite these strengths and our replication of previous cross-sectional findings,3,5,10,11 our study has limitations. The follow-up time of 3 years is a major limitation because it may not have been long enough to detect correlated changes between WMH and cortical thinning. We chose 3-year follow-up (rather than a longer time) to maximize subject retention and to be consistent with previous studies, for example, Austrian Stroke Prevention Study.1,27 Further, we have previously detected cross-sectional differences in WMH because of age within the narrow age band (<3 years) in the LBC1936 study.28 We are studying these subjects again at 6 years follow-up, and this may provide better evidence for any potentially causal relationships not identified here. We will attempt to ascertain the reasons for subjects lost to follow-up and will use full information maximum likelihood analyses, checked against analyses of completers, to minimize the effect of loss to follow-up. We defined change as individual measurements at 76 years minus measurements 6 Stroke February 2016 Downloaded from http://stroke.ahajournals.org/ by guest on March 28, 2017 Figure 3. Significance of cross-sectional (X-S) and longitudinal (L) vertex-wise associations between cortical thickness and white matter hyperintensity (WMH) volume. There were consistent patterns of negative cross-sectional associations at 73 (data not shown) and 76 years within and surrounding the Sylvian fissure extending superiorly to the parietal lobe, posteriorly to the occipital lobe, and anteriorly to the frontal lobe (left). The clear grey Q-map (right) shows that no longitudinal associations between cortical thickness and WMH volume were significant. at 73 years. We are aware that there are other ways of assessing change, for example, those often applied to cognitive variable change.29 However, the approach we used here is often applied to measure changes in brain morphology.30 Although the homogeneous nature of the LBC1936 cohort may provide increased power from having less need to control for confounding variables, for example, age and ethnicity, it limits the generalizability of our results. The longitudinal subjects we assessed here generally had higher MMSE than subjects who did not return for follow-up, and this may also limit the generalizability of our results, for example, longitudinal associations between WMH, and cortical thinning may be stronger in subjects with lower cognitive scores. We could not adequately test this here because of the small number of subjects with MMSE≤26 (N=16), and future work is required to determine whether associations are stronger in cognitively impaired subjects. Although the locations of cross-sectional associations that we (and others3,5,9–12) found between WMH and cortical thinning do not directly reflect common areas for WMH expansion, areas of associations were proximate to the tapetum of the corpus callosum fiber tracts which extend inferiorly and anteriorly into the temporal lobes.31 Further work is required to determine whether the locations of cross-sectional WMH and cortical thinning associations are because of an indirect connection through the perisylvian cortex and tapetum of the corpus callosum. Finally, it is difficult to prove or disprove a causal relationship between WMH and cortical thinning in observational studies. However, we adjusted for a number of variables known to influence WMH and cortical thinning, and although correlation is not necessarily causation, correlation is fundamental to causation.32,33 Therefore, the lack of longitudinal correlations implies the lack of a causal relationship from 73 to 76 years. Notwithstanding these limitations, we have shown that although they both worsen with age, WMH volume progression and regional cortical thinning do not seem to have a correlative/causal longitudinal relationship from 73 to 76 years. Further longitudinal studies with longer follow-up times and with more time points at different ages are required to determine whether causal relationships become apparent over longer periods of time and/or at different stages of life. The underlying cause(s) of WMH growth and cortical thinning have yet to be fully determined. Acknowledgments We thank the funders, participants, research centers, clinical, and administrative staff who contributed to the LBC1936 study (detailed fully at http://www.lothianbirthcohort.ed.ac.uk/). Sources of Funding This work was funded by a Scottish Funding Council Early Career Researcher grant to the Scottish Imaging Network—A Platform for Scientific Excellence (http://www.sinapse.ac.uk; DAD); Research into Ageing program grant (Drs Deary and Starr) and the Age UKfunded Disconnected Mind project (Drs Deary, Starr, and Wardlaw), with additional funding from the UK Medical Research Council (Drs Deary, Starr, and Wardlaw, and M.E. Bastin); and Scottish Funding Council through the Scottish Imaging Network—A Platform for Scientific Excellence (Dr Wardlaw). Dickie et al WMH Do Not Predict Cortical Thinning 7 Disclosures Dr Wardlaw reports money (grants) paid to The University of Edinburgh from Medical Research Council, Age UK, Row Fogo Charitable Trust, and Scottish Funding Council for her efforts on the LBC1936 study and various imaging projects. Dr Deary reports money (grants) paid to The University of Edinburgh from Medical Research Council and Age UK for his efforts on the LBC1936 study. Dr Deary reports money paid to him for board membership on Medical Research Council. All other authors have no disclosures. References Downloaded from http://stroke.ahajournals.org/ by guest on March 28, 2017 1. Schmidt R, Enzinger C, Ropele S, Schmidt H, Fazekas F; Austrian Stroke Prevention Study. Progression of cerebral white matter lesions: 6-year results of the Austrian Stroke Prevention Study. Lancet. 2003;361:2046–2048. 2. Sowell ER, Peterson BS, Thompson PM, Welcome SE, Henkenius AL, Toga AW. Mapping cortical change across the human life span. Nat Neurosci. 2003;6:309–315. doi: 10.1038/nn1008. 3. Tuladhar AM, Reid AT, Shumskaya E, de Laat KF, van Norden AG, van Dijk EJ, et al. Relationship between white matter hyperintensities, cortical thickness, and cognition. Stroke. 2015;46:425–432. doi: 10.1161/ STROKEAHA.114.007146. 4. Wardlaw JM, Allerhand M, Doubal FN, Valdes Hernandez M, Morris Z, Gow AJ, et al. Vascular risk factors, large-artery atheroma, and brain white matter hyperintensities. Neurology. 2014;82:1331–1338. doi: 10.1212/WNL.0000000000000312. 5. Wen W, Sachdev PS, Chen X, Anstey K. Gray matter reduction is correlated with white matter hyperintensity volume: a voxel-based morphometric study in a large epidemiological sample. Neuroimage. 2006;29:1031–1039. doi: 10.1016/j.neuroimage.2005.08.057. 6. Debette S, Markus HS. The clinical importance of white matter hyperintensities on brain magnetic resonance imaging: systematic review and meta-analysis. BMJ. 2010;341:c3666. 7. Lawrence AJ, Chung AW, Morris RG, Markus HS, Barrick TR. Structural network efficiency is associated with cognitive impairment in small-vessel disease. Neurology. 2014;83:304–311. doi: 10.1212/WNL.0000000000000612. 8. Thompson PM, Hayashi KM, de Zubicaray G, Janke AL, Rose SE, Semple J, et al. Dynamics of gray matter loss in Alzheimer’s disease. J Neurosci. 2003;23:994–1005. 9. Godin O, Maillard P, Crivello F, Alpérovitch A, Mazoyer B, Tzourio C, et al. Association of white-matter lesions with brain atrophy markers: the three-city Dijon MRI study. Cerebrovasc Dis. 2009;28:177–184. doi: 10.1159/000226117. 10. Knopman DS, Griswold ME, Lirette ST, Gottesman RF, Kantarci K, Sharrett AR, et al; ARIC Neurocognitive Investigators. Vascular imaging abnormalities and cognition: mediation by cortical volume in nondemented individuals: atherosclerosis risk in communities-neurocognitive study. Stroke. 2015;46:433–440. doi: 10.1161/STROKEAHA.114.007847. 11. Raji CA, Lopez OL, Kuller LH, Carmichael OT, Longstreth WT Jr, Gach HM, et al. White matter lesions and brain gray matter volume in cognitively normal elders. Neurobiol Aging. 2012;33:834.e837–834.e816 12. Rossi R, Boccardi M, Sabattoli F, Galluzzi S, Alaimo G, Testa C, et al. Topographic correspondence between white matter hyperintensities and brain atrophy. J Neurol. 2006;253:919–927. doi: 10.1007/ s00415-006-0133-z. 13. Jouvent E, Mangin JF, Duchesnay E, Porcher R, Düring M, Mewald Y, et al. Longitudinal changes of cortical morphology in CADASIL. Neurobiol Aging. 2012;33:1002.e29–1002.e36. doi: 10.1016/j. neurobiolaging.2011.09.013. 14. Deary IJ, Gow AJ, Taylor MD, Corley J, Brett C, Wilson V, et al. The Lothian Birth Cohort 1936: a study to examine influences on cognitive ageing from age 11 to age 70 and beyond. BMC Geriatr. 2007;7:28. doi: 10.1186/1471-2318-7-28. 15. Deary IJ, Gow AJ, Pattie A, Starr JM. Cohort profile: the Lothian Birth Cohorts of 1921 and 1936. Int J Epidemiol. 2012;41:1576–1584. doi: 10.1093/ije/dyr197. 16. Wardlaw JM, Bastin ME, Valdés Hernández MC, Maniega SM, Royle NA, Morris Z, et al. Brain aging, cognition in youth and old age and vascular disease in the Lothian Birth Cohort 1936: rationale, design and methodology of the imaging protocol. Int J Stroke. 2011;6:547–559. doi: 10.1111/j.1747-4949.2011.00683.x. 17. Hernández Mdel C, Ferguson KJ, Chappell FM, Wardlaw JM. New multispectral MRI data fusion technique for white matter lesion segmentation: method and comparison with thresholding in FLAIR images. Eur Radiol. 2010;20:1684–1691. doi: 10.1007/s00330-010-1718-6. 18. Wang X, Valdés Hernández MC, Doubal F, Chappell FM, Wardlaw JM. How much do focal infarcts distort white matter lesions and global cerebral atrophy measures? Cerebrovasc Dis. 2012;34:336–342. doi: 10.1159/000343226. 19. Wardlaw JM, Smith EE, Biessels GJ, Cordonnier C, Fazekas F, Frayne R, et al; STandards for ReportIng Vascular changes on nEuroimaging (STRIVE v1). Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. Lancet Neurol. 2013;12:822–838. doi: 10.1016/S1474-4422(13)70124-8. 20. Ad-Dab’bagh Y, Lyttelton O, Muehlboeck J, Lepage C, Einarson D, Mok K, et al. The Civet image-processing environment: A fully automated comprehensive pipeline for anatomical neuroimaging research. Poster presented at: Proceedings of the 12th Annual Meeting of the Organization for Human Brain Mapping. June 11–15, 2006; Florence, Italy. 21.Zijdenbos AP, Forghani R, Evans AC. Automatic “pipeline” analysis of 3-D MRI data for clinical trials: application to multiple sclerosis. IEEE Trans Med Imaging. 2002;21:1280–1291. doi: 10.1109/ TMI.2002.806283. 22. Karama S, Ad-Dab’bagh Y, Haier R, Deary IJ, Lyttelton OC, Lepage C, et al. Positive association between cognitive ability and cortical thickness in a representative us sample of healthy 6 to 18 year-olds. Intelligence. 2009;37:145–155. 23.Karama S, Ducharme S, Corley J, Chouinard-Decorte F, Starr JM, Wardlaw JM, et al. Cigarette smoking and thinning of the brain’s cortex. Mol Psychiatry. 2015;20:778–785. doi: 10.1038/mp.2014.187. 24. Genovese CR, Lazar NA, Nichols T. Thresholding of statistical maps in functional neuroimaging using the false discovery rate. Neuroimage. 2002;15:870–878. doi: 10.1006/nimg.2001.1037. 25.Jouvent E, Mangin JF, Porcher R, Viswanathan A, O’Sullivan M, Guichard JP, et al. Cortical changes in cerebral small vessel diseases: a 3D MRI study of cortical morphology in CADASIL. Brain. 2008;131(pt 8):2201–2208. doi: 10.1093/brain/awn129. 26. Matthews F, Brayne C; Medical Research Council Cognitive Function and Ageing Study Investigators. The incidence of dementia in England and Wales: findings from the five identical sites of the MRC CFA Study. PLoS Med. 2005;2:e193. doi: 10.1371/journal.pmed.0020193. 27. Schmidt R, Fazekas F, Kapeller P, Schmidt H, Hartung HP. MRI white matter hyperintensities: three-year follow-up of the Austrian Stroke Prevention Study. Neurology. 1999;53:132–139. 28. Aribisala BS, Morris Z, Eadie E, Thomas A, Gow A, Valdés Hernández MC, et al. Blood pressure, internal carotid artery flow parameters, and age-related white matter hyperintensities. Hypertension. 2014;63:1011– 1018. doi: 10.1161/HYPERTENSIONAHA.113.02735. 29. Deary IJ, Allerhand M, Der G. Smarter in middle age, faster in old age: a cross-lagged panel analysis of reaction time and cognitive ability over 13 years in the West of Scotland Twenty-07 Study. Psychol Aging. 2009;24:40–47. doi: 10.1037/a0014442. 30. Veldink JH, Scheltens P, Jonker C, Launer LJ. Progression of cerebral white matter hyperintensities on MRI is related to diastolic blood pressure. Neurology. 1998;51:319–320. 31.Wakana S, Jiang H, Nagae-Poetscher LM, van Zijl PC, Mori S. Fiber tract-based atlas of human white matter anatomy. Radiology. 2004;230:77–87. doi: 10.1148/radiol.2301021640. 32. Aldrich J. Correlations genuine and spurious in pearson and yule. Stat Sci. 1995:364–376. 33. Pearson K, Lee A, Yule G. On the distribution of frequency (variation and correlation) of the barometric height at divers stations. Philos Trans R Soc Lond A. 1897:423–469. Downloaded from http://stroke.ahajournals.org/ by guest on March 28, 2017 Progression of White Matter Disease and Cortical Thinning Are Not Related in Older Community-Dwelling Subjects David Alexander Dickie, Sherif Karama, Stuart J. Ritchie, Simon R. Cox, Eleni Sakka, Natalie A. Royle, Benjamin S. Aribisala, Maria Valdés Hernández, Susana Muñoz Maniega, Alison Pattie, Janie Corley, John M. Starr, Mark E. Bastin, Alan C. Evans, Ian J. Deary and Joanna M. Wardlaw Stroke. published online December 22, 2015; Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright © 2015 American Heart Association, Inc. All rights reserved. Print ISSN: 0039-2499. Online ISSN: 1524-4628 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://stroke.ahajournals.org/content/early/2015/12/22/STROKEAHA.115.011229 Free via Open Access Data Supplement (unedited) at: http://stroke.ahajournals.org/content/suppl/2016/12/20/STROKEAHA.115.011229.DC1 Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Stroke is online at: http://stroke.ahajournals.org//subscriptions/ 2 Stroke 日本語版 Vol. 11, No. 1 Full Article 地域在住の高齢者において白質病変の進行と皮質の菲薄化 は関連しない Progression of White Matter Disease and Cortical Thinning Are Not Related in Older Community-Dwelling Subjects David Alexander Dickie, PhD1,2,6; Sherif Karama, MD, PhD7,8; Stuart J. Ritchie, PhD3,4; Simon R. Cox, PhD3,4; Eleni Sakka, MSc1,2,6; Natalie A. Royle, PhD1,2,6; Benjamin S. Aribisala, PhD1,2,6; Maria Valdés Hernández, PhD1,2,6; Susana Muñoz Maniega, PhD1,2,6; Alison Pattie, BSc3; Janie Corley, MA3; John M. Starr, MD4,5; Mark E. Bastin, DPhil1,2,4,6; Alan C. Evans, PhD7; Ian J. Deary, PhD3,4; Joanna M. Wardlaw, MD1,2,4,6 1 Brain Research Imaging Centre, 2Neuroimaging Sciences, Centre for Clinical Brain Sciences, 3Department of Psychology, and 4Centre for Cognitive Ageing and Cognitive Epidemiology, 5Alzheimer Scotland Dementia Research Centre, 6The University of Edinburgh, Edinburgh, UK; Scottish Imaging Network, a Platform for Scientific Excellence (SINAPSE) Collaboration; 7Department of Neurology and Neurosurgery, McConnell Brain Imaging Center, Montreal Neurological Institute, McGill University, Montreal, QC, Canada; and 8 Department of Psychiatry, Douglas Mental Health University Institute, McGill University, Verdun, QC, Canada 背景および目的:脳全体の白質高信号病変( WMH )の容 積と局所皮質厚との間の横断的および縦断的関連を評価 した。 方法:Lothian Birth Cohort 1936 研究の地域在住の被験 者 351 例において,約 73 歳および約 76 歳の時点で WMH 容積と局所皮質厚を MRI により測定した。局所皮質厚と WMH 容積の横断的および縦断的関連について,年齢,性 別,ミニメンタルステート検査,学歴,11 歳時の知能指数, 心血管危険因子で調整した重線形回帰分析により評価し た。 結果:73 歳時および 76 歳時において,WMH 容積とシル ビウス裂の内部および周辺部の皮質厚との間に横断的関 連が認められた( rho =− 0.276,Q = 0.004 )。しかし, (1) ベースラインの WMH 容積と皮質厚の変化,( 2 )ベース ラインの皮質厚と WMH 容積の変化,( 3 )WMH 容積の 変化と皮質厚の変化,との間に有意な縦断的関連は認め られなかった。 結論: 本研究の結果,WMH 容積と皮質の菲薄化はとも に加齢により悪化し,シルビウス裂の内部および周辺部 で横断的関連がみられることが明らかになった。しかし, これらの比較的健康な高齢被験者において,73 歳時から 76 歳時の WMH 容積の変化と皮質の菲薄化との間に縦断 的関連はみられなかった。WMH の拡大と皮質の菲薄化の 根本的原因は,まだ完全には解明されていない。 Stroke . 2016; 47: 410-416. DOI: 10.1161/STROKEAHA.115.011229. 地域在住の高齢者では,MRI により,脳白質高信号 病変( WMH )の拡大と皮質の菲薄化を認めることが多 い 1-5 。このような所見の発現には顕著な個人差があるが, 頭葉構造,頭頂葉下部,楔部を含む )との間に,同様の 横断的な負の関連パターンが認められている 10。また, 小血管病の患者では,WMH 容積の増加が前頭側頭葉お WMH 容積の増加や皮質厚の減少が非常に高度にみられ よび頭頂葉下部領域の構造的な連絡性の低下に関連す る人は,脳卒中,認知症,認知機能障害および身体機能 る 7。 6-8 。こうした状況に効果的に介入す ただし,これらの研究は横断的であり,原因または結 るためには,WMH の拡大と皮質の菲薄化に関する機序 果の方向性を確認することはできない。また,WMH 容 と,両者がお互いの原因となっているか否かを理解する 積との関連が認められた皮質の菲薄化領域は,WMH が ことが非常に重要である。 最も多く発生する領域( 例えば,脳室から遠心状に広が 障害のリスクが高い 全脳の WMH 容積の拡大と, 灰白質 (GM)の容積, 密度, る領域,頭頂へと向かう上方の領域など )とは重なって 厚さの減少との間に関連があることは,横断的研究で明 いなかった 4。ベースラインで大きな WMH 容積および らかにされている 3,5,9-12。WMH 容積が大きい人は,一般 WMH 容積の顕著な拡大が,その後の局所的な皮質の菲 的に,前頭側頭葉および頭頂葉下部領域における皮質の 薄化に関連するか否かを明らかにするには,縦断的研究 厚さ 3 および密度 5,11 が比較的減少している。その他の が必要である。以前には,皮質下梗塞および白質脳症を 関心領域解析による研究でも,デフォルトモードネット 伴う常染色体優性遺伝性脳動脈症( CADASIL)患者を対 ワークにおいて全脳 WMH 容積と局所 GM 容積( 内側側 象に,皮質の形態と WMH 容積の増加との関連性を検討 PDLQLQGG 30 3 地域在住の高齢者において白質病変の進行と皮質の菲薄化は関連しない する縦断的研究が実施された 13。その結果,CADASIL において,ラクナ病変は皮質の形態学的悪化に強く関連 N=1091 (男性のN=548) 初期登録数 していたが,WMH 容積と皮質の形態学的悪化との間に 強い関連はみられなかった。 本研究では,WMH と局所皮質の菲薄化との因果関係 を明らかにするため,地域住民を対象に,WMH 容積の 変化と皮質厚の変化との関連を 73 歳から 76 歳において N=569 (男性のN=312) ベースラインで皮質厚, WMH,臨床評価, 認知機能評価を完了 縦断的に評価した 3,11。WMH と菲薄化した皮質領域との 間に因果関係があるとすれば,( 1 )ベースラインの全脳 WMH 容積と数年後の皮質厚の変化,( 2 )ベースライン の皮質厚と WMH 容積の変化,(3)WMH 容積の変化と 皮質厚の変化,との間に関連が検出されると仮定した。 N=351 (男性のN=202) ベースラインおよび 追跡調査で皮質厚, WMH,臨床評価, 認知機能評価を完了 N=351 (男性のN=202) 本研究で解析 上記の仮説を検証するため,Lothian Birth Cohort 1936 ( LBC1936 )研究の地域在住の被験者において,約 73 歳 図1 被験者登録のフローチャート。WMH:白質高信号病変。 から約 76 歳の時点において WMH 容積と皮質の菲薄化 の経過を評価した。 去の研究との一貫性を保つため 3,これらの変数( 収縮 期血圧,糖化ヘモグロビンなど )は解析に含めなかった。 方 法 高血圧などの既往歴の変数が,収縮期血圧などの測定値 研究の承認および被験者の同意 LBC1936 研究のプロトコルは,Multicentre Research の変数よりも WMH と強く関連することは既報の通りで ある 4。 Ethics Committee for Scotland ( MREC/01/0/56 )および Lothian Research Ethics Committee ( LREC/2003/2/29 ) の承認を受けた。すべての被験者から書面でインフォー 脳 MRI 画像 脳 MRI の 撮 像 パ ラ メ ー タ ー に つ い て は 以 前 に 詳 し く報告した 16。簡単に述べると,厳密な品質保証プロ ムド・コンセントを得た。 グラムで管理されている 1.5 テスラの同一の GE Signa 被験者 Horizon HDx 臨 床 画 像 診 断 装 置( General Electric, LBC1936 研究 14,15 の地域在住の被験者 351 例( 男性 Milwaukee,WI )を用い,ベースラインと追跡調査の の N = 202)を評価した。この研究では, 全脳 MRI 検査, 時点においてすべての被験者で脳 MRI 検査を実施し 臨床評価,認知機能評価がベースラインの画像検査の時 た。ベースラインと追跡調査の撮像プロトコルは同じで 点(平均年齢:72.71 ± 0.72 歳) および追跡調査の時点 (平 あ り,T1,T2,T2* 強 調 画 像 お よ び FLAIR 画 像 を 得 均年齢:76.40 ± 0.64 歳 )で実施された。これらの被験 た 16。 者は意図的に選択されておらず,単純に,脳画像の撮像 に同意し,ベースラインおよび追跡調査時点における完 全なデータセットが揃っていた被験者である( 図 1 ) 。 WMH 容積と皮質厚の測定 MRI の T1,T2,T2* 強調画像および FLAIR 画像を 認知症の可能性をスクリーニングするため,ベース 結合してセグメント化する,検証済みのマルチスペクト ラインと追跡調査の時点でミニメンタルステート検査 ル画像処理方法により,頭蓋内容積および全脳 WMH 容 ( MMSE )のスコアを記録した 14,15 。MMSE に基づく被 積を mL 単位で測定した 17-19。 験者の除外は行わず,MMSE は補正のための変数とし 皮質の厚さは,Montreal Neurological Institute で開発 て取り入れた。臨床研究施設への来院時,心血管危険因 された完全自動 Civet 画像処理パイプラインにより測定 子( VRF )として,高血圧の既往,高コレステロール血症, した 20,21。Civet は,皮質全体で 81,924 ヵ所の頂点( GM 糖尿病,喫煙状況,心血管疾患の既往,肥満指数( BMI ) と WM の表面の垂直距離 )において,その厚みを測定す を記録した。心血管疾患の既往には,自己申告による冠 る 20-23。我々は明確性を期して,この頂点を頭頂(cranial 動脈疾患,脳卒中,末梢動脈疾患,大動脈疾患も含まれ vertex )ではなく,GM と WM 表面の垂直距離と呼んで る。血液検体の採取と血圧測定も実施したが,完全なデー いる。 タセットが揃った被験者数をできるだけ増やし,かつ過 PDLQLQGG 過去に報告された方法に従い,手作業により WMH 30 4 Stroke 日本語版 Vol. 11, No. 1 容積マスクと皮質厚マップを検証した 17-19,22,23。具体 頂点における皮質の厚さの変化=従属変数,WMH 容積 的 に は,STRIVE( Standards for Reporting Vascular の変化=独立変数,とするモデルである。「 76 歳時の各 Changes on Neuroimaging )ガイドラインに従い,WMH 測定値 」−「 73 歳時の各測定値 」を,WMH および皮質 をセグメント化し,WMH マスクから皮質および皮質下 厚の変化量と定義した。 19 梗塞を手作業で除去した 。さらに,皮質厚 WMH 17-19 20-22 多重比較の補正には偽の発見率を用い,Q 値,すなわ および ち偽の発見率で補正した P 値を算出し,すべての頂点に の測定値の信頼性を検証したが,その結果に 関する回帰分析で閾値を 0.05 とした 24。他の研究者ら ついては以前に報告している。 の報告に従い 3,すべてのモデルにおいて性別,MMSE, 日齢,学歴( 教育年数 ) ,BMI,VRF で調整した。さら 統計解析 統 計 解 析 は す べ て Matrix Laboratory( MATLAB ) に,小児期( 11 歳時 )の知能指数を調整変数として加え, R2014a ( ©1994‒2014 The MathWorks,Inc )で実施した。 WMH と皮質厚との関連が,発症前の認知機能レベルの 73 歳から 76 歳における全体の平均皮質厚( 全皮質外套 影響によるものか否かを検証した。 の平均厚), 全脳 WMH 容積, 補正に用いた連続変数 ( BMI など )の変化は,対応のある t 検定で評価した。正の歪 結 果 みがみられた WMH 分布の対数変換により,結果への 影響はほとんどみられなかったため,解釈を簡略化する ベースラインのみ来院した被験者群と追跡調査 にも来院した被験者群との比較 ために最初の尺度(頭蓋内容積の比率)のまま検討した。 高血圧の既往など,補正に用いた二値変数の変化は,比 皮質厚測定の追跡調査( 76 歳時 )のために再来院し 率の z 検定を評価した。 た被験者と再来院しなかった被験者とを比較したとこ 皮 質 を 頂 点 と す る 回 帰 分 析 は SurfStat MATLAB ろ,ベースラインの時点( 73 歳 )において,全体の平均 ツールボックスで実施した( http://www.math.mcgill.ca/ , 皮 質 厚( 3.11 mm vs. 3.10 mm,t = 0.75,P = 0.46 ) keith/surfstat ) 。皮質を頂点とする回帰モデルとして次 WMH 容積( 頭蓋内容積の 0.78% vs. 0.84%,t =− 0.78, の 5 つを検証した。すなわち,( 1 )各頂点における 73 P = 0.43 ),心血管疾患の既往( 27.3% vs. 26.2%,z = 歳時の皮質の厚さ=従属変数,73 歳時の WMH 容積= P = 0.39 ),現在の喫煙( 6.5% vs. 8.2%, z =− 0.76, 0.29, 独立変数,( 2 )各頂点における 76 歳時の皮質の厚さ= P = 0.22 ),高コレステロール血症( 40.0% vs. 43.1%, 従属変数,76 歳時の WMH 容積=独立変数, ( 3)各頂点 z = − 0.71,P = 0.24 ), 高 血 圧( 46.5 % vs. 51.8 %, における皮質の厚さの変化=従属変数,73 歳時の WMH z =− 1.2,P = 0.11 ),糖尿病( 10.1% vs. 11.3%,z = 容積=独立変数,( 4 )各頂点における 73 歳時の皮質の − 0.43,P = 0.33 ) ,BMI( 27.8 vs. 27.8,t = − 0.21, 厚さ=従属変数,WMH 容積の変化=独立変数,( 5 )各 P = 0.84 ), 学 歴( 教 育 年 数,10.85 年 vs. 10.89 年, 表 1 ベースラインおよび追跡調査における認知機能,VRF,皮質厚,WMH のデータとその変化 N = 351 11 歳 小児期 IQ 73 歳 76 歳 3 年間の変化 102.21 ± 15.75 … … … 性別(%,男性) … 57.39 57.39 … 学歴,教育年数 … 10.86 ± 1.18 10.86 ± 1.18 高血圧(%,陽性者) … 46.31% 52.56% * 6.25%( z = 1.66, P = 0.049 ) 高コレステロール血症(%,陽性者) … 39.20% 46.02% * 6.82%( z = 1.83, P = 0.034 ) CVD の既往(%,陽性者) … 26.14% 32.67% 6.53%( z = 1.90, P = 0.029 ) * 喫煙(%,現喫煙者) … 6.53% 5.97% 糖尿病(%,陽性者) … 9.09% 12.22% … − 0.85%( z = − 0.31, P = 0.378 ) 3.13%( z = 1.34, P = 0.090 ) BMI … 27.57 ± 4.25 27.56 ± 4.46 − 0.02 ± 1.49( t = − 0.24, P = 0.812) MMSE … 28.97 ± 1.26 28.72 ± 1.46 * − 0.25 ± 1.48( t = − 3.01, P = 0.003) 全体の平均皮質厚,mm … 3.17 ± 0.15 3.12 ± 0.15 * − 0.05 ± 0.11( t = − 8.18, P < 0.001) WMH 容積( ICV に対する%) … 0.76 ± 0.71 10.2 ± 0.91 * 0.26 ± 0.28( t = 17.50, P < 0.001 ) BMI:肥満指数,CVD:心血管疾患,ICV:頭蓋内容積,IQ:知能指数,MMSE:ミニメンタルステート検査,WMH:白質高信号病変。 *P < 0.05。 PDLQLQGG 30 5 地域在住の高齢者において白質病変の進行と皮質の菲薄化は関連しない t =− 0.448,P = 0.65 ),11 歳時の知能指数( 101.9 vs. (表 1) 。全脳の変化と独 化に関する Cohen の d = 0.93 ) t = 1.24, P = 0.22)に有意差は認められなかった。 100.2, 立変数との関係を示す Spearman 相関マトリックス( 表 しかし,再来院した被験者群のベースラインの MMSE 2 )から,一般に高齢の被験者ほど,皮質の菲薄化がよ ス コ ア は, 再 来 院 し な か っ た 被 験 者 群 よ り 高 か っ た り顕著であることが明らかになった( ρ=− 0.13,P = 。 ( 28.9 vs. 28.6,t = 2.49,P = 0.01) 0.02 ) 。その他の独立変数ではいずれにおいても,WMH 以降の結果は,ベースラインと追跡調査の臨床データ, 認知機能データ,脳 MRI 検査データが揃っている縦断 および皮質菲薄化に対する効果( 時点の効果を超越 )は 限定的で部分的であった( 表 2 ) 。 的評価の対象被験者 351 例のみに関するものである。 ベースラインおよび追跡調査における認知機能, VRF,皮質厚,WMH のデータとその変化量 ベ ー ス ラ イ ン お よ び 追 跡 調 査 に お け る 認 知 機 能, 全体の平均皮質厚と WMH 容積との横断的およ び縦断的な総合相関性( global correlation) 全体の平均皮質厚と WMH 容積の横断的な総合相関性 ( global correlation )は,73 歳時( r =− 0.06,P = 0.27) VRF,皮質厚,WMH のデータおよびその変化を表 1 に および 76 歳時( r =− 0.08,P = 0.12 )ともに有意では 示す。高血圧,高コレステロール血症,心血管疾患の なかった。 ( 1 )73 歳時の WMH 容積と全体の平均皮質 ある被験者の比率は有意に増加し,MMSE についても ,( 2 )73 歳時の全体 厚の変化( r =− 0.07,P = 0.19 ) ベースライン( 73 歳時 )から追跡調査( 76 歳時 )までに の平均皮質厚と WMH 容積の変化( r = 0.01, P = 0.82), 有意な低下がみられた。全体において,平均皮質厚は ( 3 )WMH 容積の変化と全体の平均皮質厚の変化( r = 一般に加齢とともに減少し( 変化に関する Cohen の d = − 0.02,P = 0.67 )の対による縦断的な総合相関性に関 − 0.45 ) ,WMH 容積は一般に加齢とともに増加した( 変 しても,有意ではなかった。 表 2 全体の平均皮質厚の変化,WMH の変化,独立変数に関する Spearman 相関マトリックス rho ( P 値) Cort chng WMH chg BMI 性別 CVD DIAB HCHL HBP SMOK EDU IQ11 MMSE 年齢 Cort chng WMH chg 0.029 (0.595) BMI 性別 CVD − 0.056 − 0.043 (0.306) (0.441) 0.058 0.038 (0.292) (0.486) 0.017 (0.753) DIAB HCHL HBP EDU 0.058 0.025 (0.652) 0.050 0.055 (0.361) (0.317) 0.052 0.003 (0.962) − 0.051 0.060 年齢 0.074 − 0.148 0.149 − 0.078 * (0.159) (0.007) 0.106 0.003 (0.054) (0.956) 0.195 − 0.018 0.098 ( 0.076 ) 0.179 0.045 − 0.084 0.220 * ( < 0.001) * ( 0.001 ) 0.216 0.150 0.115 0.068 (0.417) (0.126) * ( 0.038 ) ( 0.218 ) − 0.171 0.068 0.024 ( 0.218 ) ( 0.658 ) − 0.019 − 0.084 ( 0.728 ) ( 0.126 ) − 0.029 − 0.142 − 0.006 − 0.072 (0.604) * (0.185) (0.010) ( 0.913 ) ( 0.191 ) ( 0.736 ) ( 0.756 ) ( 0.028 )* (< 0.001) * − 0.064 − 0.107 0.012 (0.835) 0.039 (0.483) 0.039 0.283 ( 0.006 ) * (< 0.001) * * (0.482) (0.002) 0.035 * (0.020) 0.088 * (0.109) (0.007) (0.183) (0.529) − 0.128 (0.964) * (0.747) (< 0.001) * (0.855) (< 0.001) 0.025 (0.273) MMSE − 0.010 (0.648) (0.355) IQ11 (0.370) (0.295) (0.350) SMOK − 0.050 − 0.002 0.073 0.104 0.011 − 0.020 ( 0.846 ) ( 0.717 ) 0.019 − 0.017 − 0.043 − 0.061 − 0.035 (0.250) (0.059) ( 0.432 ) ( 0.271 ) ( 0.521 ) ( 0.386 ) − 0.004 − 0.015 0.086 (0.941) (0.119) ( 0.783 ) 0.022 ( 0.692 ) 0.011 − 0.048 − 0.003 ( 0.849 ) ( 0.951 ) − 0.086 ( 0.119 ) − 0.121 0.528 0.234 0.378 * (< 0.001) * ( 0.052 ) ( < 0.001) − 0.048 − 0.009 − 0.014 ( 0.387 ) ( 0.864) ( 0.802 ) ( 0.164 ) − 0.077 BMI:肥満指数,Cort chng:73 歳から 76 歳における全体の平均皮質厚の変化,CVD:心血管疾患,DIAB:糖尿病,EDU:学歴,HCHL:高コレステロール,HBP:高血圧, IQ11:11 歳時の知能指数,MMSE:ミニメンタルステート検査,SMOK:喫煙,WMH chng:73 歳から 76 歳における白質高信号病変の変化。 *P < 0.05。 PDLQLQGG 30 6 Stroke 日本語版 Vol. 11, No. 1 76 歳 C-C ‒2 0 2 4 gC-wB ‒2 0 2 4 ‒2 0 2 4 wC-gB ‒2 0 2 4 皮質厚と全体の白質高信号病変( WMH )容積との皮質頂点による関連性を示す横断的( 76 歳 )および縦断的( C-C,gC-wB, wC-gB )t マップ。暖色は,WMH 容積の増加が皮質厚の減少に関連する領域を示す。これらの関連の有意性は図 3 に示す。C-C: 図2 73 歳から 76 歳における皮質厚の変化と WMH 容積の変化,gC-wB:皮質厚の変化と 73 歳時の WMH 容積,wC-gB:WMH 容 積の変化と 73 歳時の皮質厚。 局所の皮質厚および WMH 容積の横断的な皮質 を頂点とする回帰モデル 76 歳時の皮質厚と WMH 容積に関する横断的な皮質 76 歳時のシルビウス裂と周辺部における最大の横断 的関連を表した散布図( ρ=− 0.276,Q = 0.004 )を図 3 に示す。 頂点回帰モデルを図 2 に示す。皮質厚と WMH 容積との 関連性のパターンは 73 歳時と 76 歳時でほぼ同じであっ たため,73 歳時の横断データは示していない。すべて 局所の皮質厚および WMH 容積の縦断的な皮質 を頂点とする回帰モデル のモデルにおいて VRF,MMSE,学歴,性別による補 ( 1 )ベースラインの WMH 容積と皮質厚の変化( 図 2 正を行った。11 歳時の知能指数を調整変数に追加した の gC-wB ) ,( 2 )ベースラインの皮質厚と WMH 容積の ほとんどみられなかっ ことによる皮質の t マップの差は, 変化( 図 2 の wC-gB ) ,( 3 )WMH 容積の変化と皮質厚 た( データ非表示)。 の変化( 図 2 の C-C )について,皮質頂点の縦断的関連 図 2 の暖色表示は,WMH 容積の増加に関連して皮質 Q > 0.05)。したがっ 性はすべて有意ではなかった( 図 3, 厚の減少が認められた領域を示す。横断的関連の有意性 て,73 歳時の WMH 容積( または 73 歳時から 76 歳時の を図 3 の左側に示す。73 歳および 76 歳の時点において, WMH 容積の変化量 )が大きいことにより,73 歳から 76 シルビウス裂の内部から周辺部で頭頂葉,後頭葉,前頭 歳におけるいずれかの皮質領域の菲薄化が顕著であるこ 葉方向へと広がる負の横断的関連性の一定パターンが認 とは予測されなかった。また,73 歳時の皮質がより薄 められた。したがって,WMH 容積の増加が特定領域, いことにより,73 歳から 76 歳までにおける WMH の拡 すなわちシルビウス裂の内部および周辺部の皮質厚減少 大は予測されなかった。 と横断的に関連していた。上方領域( 図 2 の寒色 )では, WMH の変化と全体の平均皮質厚の変化との間の縦断 WMH 容積の増加と皮質厚の増加との関連はすべて有意 的関連は,記述上,MMSE 26 以下の被験者で大幅に顕 ではなかった。 著であったが( r =− 0.220,P = 0.41 vs. r =− 0.003, PDLQLQGG 30 地域在住の高齢者において白質病変の進行と皮質の菲薄化は関連しない X-S L 0.05 0.04 0.03 0.02 0.01 0 0.05 0.04 0.03 0.02 0.01 0 Q Q L 4 t =4.74 rho=‒0.276 Q=0.004 3.5 3 2.5 皮質厚の変化(mm) 76 歳時の皮質厚(mm) X-S 7 1 t =3.85 rho=‒0.133 Q=0.465 0.5 0 ‒0.5 ‒1 ‒1.5 2 0.01 0 0.02 0.03 0.04 0.05 76 歳時のWMH 容積(/ICV) ‒2 ‒0.005 0 0.005 0.01 0.015 0.02 0.025 WMH 容積(/ICV)の変化 皮質厚と白質高信号病変( WMH )容積との皮質頂点による横断的( X-S )および縦断的( L )関連の有意性。73 歳(データ非表示)と 図 3 76 歳の時点において,シルビウス裂の内部および周辺部で頭頂葉,後頭葉,前頭葉方向へと広がる負の横断的関連の一定パターンが 認められた( 左 ) 。無着色のグレーの Q マップ(右)は,皮質厚と WMH 容積との間に有意な縦断的関連がみられないことを示す。 P = 0.96 ),MMSE 26 以下の被験者が少なかったこと る原因が関与しており,それが横断的関連をもたらして が潜在的理由となり( N = 16 ) ,統計学的な有意差は認 いることが示唆される。一方,73 歳から 76 歳に生じる められなかった。 これらの変化には関連がないとみられ,もし,こうした 相関的な変化があるとすれば,因果関係を示す 1 つの指 標として認められたはずである。この結論は CADASIL 考 察 患者の縦断的研究と一致し,この以前の研究では,ラク 本研究では,73 ∼ 76 歳の地域在住の被験者において, ナ病変と皮質の形態の変化に強い関連を認めながらも, WMH 容積の増加とシルビウス裂周辺領域の皮質菲薄化 皮質の形態の変化と WMH 容積との間にはわずかな関連 との間に横断的関連が再現された 3,5,9-12。しかし,皮質 しか認められなかった 13,25。 のどの部分においても,( 1 )ベースラインの WMH 容積 本研究の強みは,地域在住の大規模な被験者集団にお と皮質厚の変化,( 2 )ベースラインの皮質厚と WMH 容 いて,WMH 容積と局所皮質厚との縦断的および横断的 積の変化,(3)WMH 容積の変化と皮質厚の変化につい 関連を検証できたことである。もう 1 つの強みは,小 て,縦断的関連は認められなかった。本研究で認められ 児期に知能指数検査を受けており,本研究時点で認知症 た,WMH 容積の増加とシルビウス裂領域の皮質厚の減 リスクが大きく高まる 70 代に入った同年齢の被験者を 10 少との横断的関連は,以前に行われた GM 容積 ,ボク セルベースの形態計測 5,11 3 対象としたことである 26。年齢範囲が狭いこと以外に, ,皮質厚 に関する研究の結果 LBC1936 研究の他の目新しい特徴( 全被験者が白人な と一致する。過去の研究と同様に,本研究で皮質の菲薄 ど )も,潜在的な強い交絡因子を最小限に抑えたと考え 化と WMH との関連が確認された領域は,WMH の多発 られる( 均質性の低い被験者集団では,年齢,民族の混 4 部位およびその拡大範囲 ( 脳室から遠心状に広がる領 合,地理といった因子が影響を及ぼした可能性がある ) 。 域や頭頂方向など )とは異なる。 WMH 容積と皮質厚の測定には十分に検証された定量方 本研究の結果から,WMH 容積と皮質萎縮はいずれも 法を使用し,手作業で確認の上,パイプライン処理後に 加齢とともに悪化し,これらの個人差には一部の共通す 両時点および各被験者で品質管理を行った 17,20,22。WMH PDLQLQGG 30 8 Stroke 日本語版 Vol. 11, No. 1 容積と皮質厚の測定に用いた脳 MRI の元画像は,両時 よるものか否かを評価する必要がある。最後に,WMH 点で同じプロトコルに従い,厳密に管理されている同じ と皮質の菲薄化との因果関係の有無を,観察研究で証明 画像診断装置で撮像した 16。 するのは困難である。しかし,本研究では,WMH と皮 以上の強みに加え,これまでの横断的研究結果を再現 したとはいえ 3,5,10,11 質の菲薄化に影響することが知られている多くの変数に ,本研究にも限界がある。3 年間の ついて補正した。また,相関は必ずしも因果関係を意味 追跡調査期間は WMH と皮質の菲薄化との相関的な変 しないが,因果関係の基礎である 32,33。したがって,73 化を検出するのに十分でなかった可能性があり,これ 歳から 76 歳において縦断的関連が認められなかったこ が大きな限界の 1 つである。被験者をできるだけ維持 とは,因果関係が存在しないことを示唆している。 し,Austrian Stroke Prevention Study などの過去の研 以上の限界はあるものの,両者について加齢による悪 究 1,27 との一致性もできるだけ高めるため,本研究では 化がみられ,一方,73 歳から 76 歳において WMH 容積 追跡調査期間を( より長期ではなく )3 年間とした。ま の進行と局所皮質の菲薄化に関し,相関性/因果関係を た,我々は以前に,WMH LBC1936 研究の狭い年齢範 示す縦断的関連はないと考えられた。さらに長い期間や 28 囲により( 3 歳未満 ) ,WMH に横断的な差を検出し 異なるライフステージでの因果関係の有無を明らかにす ている。現在再びこれらの被験者を対象に 6 年後時点 るためには,追跡調査期間を延ばし,評価時点の年齢を の追跡調査を実施中であるが,これによって,本研究で より多く設定した縦断的研究が必要である。WMH の拡 特定されなかった潜在的な因果関係の良好なエビデン 大と皮質の菲薄化の根本原因はまだ完全には解明されて スが明らかになる可能性がある。今後は,追跡不能例の いない。 理由を確認の上,完全な情報による最大尤度の解析法を 使用し,試験完了例の解析と照らし合わせることで,追 謝辞 跡不能例の影響を最小限に抑える予定である。本研究で LBC1936 研 究( 詳 細:http://www.lothianbirthcohort. は,「 76 歳時の各測定値」−「73 歳時の各測定値 」を「 変 ed.ac.uk/ )にご協力いただいた資金提供者,被験者,研 化 」と定義した。我々は,これ以外にも変化の評価方法 究施設,臨床スタッフおよび運営スタッフに感謝する。 は,認知機能の変数の変化によく使用されるもの 29 を 含め,複数あることを認識している。しかし,本研究で 使用したアプローチは,脳の形態変化の測定にしばし 30 研究費の財源 本 研 究 は Scottish Funding Council Early Career ば使用されている 。LBC1936 研究のコホートは均質 Researcher か ら Scottish Imaging Network̶A Platform 性が高く,年齢や民族などの交絡変数を調整する必要性 for Scientific Excellence( http: / / www. sinapse. ac. uk, が低いため,検出力は高いが,本研究結果の一般化でき DAD )への助成金,Research into Ageing プログラム助 る可能性は制限されている。本研究で縦断的に評価し 成金( Dr Deary および Dr Starr ) ,Age UK の支援する た被験者は全般に,追跡調査のため再来院しなかった Disconnected Mind プロジェクト( Dr Deary,Dr Starr, 被験者に比べ,MMSE が高かった。例えば,認知機能 Dr Wardlaw )による支援を受けた。また,UK Medical スコアが低い被験者では,WMH と皮質の菲薄化との間 Research Council( Dr Deary,Dr Starr,Dr Wardlaw, に,より強い縦断的関連を認める可能性があり,これも Dr M.E. Bastin) ,Scottish Funding Council から Scottish 本研究結果を一般化できる可能性を低下させていること Imaging Network̶A Platform for Scientific Excellence が考えられる。本研究では,MMSE 26 以下の被験者数 への助成金(Dr Wardlaw)による支援を受けた。 ( N = 16 )が少なかったため,この点を十分に検証する ことはできなかった。今後の研究では,認知機能が低 情報開示 下した被験者で関連がより強くなるか否かを確認する Dr Wardlaw は,LBC1936 研究およびさまざまな画 必要がある。本研究(および他の研究 3,5,9-12 )で WMH と 像研究プロジェクトへの協力に対し,Medical Research 皮質の菲薄化との間に横断的関連が認められた位置は, Council, Age UK, Row Fogo Charitable Trust, WMH の拡大がよくみられる領域を直接反映していな Scottish Funding Council か ら エ ジ ン バ ラ 大 学 へ の 金 いが,関連が示された位置は,側頭葉へと下方および 銭( 助成金 )の支払いがあったことを報告している。Dr 前方に延びる脳梁神経束の壁板( tapetum )に最も近かっ Deary は,LBC1936 研 究 へ の 協 力 に 対 し,Medical 31 た 。今後の研究により,WMH と皮質の菲薄化との間 Research Council および Age UK からエジンバラ大学へ に横断的関連が認められた位置について,その理由がシ の金銭(助成金 )の支払いがあったことを報告している。 ルビウス裂周囲の皮質および脳梁の壁板の間接的連絡に Dr Deary は,Medical Research Council の委員としての PDLQLQGG 30 地域在住の高齢者において白質病変の進行と皮質の菲薄化は関連しない 活動に対し,本人への金銭の支払いがあったことを報告 している。その他の著者からの情報開示はない。 References 1. Schmidt R, Enzinger C, Ropele S, Schmidt H, Fazekas F; Austrian Stroke Prevention Study. Progression of cerebral white matter lesions: 6-year results of the Austrian Stroke Prevention Study. Lancet. 2003;361:2046–2048. 2. Sowell ER, Peterson BS, Thompson PM, Welcome SE, Henkenius AL, Toga AW. Mapping cortical change across the human life span. Nat Neurosci. 2003;6:309–315. doi: 10.1038/nn1008. 3. Tuladhar AM, Reid AT, Shumskaya E, de Laat KF, van Norden AG, van Dijk EJ, et al. Relationship between white matter hyperintensities, cortical thickness, and cognition. Stroke. 2015;46:425–432. doi: 10.1161/ STROKEAHA.114.007146. 4. Wardlaw JM, Allerhand M, Doubal FN, Valdes Hernandez M, Morris Z, Gow AJ, et al. Vascular risk factors, large-artery atheroma, and brain white matter hyperintensities. Neurology. 2014;82:1331–1338. doi: 10.1212/WNL.0000000000000312. 5. Wen W, Sachdev PS, Chen X, Anstey K. Gray matter reduction is correlated with white matter hyperintensity volume: a voxel-based morphometric study in a large epidemiological sample. Neuroimage. 2006;29:1031–1039. doi: 10.1016/j.neuroimage.2005.08.057. 6. Debette S, Markus HS. The clinical importance of white matter hyperintensities on brain magnetic resonance imaging: systematic review and meta-analysis. BMJ. 2010;341:c3666. 7. Lawrence AJ, Chung AW, Morris RG, Markus HS, Barrick TR. Structural network efficiency is associated with cognitive impairment in small-vessel disease. Neurology. 2014;83:304–311. doi: 10.1212/WNL.0000000000000612. 8. Thompson PM, Hayashi KM, de Zubicaray G, Janke AL, Rose SE, Semple J, et al. Dynamics of gray matter loss in Alzheimer’s disease. J Neurosci. 2003;23:994–1005. 9. Godin O, Maillard P, Crivello F, Alpérovitch A, Mazoyer B, Tzourio C, et al. Association of white-matter lesions with brain atrophy markers: the three-city Dijon MRI study. Cerebrovasc Dis. 2009;28:177–184. doi: 10.1159/000226117. 10. Knopman DS, Griswold ME, Lirette ST, Gottesman RF, Kantarci K, Sharrett AR, et al; ARIC Neurocognitive Investigators. Vascular imaging abnormalities and cognition: mediation by cortical volume in nondemented individuals: atherosclerosis risk in communities-neurocognitive study. Stroke. 2015;46:433–440. doi: 10.1161/STROKEAHA.114.007847. 11. Raji CA, Lopez OL, Kuller LH, Carmichael OT, Longstreth WT Jr, Gach HM, et al. White matter lesions and brain gray matter volume in cognitively normal elders. Neurobiol Aging. 2012;33:834.e837–834.e816 12. Rossi R, Boccardi M, Sabattoli F, Galluzzi S, Alaimo G, Testa C, et al. Topographic correspondence between white matter hyperintensities and brain atrophy. J Neurol. 2006;253:919–927. doi: 10.1007/ s00415-006-0133-z. 13. Jouvent E, Mangin JF, Duchesnay E, Porcher R, Düring M, Mewald Y, et al. Longitudinal changes of cortical morphology in CADASIL. Neurobiol Aging. 2012;33:1002.e29–1002.e36. doi: 10.1016/j. neurobiolaging.2011.09.013. 14. Deary IJ, Gow AJ, Taylor MD, Corley J, Brett C, Wilson V, et al. The Lothian Birth Cohort 1936: a study to examine influences on cognitive ageing from age 11 to age 70 and beyond. BMC Geriatr. 2007;7:28. doi: 10.1186/1471-2318-7-28. 15. Deary IJ, Gow AJ, Pattie A, Starr JM. Cohort profile: the Lothian Birth Cohorts of 1921 and 1936. Int J Epidemiol. 2012;41:1576–1584. doi: 10.1093/ije/dyr197. 16. Wardlaw JM, Bastin ME, Valdés Hernández MC, Maniega SM, Royle NA, Morris Z, et al. Brain aging, cognition in youth and old age and vascular disease in the Lothian Birth Cohort 1936: rationale, design and PDLQLQGG 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 9 methodology of the imaging protocol. Int J Stroke. 2011;6:547–559. doi: 10.1111/j.1747-4949.2011.00683.x. Hernández Mdel C, Ferguson KJ, Chappell FM, Wardlaw JM. New multispectral MRI data fusion technique for white matter lesion segmentation: method and comparison with thresholding in FLAIR images. Eur Radiol. 2010;20:1684–1691. doi: 10.1007/s00330-010-1718-6. Wang X, Valdés Hernández MC, Doubal F, Chappell FM, Wardlaw JM. How much do focal infarcts distort white matter lesions and global cerebral atrophy measures? Cerebrovasc Dis. 2012;34:336–342. doi: 10.1159/000343226. Wardlaw JM, Smith EE, Biessels GJ, Cordonnier C, Fazekas F, Frayne R, et al; STandards for ReportIng Vascular changes on nEuroimaging (STRIVE v1). Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. Lancet Neurol. 2013;12:822–838. doi: 10.1016/S1474-4422(13)70124-8. Ad-Dab’bagh Y, Lyttelton O, Muehlboeck J, Lepage C, Einarson D, Mok K, et al. The Civet image-processing environment: A fully automated comprehensive pipeline for anatomical neuroimaging research. Poster presented at: Proceedings of the 12th Annual Meeting of the Organization for Human Brain Mapping. June 11–15, 2006; Florence, Italy. Zijdenbos AP, Forghani R, Evans AC. Automatic “pipeline” analysis of 3-D MRI data for clinical trials: application to multiple sclerosis. IEEE Trans Med Imaging. 2002;21:1280–1291. doi: 10.1109/ TMI.2002.806283. Karama S, Ad-Dab’bagh Y, Haier R, Deary IJ, Lyttelton OC, Lepage C, et al. Positive association between cognitive ability and cortical thickness in a representative us sample of healthy 6 to 18 year-olds. Intelligence. 2009;37:145–155. Karama S, Ducharme S, Corley J, Chouinard-Decorte F, Starr JM, Wardlaw JM, et al. Cigarette smoking and thinning of the brain’s cortex. Mol Psychiatry. 2015;20:778–785. doi: 10.1038/mp.2014.187. Genovese CR, Lazar NA, Nichols T. Thresholding of statistical maps in functional neuroimaging using the false discovery rate. Neuroimage. 2002;15:870–878. doi: 10.1006/nimg.2001.1037. Jouvent E, Mangin JF, Porcher R, Viswanathan A, O’Sullivan M, Guichard JP, et al. Cortical changes in cerebral small vessel diseases: a 3D MRI study of cortical morphology in CADASIL. Brain. 2008;131(pt 8):2201–2208. doi: 10.1093/brain/awn129. Matthews F, Brayne C; Medical Research Council Cognitive Function and Ageing Study Investigators. The incidence of dementia in England and Wales: findings from the five identical sites of the MRC CFA Study. PLoS Med. 2005;2:e193. doi: 10.1371/journal.pmed.0020193. Schmidt R, Fazekas F, Kapeller P, Schmidt H, Hartung HP. MRI white matter hyperintensities: three-year follow-up of the Austrian Stroke Prevention Study. Neurology. 1999;53:132–139. Aribisala BS, Morris Z, Eadie E, Thomas A, Gow A, Valdés Hernández MC, et al. Blood pressure, internal carotid artery flow parameters, and age-related white matter hyperintensities. Hypertension. 2014;63:1011– 1018. doi: 10.1161/HYPERTENSIONAHA.113.02735. Deary IJ, Allerhand M, Der G. Smarter in middle age, faster in old age: a cross-lagged panel analysis of reaction time and cognitive ability over 13 years in the West of Scotland Twenty-07 Study. Psychol Aging. 2009;24:40–47. doi: 10.1037/a0014442. Veldink JH, Scheltens P, Jonker C, Launer LJ. Progression of cerebral white matter hyperintensities on MRI is related to diastolic blood pressure. Neurology. 1998;51:319–320. Wakana S, Jiang H, Nagae-Poetscher LM, van Zijl PC, Mori S. Fiber tract-based atlas of human white matter anatomy. Radiology. 2004;230:77–87. doi: 10.1148/radiol.2301021640. Aldrich J. Correlations genuine and spurious in pearson and yule. Stat Sci. 1995:364–376. Pearson K, Lee A, Yule G. On the distribution of frequency (variation and correlation) of the barometric height at divers stations. Philos Trans R Soc Lond A. 1897:423–469. 30