r Human Brain Mapping 000:000���000 (2011) r Cortical Thickness is Linked to Executive Functioning in Adulthood and Aging Agnieszka Z. Burzynska,1* Irene E. Nagel,2 Claudia Preuschhof,2 Sebastian Gluth,1 Lars Backman,1,3 �� Shu-Chen Li,1 Ulman Lindenberger,1 and Hauke R. Heekeren1,2,4 1The Center for Lifespan Psychology, Max Planck Institute for Human Development, Lentzeallee 94, 14195 Berlin, Germany 2Department of Educational Science and Psychology, Freie Universitat �� Berlin, 14195 Berlin, Germany 3Aging Research Center, Karolinska Institute, Gavlegatan �� 16, 11330, Stockholm, Sweden 4Max Planck Institute for Human Cognitive and Brain Sciences, Stephanstra e 1a, 04103 Leipzig, Germany r r Abstract: Executive functions that are dependent upon the frontal-parietal network decline consider- ably during the course of normal aging. To delineate neuroanatomical correlates of age-related execu- tive impairment, we investigated the relation between cortical thickness and executive functioning in 73 younger (20���32 years) and 56 older (60���71 years) healthy adults. Executive functioning was assessed using the Wisconsin Card Sorting Test (WCST). Cortical thickness was measured at each location of the cortical mantle using surface-based segmentation procedures on high-resolution T1-weighted mag- netic resonance images. For regions involved in WCST performance, such as the lateral prefrontal and parietal cortices, we found that thicker cortex was related to higher accuracy. Follow-up ROI-based analyses revealed that these associations were stronger in older than in younger adults. Moreover, among older adults, high and low performers differed in cortical thickness within regions generally linked to WCST performance. Our results indicate that the structural cortical correlates of executive functioning largely overlap with previously identified functional patterns. We conclude that structural preservation of relevant brain regions is associated with higher levels of executive performance in old age, and underscore the need to consider the heterogeneity of brain aging in relation to cognitive func- tioning. Hum Brain Mapp 00:000���000, 2011. V C 2011 Wiley-Liss, Inc. Key words: WCST right DLPFC prefrontal cortex parietal cortex fronto-parietal network structure- function relationship performance level lifespan healthy aging r r Additional Supporting Information may be found in the online version of this article. Contract grant sponsors: The Max Planck Society Contract grant number: M.FE.A.BILD0005 Contract grant sponsor: The German Federal Ministry for Research Contract grant number: 01GO0501 Contract grant sponsor: The German Research Council Contract grant sponsor: The Swedish Research Council Contract grant number: 521-2007-2829 Contract grant sponsors: Swedish Brain Power Alexander von Humboldt Research Award Jochnick Foundation The Center for Advanced Study in the Behavioral Sci- ences at Stanford University The International Max Planck Research School on The Life Course: Evolutionary and Ontoge- netic Dynamics (LIFE). *Correspondence to: Agnieszka Z. Burzynska, Max Planck Insti- tute for Human Development, Lentzeallee 94, D-14195 Berlin, Ger- many. E-mail: burzynska@mpib-berlin.mpg.de Received for publication 10 January 2010 Revised 15 November 2010 Accepted 18 February 2011 DOI: 10.1002/hbm.21311 Published online in Wiley Online Library(wileyonlinelibrary.com). V C 2011 Wiley-Liss, Inc.

INTRODUCTION Normal aging is accompanied by decline in many cogni- tive domains, especially those related to fluid abilities [Baltes, 1987 Cattell, 1971 Horn, 1968 Lindenberger, 2001], such as reasoning and executive functioning [Back- �� man et al., 2000 Li et al., 2004 Rhodes, 2004 West, 1996]. Executive functions comprise several cognitive abilities, such as working memory, selective attention, inhibition, set shifting, and task switching [Miyake et al., 2000 Teuber 1972]. Conjointly, these functions enable flexible and goal-directed behavior in a changing environment. Age-related decline in executive functioning may impose difficulties in everyday situations, such as social contacts, street traffic, and diet or drug compliance. To assess executive functioning, we administered the Wisconsin Card Sorting Test (WCST), a standard neuro- psychological test, which requires attention, feedback- based updating of information in working memory, inhibi- tion of prepotent responses, and shifting of mental set [Grant and Berg 1948 Heaton et al., 1993]. In healthy adults, performing the WCST results in increased func- tional brain activity within a distributed fronto-parietal network [see Buchsbaum et al., 2005 Nyhus and Barcelo, 2009, for reviews]. The right prefrontal (PFC) and parietal cortices are most consistently recruited during the WCST, with less consistent findings regarding activation of the temporal and occipital lobe, as well as subcortical gray matter [Buchsbaum et al., 2005 Nyhus and Barcelo, 2009]. Parallel to cognitive decline, advancing age is related to a structural degeneration of the cerebral cortex as indi- cated by widespread reductions in cortical thickness [Fjell et al., 2009 Hutton et al., 2009 Salat et al., 2004], gray mat- ter density [Sowell et al., 2003], and gray matter volume [Courchesne et al., 2000 Good et al., 2001 Raz et al., 1997, 2000, 2005 Walhovd et al., 2009]. So far, attempts at link- ing age-related structural alterations to executive function- ing have yielded mixed results. Fjell et al. [2006], Ziegler et al. [2010], and Van Petten et al. [2004] reported no corre- lation between cortical thickness and composite scores of various executive measures, whereas Kochunov et al. [2009] found that thicker cortex in many regions was related to better executive performance in adults between 30 and 90 years, but not in adults between 19 and 26 years. These correlations, however, were no longer signifi- cant after controlling for age [Kochunov et al., 2009]. In 87 participants aged 20 to 77 years, Raz et al. [2008] observed that, after accounting for the effects of age, sex, and vascu- lar risk factors, the orbitofrontal cortex and the prefrontal white-matter volumes as well as the 5-year change in ento- rhinal cortex volume predicted fluid intelligence level, assessed by Cattell Culture Fair Intelligence and Letter Sets tests. Hartberg et al. [2010] found that thicker cortex in superior temporal, superior frontal, and inferior frontal gyri was related to fewer perseverative errors in the WCST among schizophrenic patients and healthy adults 20���56 years old. Elderkin-Thompson et al. [2008] found both positive and negative correlations between volume of the frontal lobe gyri and various executive function scores in adults aged 61���88 years. Specifically, greater anterior cingulate volume was associated with less time needed to complete the Stroop Inhibition Task, whereas larger orbito- frontal volumes were associated with poorer verbal flu- ency. Other studies show that, in older adults, smaller PFC gray-matter volume [Gunning-Dixon and Raz, 2003] and smaller right frontal-lobe volume [Hanninen et al., 1997] were related to a higher number of perseverative errors in the WCST, and that larger frontal-lobe volume was positively linked to the number of completed WCST categories [Schretlen et al., 2000]. At least some of the above findings suggest that struc- ture-function relationships are more pronounced in late than in early adulthood. A healthy mature brain may func- tion above a critical threshold despite some inter-individ- ual variability in brain structure. Conversely, in the presence of biological constraints on information process- ing mechanisms (i.e., when decline in biological resources reaches a certain threshold), such as during aging, the extent of decline in brain structure may be more predictive of cognitive performance [Almeida et al., 2008 Li et al., 2004 Lindenberger et al., 2008 Nagel et al., 2008 Sullivan and Pfefferbaum, 2006]. Despite the associations described above, many studies using the WCST have reported negative results. This might reflect relatively small sample sizes, Elderkin- Thompson et al., 2008, n �� 23 Gur et al., 1998, n �� 17 Sanfilipo et al., 2002, n �� 27, or stem from combining the WCST with other tests, whose neural substrates may not be overlapping [Van Petten et al., 2004]. In addition to issues regarding sample size and use of composite scores of executive functioning, the studies reporting negative results are affected by at least one of the following limita- tions: (a) use of volumetric measures, which confound cortical thickness, surface area, and folding [Im et al., 2006a,b] a direct comparison of volumetric measures based on voxel-based morphometry and cortical thickness measures in 48 healthy adults aged 22���60 years revealed that volumetric measures were less sensitive, had lower signal-to-noise ratio, lower T-scores, and were more con- founded by overall brain size than thickness measures [Hutton et al. 2009] (b) gross definition of frontal volume or frontal gyri, pooling together functionally dissociable areas, such as the dorsal and ventrolateral PFC, the orbito- frontal cortex, and premotor areas and (c) investigation of frontal-lobe volume only, although the WCST is no longer considered a specific test of prefrontal function [Nyhus and Barcelo, 2009]. In this study, we investigated the structural correlates of executive functioning in the adult human brain. Specifi- cally, our goal was twofold: (a) to better characterize the relations between cortical thickness of brain regions associ- ated with the WCST (as defined in previous functional neuroimaging studies) [Buschbaum et al., 2005 Nyhus and Barcelo, 2009] and executive functioning and (b) to r Burzynska et al. r r 2 r