Cerebral Cortexdoi:10.1093/cercor/bhp026Distinct Genetic Influences on CorticalSurface Area and Cortical ThicknessMatthew S. Panizzon1, Christine Fennema-Notestine1,2, LisaT. Eyler1,3, Terry L. Jernigan1,2, Elizabeth Prom-Wormley4,Michael Neale4, Kristen Jacobson5, Michael J. Lyons6, MichaelD. Grant6, Carol E. Franz1, Hong Xian7, Ming Tsuang1,8,9,Bruce Fischl10, Larry Seidman11, Anders Dale1,12and WilliamS. Kremen1,3,81Department of Psychiatry,2Department of Radiology,University of California, San Diego, La Jolla, CA 92093, USA,3SanDiego Veterans Administration Health Care System, San Diego,CA 92161, USA,4Virginia Institute for Psychiatric andBehavioral Genetics, Virginia Commonwealth University Schoolof Medicine, Richmond, VA 23219, USA,5Department ofPsychiatry, University of Chicago, Chicago, IL 60637, USA,6Department of Psychology, Boston University, Boston, MA02215, USA,7Department of Psychiatry, Washington UniversitySchool of Medicine, St Louis, MO 63130, USA,8Center forBehavioral Genomics, University of California, San Diego, LaJolla, CA 92093, USA,9Harvard Institute of PsychiatricEpidemiology and Genetics, Harvard Medical School and Schoolof Public Health, Boston, MA 02115, USA,10Department ofRadiology, Massachusetts General Hospital, Boston, MA 02129,USA,11Department of Psychiatry, Harvard Medical School, Boston,MA 02115, USA and12Department of Neurosciences, Universityof California, San Diego, La Jolla, CA 92093, USANeuroimaging studies examining the effects of aging andneuropsychiatric disorders on the cerebral cortex have largelybeen based on measures of cortical volume. Given that corticalvolume is a product of thickness and surface area, it is plausiblethat measures of volume capture at least 2 distinct sets of geneticinfluences. The present study aims to examine the geneticrelationships between measures of cortical surface area andthickness. Participants were men in the Vietnam Era Twin Study ofAging (110 monozygotic pairs and 92 dizygotic pairs). Mean agewas 55.8 years (range: 51--59). Bivariate twin analyses wereutilized in order to estimate the heritability of cortical surface areaand thickness, as well as their degree of genetic overlap. Totalcortical surface area and average cortical thickness were bothhighly heritable (0.89 and 0.81, respectively) but were essentiallyunrelated genetically (genetic correlation 5 0.08). This pattern wassimilar at the lobar and regional levels of analysis. These resultsdemonstrate that cortical volume measures combine at least 2distinct sources of genetic influences. We conclude that usingvolume in a genetically informative study, or as an endophenotypefor a disorder, may confound the underlying genetic architecture ofbrain structure.Keywords: cortical volume, genetic correlation, heritability, magneticresonance imaging, twin studyIntroductionOver the past decade, the field of neuroimaging has madeextraordinary advances, leading to increased precision in howthe living brain can be measured. Although researchers wereonce limited to gross measures of volume or cross-sectionalsurface areas, contiguous maps of cortical thickness measure-ments can now be generated, greatly expanding the realm ofhypotheses to be tested. Yet, as the approaches to measure-ment have changed, there has been remarkably little examina-tion of how these different methods of measuring brainstructure may capture different aspects of the underlyingneural architecture.Nowhere is this trend more evident than in neuroimagingstudies examining the effects of aging and neuropsychiatricdisorders on the cerebral cortex. Until relatively recently, thesestudies have largely been based on measures of volume (forreviews, see Wright et al. 2000; Sheline 2003; Zakzanis et al.2003; Raz et al. 2005; Ellison-Wright et al. 2008). Aided largelyby methods of high-resolution, semiautomated measurement,researchers have begun to examine cortical thickness asa relevant structural measure. However, examination of theliterature demonstrates that measures of volume and thicknessare often utilized interchangeably as descriptors of corticalstructure, whereas few studies have acknowledged or exam-ined how they may be related to or distinguished from oneanother (Dickerson et al. 2007; Goghari et al. 2007; Wang et al.2007; Fornito et al. 2008; Venkatasubramanian et al. 2008).By definition, volume is the product of surface area andthickness; therefore, measures of cortical volume could combinestructural properties that are unique to cortical surface area andunique to cortical thickness. The cellular architecture of thecortex itself suggests that this may be the case. It has been welldocumented that neurons within the cerebral cortex areorganized into ontogenetic columns that run perpendicular tothe surface of the brain (Mountcastle 1997). The radial unithypothesis of cortical development argues that the cells withina column share a common origin and migrate to their locationwithin the cortex during development (Rakic 1988, 1995, 2007).It also postulates that the size of the cortical surface area is drivenby the number of columns, whereas cortical thickness isinfluenced by the number of cells within a column (Rakic1988). A recent extension of the radial unit hypothesis, referredto as the radial amplification hypothesis, highlights the contri-bution of intermediate progenitor cells to neurogenesis and howtheir manipulation within animal models results in changes toÓ The Author 2009. Published by Oxford University Press. All rights reserved.For permissions, please e-mail: journals.permissions @oxf ordjour nals.o rg Cerebral Cortex Advance Access published March 18, 2009cortical thickness but not surface area (Pontious et al. 2008). Inaddition, studies of interindividual variation in adult brain sizehave found that differences in cortical gray matter volume aredriven almost exclusively by differences in the cortical surfacearea rather than cortical thickness, suggesting little covariationbetween the 2 measures (Pakkenberg and Gundersen 1997; Imet al. 2008). Such evidence suggests that surface area andthickness are distinct rather than redundant features of corticalstructure.If cortical surface area and cortical thickness are indeeddriven by distinct cellular mechanisms, then it is reasonable topropose that they will have distinct genetic etiologies. Twinstudies have clearly demonstrated significant genetic contribu-tions to variation in brain structure (for reviews, see Glahn et al.2007;