Advertisement

Genomic Studies Across the Lifespan Point to Early Mechanisms Determining Subcortical Volumes

  • Quentin Le Grand
    Affiliations
    University of Bordeaux, INSERM, Bordeaux Population Health Center, UMR1219, Bordeaux, France
    Search for articles by this author
  • Claudia L. Satizabal
    Affiliations
    Glenn Biggs Institute for Alzheimer’s and Neurodegenerative Diseases, UT Health San Antonio, San Antonio, Texas

    Department of Population Health Sciences, UT Health San Antonio, San Antonio, Texas

    Framingham Heart Study, Framingham, Massachusetts

    Department of Neurology, Boston University School of Medicine, Boston, Massachusetts
    Search for articles by this author
  • Muralidharan Sargurupremraj
    Affiliations
    University of Bordeaux, INSERM, Bordeaux Population Health Center, UMR1219, Bordeaux, France

    Glenn Biggs Institute for Alzheimer’s and Neurodegenerative Diseases, UT Health San Antonio, San Antonio, Texas
    Search for articles by this author
  • Aniket Mishra
    Affiliations
    University of Bordeaux, INSERM, Bordeaux Population Health Center, UMR1219, Bordeaux, France
    Search for articles by this author
  • Aicha Soumaré
    Affiliations
    University of Bordeaux, INSERM, Bordeaux Population Health Center, UMR1219, Bordeaux, France
    Search for articles by this author
  • Alexandre Laurent
    Affiliations
    University of Bordeaux, Institute of Neurodegenerative Diseases, UMR5293, Neurofunctional imaging group, Bordeaux, France

    CNRS, Institute of Neurodegenerative Diseases, UMR5293, Neurofunctional imaging group, Bordeaux, France

    CEA, Institute of Neurodegenerative Diseases, UMR5293, Neurofunctional imaging group, Bordeaux, France
    Search for articles by this author
  • Fabrice Crivello
    Affiliations
    University of Bordeaux, Institute of Neurodegenerative Diseases, UMR5293, Neurofunctional imaging group, Bordeaux, France

    CNRS, Institute of Neurodegenerative Diseases, UMR5293, Neurofunctional imaging group, Bordeaux, France

    CEA, Institute of Neurodegenerative Diseases, UMR5293, Neurofunctional imaging group, Bordeaux, France
    Search for articles by this author
  • Ami Tsuchida
    Affiliations
    University of Bordeaux, Institute of Neurodegenerative Diseases, UMR5293, Neurofunctional imaging group, Bordeaux, France

    CNRS, Institute of Neurodegenerative Diseases, UMR5293, Neurofunctional imaging group, Bordeaux, France

    CEA, Institute of Neurodegenerative Diseases, UMR5293, Neurofunctional imaging group, Bordeaux, France
    Search for articles by this author
  • Jean Shin
    Affiliations
    Department of Physiology, Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada

    Department of Nutritional Sciences, Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Mélissa Macalli
    Affiliations
    University of Bordeaux, INSERM, Bordeaux Population Health Center, UMR1219, Bordeaux, France
    Search for articles by this author
  • Baljeet Singh
    Affiliations
    Imaging of Dementia and Aging Laboratory, Department of Neurology, University of California Davis, Davis, California
    Search for articles by this author
  • Alexa S. Beiser
    Affiliations
    Framingham Heart Study, Framingham, Massachusetts

    Department of Neurology, Boston University School of Medicine, Boston, Massachusetts

    Department of Biostatistics, Boston University School of Public Health, Boston, Massachusetts
    Search for articles by this author
  • Charles DeCarli
    Affiliations
    Imaging of Dementia and Aging Laboratory, Department of Neurology, University of California Davis, Davis, California
    Search for articles by this author
  • Evan Fletcher
    Affiliations
    Imaging of Dementia and Aging Laboratory, Department of Neurology, University of California Davis, Davis, California
    Search for articles by this author
  • Tomas Paus
    Affiliations
    Department of Psychiatry, University of Toronto, Toronto, Ontario, Canada

    Department of Psychology, University of Toronto, Toronto, Ontario, Canada

    Department of Psychiatry, Centre Hospitalier Universitaire Sainte-Justine, Faculty of Medicine, University of Montreal, Montreal, Quebec, Canada
    Search for articles by this author
  • Mark Lathrop
    Affiliations
    McGill Genome Center, McGill University, Montreal, Quebec, Canada
    Search for articles by this author
  • Hieab H.H. Adams
    Affiliations
    Department of Radiology and Nuclear Medicine, Erasmus University Medical Center, Rotterdam, the Netherlands

    Department of Clinical Genetics, Erasmus University Medical Center, Rotterdam, the Netherlands
    Search for articles by this author
  • Joshua C. Bis
    Affiliations
    Cardiovascular Health Research Unit, Department of Medicine, University of Washington, Seattle, Washington
    Search for articles by this author
  • Sudha Seshadri
    Affiliations
    Glenn Biggs Institute for Alzheimer’s and Neurodegenerative Diseases, UT Health San Antonio, San Antonio, Texas

    Department of Population Health Sciences, UT Health San Antonio, San Antonio, Texas

    Framingham Heart Study, Framingham, Massachusetts

    Department of Neurology, Boston University School of Medicine, Boston, Massachusetts
    Search for articles by this author
  • Christophe Tzourio
    Affiliations
    University of Bordeaux, INSERM, Bordeaux Population Health Center, UMR1219, Bordeaux, France

    Bordeaux University Hospital, Department of Medical Informatics, Bordeaux, France
    Search for articles by this author
  • Bernard Mazoyer
    Affiliations
    University of Bordeaux, Institute of Neurodegenerative Diseases, UMR5293, Neurofunctional imaging group, Bordeaux, France

    CNRS, Institute of Neurodegenerative Diseases, UMR5293, Neurofunctional imaging group, Bordeaux, France

    CEA, Institute of Neurodegenerative Diseases, UMR5293, Neurofunctional imaging group, Bordeaux, France

    Bordeaux University Hospital, Department of Neuroradiology, Bordeaux, France
    Search for articles by this author
  • Stéphanie Debette
    Correspondence
    Address correspondence to Stéphanie Debette, M.D., Ph.D.
    Affiliations
    University of Bordeaux, INSERM, Bordeaux Population Health Center, UMR1219, Bordeaux, France

    Bordeaux University Hospital, Department of Neurology, Institute of Neurodegenerative Diseases, Bordeaux, France
    Search for articles by this author
Published:October 23, 2021DOI:https://doi.org/10.1016/j.bpsc.2021.10.011

      Abstract

      Background

      Subcortical brain structures play a key role in pathological processes of age-related neurodegenerative disorders. Mounting evidence also suggests that early-life factors may have an impact on the development of common late-life neurological diseases, including genetic factors that can influence both brain maturation and neurodegeneration.

      Methods

      Using large population-based brain imaging datasets across the lifespan (N ≤ 40,628), we aimed to 1) estimate the heritability of subcortical volumes in young (18–35 years), middle (35–65 years), and older (65+ years) age, and their genetic correlation across age groups; 2) identify whether genetic loci associated with subcortical volumes in older persons also show associations in early adulthood, and explore underlying genes using transcriptome-wide association studies; and 3) explore their association with neurological phenotypes.

      Results

      Heritability of subcortical volumes consistently decreased with increasing age. Genetic risk scores for smaller caudate nucleus, putamen, and hippocampus volume in older adults were associated with smaller volumes in young adults. Individually, 10 loci associated with subcortical volumes in older adults also showed associations in young adults. Within these loci, transcriptome-wide association studies showed that expression of several genes in brain tissues (especially MYLK2 and TUFM) was associated with subcortical volumes in both age groups. One risk variant for smaller caudate nucleus volume (TUFM locus) was associated with lower cognitive performance. Genetically predicted Alzheimer’s disease was associated with smaller subcortical volumes in middle and older age.

      Conclusions

      Our findings provide novel insights into the genetic determinants of subcortical volumes across the lifespan. More studies are needed to decipher the underlying biology and clinical impact.

      Keywords

      Subcortical structures (hippocampus, caudate nucleus, putamen, pallidum, accumbens, amygdala, and thalamus) are involved in many neural processes, from autonomic and sensorimotor functions to memory, decision making, and processing of reward and threat signals (
      • Yin H.H.
      • Knowlton B.J.
      The role of the basal ganglia in habit formation.
      ,
      • McDonald A.J.
      • Mott D.D.
      Functional neuroanatomy of amygdalohippocampal interconnections and their role in learning and memory: Amygdalohippocampal Interconnections.
      ,
      • Hikosaka O.
      • Kim H.F.
      • Yasuda M.
      • Yamamoto S.
      Basal ganglia circuits for reward value–guided behavior.
      ,
      • Salzman C.D.
      • Fusi S.
      Emotion, cognition, and mental state representation in amygdala and prefrontal cortex.
      ,
      • Floresco S.B.
      The nucleus accumbens: An interface between cognition, emotion, and action.
      ). These structures also play a key role in pathological processes of age-related neurodegenerative disorders, including Parkinson’s disease (PD) and Alzheimer’s disease (AD). Volumes of subcortical structures can be measured quantitatively in large population-based samples using brain magnetic resonance imaging (MRI) and were shown to decrease with aging and neurodegenerative diseases (
      • Dima D.
      • Modabbernia A.
      • Papachristou E.
      • Doucet G.E.
      • Agartz I.
      • Aghajani M.
      • et al.
      Subcortical volumes across the lifespan: Data from 18,605 healthy individuals aged 3-90 years.
      ,
      • Coupé P.
      • Manjón J.V.
      • Lanuza E.
      • Catheline G.
      Lifespan changes of the human brain in Alzheimer’s disease.
      ,
      • de Jong L.W.
      • van der Hiele K.
      • Veer I.M.
      • Houwing J.J.
      • Westendorp R.G.J.
      • Bollen E.L.E.M.
      • et al.
      Strongly reduced volumes of putamen and thalamus in Alzheimer’s disease: An MRI study.
      ,
      • Mak E.
      • Su L.
      • Williams G.B.
      • Firbank M.J.
      • Lawson R.A.
      • Yarnall A.J.
      • et al.
      Baseline and longitudinal grey matter changes in newly diagnosed Parkinson’s disease: ICICLE-PD study.
      ,
      • Carriere N.
      • Besson P.
      • Dujardin K.
      • Duhamel A.
      • Defebvre L.
      • Delmaire C.
      • Devos D.
      Apathy in Parkinson’s disease is associated with nucleus accumbens atrophy: A magnetic resonance imaging shape analysis.
      ,
      • Möller C.
      • Dieleman N.
      • van der Flier W.M.
      • Versteeg A.
      • Pijnenburg Y.
      • Scheltens P.
      • et al.
      More atrophy of deep gray matter structures in frontotemporal dementia compared to Alzheimer’s disease.
      ,
      • Graff-Radford J.
      • Williams L.
      • Jones D.T.
      • Benarroch E.E.
      Caudate nucleus as a component of networks controlling behavior.
      ,
      • Pasquini J.
      • Durcan R.
      • Wiblin L.
      • Gersel Stokholm M.
      • Rochester L.
      • Brooks D.J.
      • et al.
      Clinical implications of early caudate dysfunction in Parkinson’s disease.
      ,
      • Luo X.
      • Mao Q.
      • Shi J.
      • Wang X.
      • Li C.-S.R.
      Putamen gray matter volumes in neuropsychiatric and neurodegenerative disorders.
      ,
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ).
      Mounting evidence suggests that factors already present at an early age may have an impact on the development of late-life neurological diseases, likely owing to a complex combination of genetic and environmental factors that influence both brain maturation and neurodegeneration (
      • Glahn D.C.
      • Kent J.W.
      • Sprooten E.
      • Diego V.P.
      • Winkler A.M.
      • Curran J.E.
      • et al.
      Genetic basis of neurocognitive decline and reduced white-matter integrity in normal human brain aging.
      ,
      • Hunter D.J.
      Gene–environment interactions in human diseases.
      ). Identifying early-life determinants of neurological diseases occurring in older age could therefore be crucial for developing more efficient prevention strategies. MRI-based measures of subcortical brain structures represent a powerful intermediate phenotype to decipher these early determinants, as they show alterations many years before the onset of clinical symptoms (
      • Jack C.R.
      • Holtzman D.M.
      Biomarker modeling of Alzheimer’s disease.
      ).
      Recently, genome-wide association studies (GWASs) have enable substantial progress in deciphering genes underlying variations in volumes of brain structures, identifying 38 independent genetic risk loci for subcortical volumes (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ,
      • Hibar D.P.
      • Stein J.L.
      • Renteria M.E.
      • Arias-Vasquez A.
      • Desrivières S.
      • Jahanshad N.
      • et al.
      Common genetic variants influence human subcortical brain structures.
      ,
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ,
      • Bis J.C.
      • DeCarli C.
      • Smith A.V.
      • van der Lijn F.
      • Crivello F.
      • Fornage M.
      • et al.
      Common variants at 12q14 and 12q24 are associated with hippocampal volume.
      ). To date, most of these studies have focused on middle-aged and older adults. A recent study found that AD genetic risk variants were associated with smaller hippocampal volume in young adults, suggesting that genetic determinants of late-life neurodegenerative diseases may already be associated with subtle variations in brain structure early in life (
      • Foley S.F.
      • Tansey K.E.
      • Caseras X.
      • Lancaster T.
      • Bracht T.
      • Parker G.
      • et al.
      Multimodal brain imaging reveals structural differences in Alzheimer’s disease polygenic risk carriers: A study in healthy young adults.
      ). However, genetic determinants of MRI-based measures may also change over time, either because of different underlying physiological or pathological processes between young and older persons or because of time-varying exposures to environmental risk factors across the lifespan (
      • Glahn D.C.
      • Kent J.W.
      • Sprooten E.
      • Diego V.P.
      • Winkler A.M.
      • Curran J.E.
      • et al.
      Genetic basis of neurocognitive decline and reduced white-matter integrity in normal human brain aging.
      ,
      • Brouwer R.M.
      • Panizzon M.S.
      • Glahn D.C.
      • Hibar D.P.
      • Hua X.
      • Jahanshad N.
      • et al.
      Genetic influences on individual differences in longitudinal changes in global and subcortical brain volumes: Results of the ENIGMA plasticity working group: Heritability Estimates of Brain Changes.
      ,
      • Chouinard-Decorte F.
      • McKay D.R.
      • Reid A.
      • Khundrakpam B.
      • Zhao L.
      • Karama S.
      • et al.
      Heritable changes in regional cortical thickness with age.
      ).
      Exploring the shared genetic variation underpinning early and late structural brain variations, specifically of subcortical brain volumes, could provide important novel insight into the time course of structural brain changes throughout the adult lifespan and into the mechanisms underlying brain aging and their connection with factors modulating brain maturation. This could contribute to exploring early biological processes driving late-onset neurodegenerative diseases and open new avenues for preventive approaches.
      Using large population-based brain imaging datasets across the adult lifespan, we aimed to 1) estimate the heritability of subcortical volumes from young adulthood to older age, and assess the amount of shared genetic variation across age groups; 2) identify whether genetic loci associated with subcortical volumes in older age also show association with subcortical volumes in early adulthood, and explore underlying genes using transcriptome-wide association studies; and 3) explore the clinical significance of these results in relation with cognition, AD, and PD.

      Methods and Materials

      Study Population

      Analyses were conducted in three age groups: young (18–35 years), middle-aged (35–65 years), and older (65+ years) adults. The cutoff of 35 years of age (also corresponding to the upper age limit of i-Share [Internet-based Students HeAlth Research Enterprise] participants) was based on previous neuroimaging studies indicating that some brain structures, including subcortical structures, reach a peak volume around this age (
      • Dima D.
      • Modabbernia A.
      • Papachristou E.
      • Doucet G.E.
      • Agartz I.
      • Aghajani M.
      • et al.
      Subcortical volumes across the lifespan: Data from 18,605 healthy individuals aged 3-90 years.
      ,
      • Pomponio R.
      • Erus G.
      • Habes M.
      • Doshi J.
      • Srinivasan D.
      • Mamourian E.
      • et al.
      Harmonization of large MRI datasets for the analysis of brain imaging patterns throughout the lifespan.
      ). The cutoff of 65 years of age (also the lower age limit for the 3C-Dijon [Three-City Dijon] participants) corresponds to an age after which aging features on brain imaging tend to accelerate more steeply (
      • Dima D.
      • Modabbernia A.
      • Papachristou E.
      • Doucet G.E.
      • Agartz I.
      • Aghajani M.
      • et al.
      Subcortical volumes across the lifespan: Data from 18,605 healthy individuals aged 3-90 years.
      ,
      • Pomponio R.
      • Erus G.
      • Habes M.
      • Doshi J.
      • Srinivasan D.
      • Mamourian E.
      • et al.
      Harmonization of large MRI datasets for the analysis of brain imaging patterns throughout the lifespan.
      ).

      Heritability Analyses

      Heritability analyses were conducted on individual-level data from three cohorts.

      Heritability Analyses: Population-Based Cohort Studies of Unrelated Individuals

      The i-Share study is an ongoing prospective population-based cohort study of French-speaking students (
      • Montagni I.
      • Guichard E.
      • Kurth T.
      Association of screen time with self-perceived attention problems and hyperactivity levels in French students: A cross-sectional study.
      ). A total of 1777 participants 18 to 35 years of age had both brain MRI and genome-wide genotypes (Supplemental Methods).
      The 3C-Dijon study is a population-based cohort study (
      3C Study Group
      Vascular factors and risk of dementia: Design of the three-city study and baseline characteristics of the study population.
      ). A total of 1440 participants 65 years of age and older had both brain MRI and genome-wide genotypes, excluding participants with dementia, stroke history, or brain tumors at baseline (Supplemental Methods).

      Heritability Analyses: Family-Based Cohort Studies

      The Framingham Heart Study is a community-based cohort study comprising three generations of participants. A total of 1999 participants 36 to 64 years of age and 1828 participants 65 years of age and older had both brain MRI and genome-wide genotypes, excluding participants with stroke history or other neurologic disorders confounding the assessment of brain volumes at the time of MRI (Supplemental Methods).

      Analyses of Shared Genetic Variation Across the Lifespan

      To analyze genetic associations with subcortical volumes in young adults, we used the aforementioned 1777 i-Share participants with high-quality brain MRI and genome-wide genotype data. To derive genome-wide–significant associations with subcortical structures in middle-aged to older adults, we used summary statistics of the largest published meta-analyses of subcortical volumes GWASs (detailed in Supplemental Methods) for optimal power (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ,
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ). In secondary analyses, we also used smaller meta-analyses without a subset of cohorts comprising younger age groups, leading to a sample size of 19,555 participants (detailed in Supplemental Methods). For hippocampal volume, we used a previously published GWAS meta-analysis (
      • Bis J.C.
      • DeCarli C.
      • Smith A.V.
      • van der Lijn F.
      • Crivello F.
      • Fornage M.
      • et al.
      Common variants at 12q14 and 12q24 are associated with hippocampal volume.
      ). Participants with prevalent dementia, stroke, or other neurological pathologies potentially influencing brain measurements at the time of MRI were excluded in all but one meta-analysis (
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ).
      This study was approved by the ethics committees of participating studies, and written informed consent was obtained from all participants.

      MRI Acquisition and Phenotyping

      MRI acquisition parameters and phenotyping methods in individual cohorts are presented in the Supplemental Methods and have been described in detail (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ,
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ,
      • Bis J.C.
      • DeCarli C.
      • Smith A.V.
      • van der Lijn F.
      • Crivello F.
      • Fornage M.
      • et al.
      Common variants at 12q14 and 12q24 are associated with hippocampal volume.
      ,
      • Maillard P.
      • Delcroix N.
      • Crivello F.
      • Dufouil C.
      • Gicquel S.
      • Joliot M.
      • et al.
      An automated procedure for the assessment of white matter hyperintensities by multispectral (T1, T2, PD) MRI and an evaluation of its between-centre reproducibility based on two large community databases.
      ).
      We defined subcortical volumes by the total (left + right) gray matter volumes (cm3) of the accumbens, amygdala, caudate nucleus, hippocampus, pallidum, putamen, and thalamus. The accumbens volume was not available in the 3C-Dijon study.

      Genotyping, Quality Control, and Imputation

      Genome-wide genotyping platforms, quality control, and imputation procedures are described in the Supplemental Methods.

      Statistical Analyses

      Analyses performed in this study are summarized in Figure 1.
      Figure thumbnail gr1
      Figure 1Study workflow and samples summary. Samples of young adults (18–35 years) are presented in blue, of middle-aged (35–65 years) to older (65+ years) adults in green, and of older adults in red, and disease traits are presented in purple. ∗GWAS of subcortical volumes except hippocampal volume after removing ENIGMA and CHARGE cohorts with some young participants (C.L. Satizabal, Ph.D., et al., unpublished data, January 2021). 3C-Dijon, Three-City Dijon; CHARGE, Cohorts for Heart and Aging Research in Genomic Epidemiology; ENIGMA, Enhancing NeuroImaging Genetics through Meta-Analysis; FHS, Framingham Heart Study; GRM, genetic relationship matrix; GRS, genetic risk score; GSMR, generalized summary data–based Mendelian randomization; GWAS, genome-wide association study; i-Share, Internet-based Students HeAlth Research Enterprise; LD, linkage disequilibrium; SNP, single nucleotide polymorphism; TWAS, transcriptome-wide association study.

      Heritability Analyses

      Heritability analyses were adjusted for age, sex, total intracranial volume, and the first four principal components of population stratification.

      Heritability Analyses: Population-Based Cohort Studies of Unrelated Individuals

      To estimate heritability for each subcortical volume in i-Share and 3C-Dijon studies, we used GCTA (v1.26.0) to estimate the proportion of phenotype variance explained by genome-wide single nucleotide polymorphisms (SNPs) (Figure 1; Supplemental Methods) (
      • Yang J.
      • Lee S.H.
      • Goddard M.E.
      • Visscher P.M.
      GCTA: A tool for genome-wide complex trait analysis.
      ).

      Heritability Analyses: Family-Based Cohort Studies

      To estimate heritability for subcortical volumes in the Framingham Heart Study, we used the variance component model implemented in SOLAR accounting for familial relationships to determine the ratio of the genetic variance to the phenotypic variance (
      • Almasy L.
      • Blangero J.
      Multipoint quantitative-trait linkage analysis in general pedigrees.
      ). Mixed models were fit including fixed effects for covariates and additive effects for additive polygenetic and residual error terms.

      Genetic Correlation Analyses

      We used linkage disequilibrium (LD) score regression to estimate the genetic correlation between subcortical volumes in young and older adults (Figure 1) (
      • Bulik-Sullivan B.K.
      • Loh P.-R.
      • Finucane H.K.
      • Ripke S.
      • Yang J.
      • Patterson N.
      • et al.
      LD Score regression distinguishes confounding from polygenicity in genome-wide association studies.
      ,
      • Bulik-Sullivan B.
      • Finucane H.K.
      • Anttila V.
      • Gusev A.
      • Day F.R.
      • Loh P.-R.
      • et al.
      An atlas of genetic correlations across human diseases and traits.
      ). For the older population, we first used the latest, largest published GWAS meta-analyses of subcortical brain volumes (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ,
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ). To confirm that the effects were not driven by the youngest individuals in the GWAS meta-analyses, we conducted secondary analyses using subcortical brain volumes GWAS meta-analyses without cohorts or consortia comprising younger participants (as described for analyses of shared genetic variation across the lifespan). For young adults, we performed a GWAS of subcortical volumes in the i-Share study, adjusting for age, sex, total intracranial volume, and the first four principal components of population stratification. We used a linear mixed model implemented in GCTA-MLMA-LOCO (
      • Yang J.
      • Zaitlen N.A.
      • Goddard M.E.
      • Visscher P.M.
      • Price A.L.
      Advantages and pitfalls in the application of mixed-model association methods.
      ). We used a method implemented in R (matSPDlite) (v3.6.1; R Foundation for Statistical Computing) to correct for multiple testing (
      • Li J.
      • Ji L.
      Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix.
      ). Among the seven subcortical volumes studied, the number of independent tests was estimated at 4.0 based on the correlation matrix between all phenotypes, leading to a Bonferroni-corrected significance threshold of p < 1.25 × 10−2.

      Single-Variant Analyses and Genetic Risk Score Approaches

      First, we selected published genome-wide–significant associations in middle-aged to older adults (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ,
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ) and looked up associations with the corresponding phenotype in young adults (i-Share study) for the lead SNP and nearby variants (±250 kb) in moderate to high LD (LD-r2 > 0.5). To define significance thresholds, we corrected for the four independent phenotypes in the i-Share study and the number of independent loci tested for each subcortical volume (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      Second, we generated genetic risk scores (GRSs) for lower subcortical volumes in young adults (i-Share study) by summing the number of published genome-wide–significant independent risk alleles identified from subcortical volume GWASs in middle-aged to older adults, weighting each risk allele by the regression coefficient for the corresponding SNP in the published GWAS (Supplemental Methods). Associations were tested using linear regression models in R and adjusted for age, sex, total intracranial volume, and the first four principal components of population stratification. To account for multiple testing, we corrected for four independent phenotypes (p < 1.25 × 10−2). As a sensitivity analysis, GRS analyses were repeated using summary statistics of the subcortical volumes GWASs conducted after removing cohorts with young participants as described above (Figure 1).

      Transcriptome-wide Association Study

      To explore genes underlying genetic associations with subcortical volumes across the lifespan, we performed transcriptome-wide association studies (TWASs) using TWAS-Fusion (Figure 1; Supplemental Methods) (
      • Gusev A.
      • Ko A.
      • Shi H.
      • Bhatia G.
      • Chung W.
      • Penninx B.W.J.H.
      • et al.
      Integrative approaches for large-scale transcriptome-wide association studies.
      ). We used summary statistics from the aforementioned published GWAS meta-analyses of subcortical volumes (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ,
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ) and 17 publicly available gene expression quantitative trait loci reference panels from blood, brain, and peripheral nerve tissues (Supplemental Methods). We used TWAS-Fusion to estimate the TWAS association statistics between predicted expression and each subcortical volume (
      • Gusev A.
      • Ko A.
      • Shi H.
      • Bhatia G.
      • Chung W.
      • Penninx B.W.J.H.
      • et al.
      Integrative approaches for large-scale transcriptome-wide association studies.
      ). Transcriptome-wide–significant genes were determined in each tissue expression panel after correcting for the average number of features (3793.5 genes) and 4 independent phenotypes (p < 3.30 × 10−6). Transcriptome-wide–significant genes were then tested in conditional analysis (TWAS-Fusion) (
      • Gusev A.
      • Ko A.
      • Shi H.
      • Bhatia G.
      • Chung W.
      • Penninx B.W.J.H.
      • et al.
      Integrative approaches for large-scale transcriptome-wide association studies.
      ). Next, to ensure that observed associations did not reflect random correlation between gene expression and noncausal variants associated with subcortical volumes, we performed a colocalization analysis (COLOC) on the conditionally significant genes (p < .05) to estimate the posterior probability of a shared causal variant between the gene expression and trait association (PP4) (
      • Giambartolomei C.
      • Vukcevic D.
      • Schadt E.E.
      • Franke L.
      • Hingorani A.D.
      • Wallace C.
      • Plagnol V.
      Bayesian test for colocalisation between pairs of genetic association studies using summary statistics.
      ). Genes presenting a PP4 ≥0.75, for which expression quantitative trait loci did not reach genome-wide significance in association with subcortical volumes, and were not in LD (r2 < 0.01), with a lead SNP of genome-wide–significant risk loci for subcortical volumes, were considered as novel.
      In order to test colocalized associations in a younger population, we used the individual-level prediction of the gene expression option implemented in the Fusion software to generate expression weights in the i-Share study (Supplemental Methods). The significance threshold accounted for the number of genes colocalized in ≥1 tissue for each phenotype (accumbens: p < .05; amygdala: p < .05; caudate: p < 2.17 × 10−3; hippocampus: p < 1.67 × 10−2; pallidum: p < 2.94 × 10−3; putamen: p < 3.85 × 10−3; thalamus: p < .05).

      Lifetime Brain Gene Expression Profile

      We examined the spatiotemporal expression pattern of genes in loci reaching genome-wide significance in the subcortical volumes GWASs, TWAS-COLOC significance, and at least nominal significance in the i-Share TWAS. We used a public database (https://hbatlas.org/) comprising genome-wide exon-level transcriptome data from 1340 tissue samples from 16 brain regions of 57 postmortem human brains, from embryonic development to late adulthood (
      • Kang H.J.
      • Kawasawa Y.I.
      • Cheng F.
      • Zhu Y.
      • Xu X.
      • Li M.
      • et al.
      Spatio-temporal transcriptome of the human brain.
      ).

      Clinical Correlates

      First, we explored the relation of loci associated with subcortical volumes in young and older populations or at least nominally significant in TWAS in the i-Share study (Table 1) with AD, PD, and general cognitive function using summary statistics of the latest, largest published GWASs (
      • Schwartzentruber J.
      • Cooper S.
      • Liu J.Z.
      • Barrio-Hernandez I.
      • Bello E.
      • Kumasaka N.
      • et al.
      Genome-wide meta-analysis, fine-mapping and integrative prioritization implicate new Alzheimer’s disease risk genes.
      ,
      • Nalls M.A.
      • Blauwendraat C.
      • Vallerga C.L.
      • Heilbron K.
      • Bandres-Ciga S.
      • Chang D.
      • et al.
      Identification of novel risk loci, causal insights, and heritable risk for Parkinson’s disease: A meta-analysis of genome-wide association studies.
      ,
      • Davies G.
      • Lam M.
      • Harris S.E.
      • Trampush J.W.
      • Luciano M.
      • Hill W.D.
      • et al.
      Study of 300,486 individuals identifies 148 independent genetic loci influencing general cognitive function.
      ). Accounting for 10 independent loci (r2 < 0.5) and three traits, the significance threshold was p < 1.67 × 10−3.
      Table 1Association of Genome-wide–Significant Variants for Subcortical Volumes in Older Adults With the Same Volumes in Young Adults
      SNP
      For each locus associated at genome-wide–significant level with at least one subcortical structure in Satizabal et al. (15) and Hibar et al. (20), associations of the lead SNP and nearby variants (±250 kb) in moderate to high LD (LD-r2 > 0.5) with the corresponding phenotype were tested in young adults. Only the top SNP of each locus is presented in this table.
      ChromosomePositionA1FrequencyNearest GenesLead SNP From GWASSingle-Variant AnalysesTWAS Analyses
      TWAS analyses based on the summary statistics from Satizabal et al. (15) for subcortical volumes (except hippocampal volume) and from Hibar et al. (20) for hippocampal volume and results from the i-Share TWAS (n = 1586).
      i-Share Cohort (n = 1777)Gene Colocalized in the Locus
      A gene was considered as in the same locus than the top SNP from the GWAS if at least one of its expression quantitative trait loci was in LD (r2 > 0.01) with the top SNP.
      BetaSEp
      Amygdala
       rs1111129312q23.2102,921,296C0.18IGF1, LINC00485rs11111293−0.040.014.50 × 10−3
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      Caudate Nucleus
       rs109099011p36.323,131,235T0.28PRDM16rs28171450.130.038.03 × 10−6
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
       rs1083089411q14.392,018,778T0.41MIR4490, FAT3rs3133370−0.080.025.98 × 10−4
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
       rs195335314q22.356,189,751A0.32KTN1, RPL13AP3rs148470213−0.100.031.92 × 10−4
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
       rs411566816p11.228,607,532A0.27SULT1A2rs1987471−0.100.031.45 × 10−4
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      CCDC101,
      Significant in the i-Share TWAS.
      NPIPB7,
      Significant in the i-Share TWAS.
      NPIPB9,
      Significant in the i-Share TWAS.
      SULT1A1, TUFM,
      Significant in the i-Share TWAS.
      RP11-1348G14.4,
      Significant in the i-Share TWAS.
      RP11-22P6.2
      Significant in the i-Share TWAS.
       rs106279420q11.2130,381,758C0.34TPX2rs6060983−0.080.021.59 × 10−3MYLK2
      Significant in the i-Share TWAS.
      ,
      Significant after multiple testing correction (accumbens: p < .05; amygdala: p < .05; caudate: p < 2.17 × 10−3; hippocampus: p < 1.67 × 10−2; pallidum: p < 2.94 × 10−3; putamen: p < 3.85 × 10−3; thalamus: p < .05).
      Hippocampus
       rs1717800612q14.365,718,299G0.09MSRB3rs61921502−0.150.031.72 × 10−6
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
       rs11320521612q24.22117,326,943A0.09HRK, FBXW8rs779563140.120.031.41 × 10−4
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      Pallidum
       rs94527014q22.356,200,473G0.44KTN1, RPL13AP3rs10129414−0.040.012.29 × 10−4
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
       rs11381854620q11.2130,369,090T0.24TPX2rs10439607−0.030.018.46 × 10−3ENTPD6,
      Significant in the i-Share TWAS.
      FRG1B,
      Significant in the i-Share TWAS.
      ,
      Significant after multiple testing correction (accumbens: p < .05; amygdala: p < .05; caudate: p < 2.17 × 10−3; hippocampus: p < 1.67 × 10−2; pallidum: p < 2.94 × 10−3; putamen: p < 3.85 × 10−3; thalamus: p < .05).
      MLLT10P1,
      Significant in the i-Share TWAS.
      ,
      Significant after multiple testing correction (accumbens: p < .05; amygdala: p < .05; caudate: p < 2.17 × 10−3; hippocampus: p < 1.67 × 10−2; pallidum: p < 2.94 × 10−3; putamen: p < 3.85 × 10−3; thalamus: p < .05).
      PYGB
      Significant in the i-Share TWAS.
      Putamen
       rs74451695q14.387,703,099G0.25TMEM161B-AS1rs2410767−0.080.031.25 × 10−2CTC-498M16.4,
      Significant in the i-Share TWAS.
      TMEM161B,
      Significant in the i-Share TWAS.
      TMEM161B-AS1
      Significant in the i-Share TWAS.
       rs1280026411q23.3117,396,269A0.18DSCAML1rs35200015−0.160.041.30 × 10−5
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
       rs801717214q22.356,199,048A0.44KTN1, RPL13AP3rs945270−0.150.031.43 × 10−7
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
       rs606095420q11.2130,383,187T0.34TPX2rs6087771−0.080.035.05 × 10−3FRG1B,
      Significant in the i-Share TWAS.
      MLLT10P1,
      Significant in the i-Share TWAS.
      MYLK2
      Significant in the i-Share TWAS.
      ,
      Significant after multiple testing correction (accumbens: p < .05; amygdala: p < .05; caudate: p < 2.17 × 10−3; hippocampus: p < 1.67 × 10−2; pallidum: p < 2.94 × 10−3; putamen: p < 3.85 × 10−3; thalamus: p < .05).
      Only loci significant in this analysis or in the TWAS in the i-Share study are presented here.
      GWAS, genome-wide association study; i-Share, Internet-based Students HeAlth Research Enterprise; LD, linkage disequilibrium; SNP, single nucleotide polymorphism; TWAS, transcriptome-wide association study.
      a For each locus associated at genome-wide–significant level with at least one subcortical structure in Satizabal et al. (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ) and Hibar et al. (
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ), associations of the lead SNP and nearby variants (±250 kb) in moderate to high LD (LD-r2 > 0.5) with the corresponding phenotype were tested in young adults. Only the top SNP of each locus is presented in this table.
      b TWAS analyses based on the summary statistics from Satizabal et al. (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ) for subcortical volumes (except hippocampal volume) and from Hibar et al. (
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ) for hippocampal volume and results from the i-Share TWAS (n = 1586).
      c A gene was considered as in the same locus than the top SNP from the GWAS if at least one of its expression quantitative trait loci was in LD (r2 > 0.01) with the top SNP.
      d Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      e Significant in the i-Share TWAS.
      f Significant after multiple testing correction (accumbens: p < .05; amygdala: p < .05; caudate: p < 2.17 × 10−3; hippocampus: p < 1.67 × 10−2; pallidum: p < 2.94 × 10−3; putamen: p < 3.85 × 10−3; thalamus: p < .05).
      Second, we tested whether genetically predicted AD or PD have an impact on subcortical volumes in the general population in young, middle-aged, and older adults using the generalized summary data–based Mendelian randomization tool implemented in GCTA (Supplemental Methods) (
      • Yang J.
      • Lee S.H.
      • Goddard M.E.
      • Visscher P.M.
      GCTA: A tool for genome-wide complex trait analysis.
      ,
      • Zhu Z.
      • Zheng Z.
      • Zhang F.
      • Wu Y.
      • Trzaskowski M.
      • Maier R.
      • et al.
      Causal associations between risk factors and common diseases inferred from GWAS summary data.
      ). The significance threshold accounted for four independent subcortical volumes and two diseases (p < 6.25 × 10−3).

      Data Availability

      Data supporting these results can be made available upon reasonable request from the corresponding author.

      Results

      Heritability and Genetic Correlation of Subcortical Volumes Across the Lifespan

      SNP-heritability estimates for all subcortical volumes in young, middle-aged, and older adults are presented in Figure 2 and Table S1. We observed a decreasing trend of the average SNP-heritability estimates of subcortical volumes in cohorts of increasing age (Figure 2A), and a similar decreasing trend for most individual subcortical volumes (Figure 2B), both in unrelated and in family-based population-based studies.
      Figure thumbnail gr2
      Figure 2Heritability of subcortical volumes in young, middle-aged, and older adults. (A) The box plots represent the distribution of the heritability of subcortical volumes. The main rectangle represents the interquartile range and the horizontal line is the median. (B) Plot of the estimated SNP-heritability for each subcortical volume, in each age group. For the unrelated population, “young adults” refers to the i-Share cohort (18–35 years, n = 1528) and “older adults” refers to the 3C-Dijon cohort (65+ years, n = 1396). For the family-based population, all the groups came from the Framingham Heart Study (three generations). “Middle” or “middle-aged adults” refers to 36–64 years of age (n = 1999) and “old” or “older adults” for 65+ years of age (n = 1828). 3C-Dijon, Three-City Dijon; i-Share, Internet-based Students HeAlth Research Enterprise; SNP, single nucleotide polymorphism.
      Most subcortical volumes were genetically correlated with each other among middle-aged to older adults, while in young adults genetic correlations reached significance after correction for multiple testing between the hippocampus and amygdala, pallidum and caudate nucleus, and putamen and pallidum. Genetic correlations between young and middle-aged to older adults for the same individual subcortical volumes were significant for the caudate nucleus and the pallidum (Figure 3). These results were similar when using the published GWAS meta-analyses of subcortical volumes in middle-aged to older adults (Figure 3A), and the secondary GWAS meta-analyses in older adults (Figure 3B).
      Figure thumbnail gr3
      Figure 3Heatmap of the genetic correlation between subcortical volumes in young adults (“young”), in the older population—(A) “old” and (B) “old only”—and between the two populations for equivalent structures. ∗Nominally significant. ∗∗Significant after multiple testing correction (p < 1.25 × 10−2); larger colored squares correspond to more significant p values. In each square, the numbers correspond to the coefficient of the genetic correction (rg). “Young” refers to the i-Share cohort; “middle to old” refers to subcortical volumes from the latest largest GWASs (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ,
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ); “old” refers to the GWAS of subcortical volumes except hippocampal volume after removing ENIGMA and CHARGE cohorts with some young participants and for the CHARGE GWAS of hippocampal volume (
      • Bis J.C.
      • DeCarli C.
      • Smith A.V.
      • van der Lijn F.
      • Crivello F.
      • Fornage M.
      • et al.
      Common variants at 12q14 and 12q24 are associated with hippocampal volume.
      ). CHARGE, Cohorts for Heart and Aging Research in Genomic Epidemiology; ENIGMA, Enhancing NeuroImaging Genetics through Meta-Analysis; GWAS, genome-wide association study; i-Share, Internet-based Students HeAlth Research Enterprise.

      Genetic Associations With Subcortical Volumes Across the Lifespan

      Out of the 38 genome-wide–significant loci for subcortical volumes in middle-aged to older adults, 10 were significantly associated with the same volume, in the same direction, in young adults (Table 1; Table S2): four in the caudate nucleus; two in the hippocampus and the putamen, respectively; and one in the amygdala and the pallidum. The most significant association in young adults was for rs8017172 (p = 1.43 × 10−7), near KTN1, with putamen volume.
      GRSs for smaller subcortical volumes derived from GWASs in middle-aged to older adults were significantly associated with smaller volumes of the same structure in young adults for the putamen, caudate nucleus, hippocampus, amygdala, and pallidum (Table 2). The most significant association was observed for the putamen (p = 5.04 × 10−11). When deriving the GRS from the secondary meta-analyses of GWASs in older adults exclusively, associations remained significant for the putamen, caudate nucleus, and hippocampus.
      Table 2Association of GRS (From Genome-wide–Significant Variants) of Subcortical Volumes in Older Adults With the Same Volumes in Young Adults
      GRSi-Share Cohort (n = 1586)
      BetaSEp
      Accumbens
       GRS from full GWAS
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) and from Hibar et al. (20) for hippocampal volume.
      −0.040.043.56 × 10−1
       GRS from GWAS old only
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) after excluding cohorts containing young participants and from Bis et al. (21) for hippocampal volume.
      −0.050.084.69 × 10−1
      Amygdala
       GRS from full GWAS
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) and from Hibar et al. (20) for hippocampal volume.
      −0.610.149.47 × 10−6
      Significant results after multiple testing correction (p < 1.25 × 10−2).
       GRS from GWAS old only
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) after excluding cohorts containing young participants and from Bis et al. (21) for hippocampal volume.
      −0.230.151.29 × 10−1
      Caudate Nucleus
       GRS from full GWAS
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) and from Hibar et al. (20) for hippocampal volume.
      −0.730.149.76 × 10−8
      Significant results after multiple testing correction (p < 1.25 × 10−2).
       GRS from GWAS old only
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) after excluding cohorts containing young participants and from Bis et al. (21) for hippocampal volume.
      −0.990.212.31 × 10−6
      Significant results after multiple testing correction (p < 1.25 × 10−2).
      Hippocampus
       GRS from full GWAS
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) and from Hibar et al. (20) for hippocampal volume.
      −0.460.092.08 × 10−7
      Significant results after multiple testing correction (p < 1.25 × 10−2).
       GRS from GWAS old only
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) after excluding cohorts containing young participants and from Bis et al. (21) for hippocampal volume.
      −1.120.217.74 × 10−8
      Significant results after multiple testing correction (p < 1.25 × 10−2).
      Pallidum
       GRS from full GWAS
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) and from Hibar et al. (20) for hippocampal volume.
      −0.280.076.15 × 10−5
      Significant results after multiple testing correction (p < 1.25 × 10−2).
       GRS from GWAS old only
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) after excluding cohorts containing young participants and from Bis et al. (21) for hippocampal volume.
      −0.160.101.20 × 10−1
      Putamen
       GRS from full GWAS
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) and from Hibar et al. (20) for hippocampal volume.
      −0.670.105.04 × 10−11
      Significant results after multiple testing correction (p < 1.25 × 10−2).
       GRS from GWAS old only
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) after excluding cohorts containing young participants and from Bis et al. (21) for hippocampal volume.
      −1.050.215.92 × 10−7
      Significant results after multiple testing correction (p < 1.25 × 10−2).
      Thalamus
       GRS from full GWAS
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) and from Hibar et al. (20) for hippocampal volume.
      −0.210.315.01 × 10−1
       GRS from GWAS old only
      GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (15) after excluding cohorts containing young participants and from Bis et al. (21) for hippocampal volume.
      0.070.548.98 × 10−1
      GRS, genetic risk score; GWAS, genome-wide association study; i-Share, Internet-based Students HeAlth Research Enterprise; SNP, single nucleotide polymorphism.
      a GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ) and from Hibar et al. (
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ) for hippocampal volume.
      b GRSs generated using the SNPs with p < 5 × 10−8 from the summary statistics of the GWAS of subcortical volumes from Satizabal et al. (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ) after excluding cohorts containing young participants and from Bis et al. (
      • Bis J.C.
      • DeCarli C.
      • Smith A.V.
      • van der Lijn F.
      • Crivello F.
      • Fornage M.
      • et al.
      Common variants at 12q14 and 12q24 are associated with hippocampal volume.
      ) for hippocampal volume.
      c Significant results after multiple testing correction (p < 1.25 × 10−2).
      We then sought to explore putative causal genes underlying genome-wide–significant associations with subcortical volumes using TWASs, initially based on published GWAS meta-analyses in middle-aged and older adults (Figure 1). Among all genes showing transcriptome-wide–significant associations, 54 presented a high colocalization posterior probability of sharing a causal variant between the gene expression and trait association (COLOC PP4 ≥ 0.75) with at least one subcortical volume, mostly in brain tissues (Figure S1; Table S3): 23 with the caudate nucleus, 17 with the pallidum, 13 with the putamen, 3 with the hippocampus, and 1 respectively with the nucleus accumbens, amygdala, and thalamus. Among these 54 genes, 30 were in loci that did not reach genome-wide significance in the subcortical volumes GWASs and can be considered as novel (Table S3). Although we lacked power to detect transcriptome-wide associations in the i-Share study, we detected significant signals after multiple testing correction for 4 of the 54 colocalized genes described above, all at the genome-wide–significant subcortical volume GWAS locus chr20q11.21 (Figure 4; Figure S1, Table S3): lower expression of MYLK2 in the putamen was associated with smaller caudate nucleus and putamen volume and higher expression of FRG1B, MLLT10P1, and RP4-610C12.4 in the basal ganglia and cerebellum was associated with smaller pallidum volume. With an exploratory purpose, we also considered 13 additional genes showing nominally significant associations in the young adult TWASs (Figure 4), including seven genes/transcripts for the caudate nucleus (at chr16p11.2-12.1: CCDC101, NPIPB7, NPIPB9, TUFM, EIF3C, RP11-1348G14.4, RP11-22P6.2), two for the pallidum (at chr20p11.21: ENTPD6, PYGB), and four for the putamen (at chr5q14.3: CTC-498M16.4, TMEM161B, TMEM161B-AS1; and at chr20q11.21: MLLT10P1). Of note, the chr16p11.2-12.1 locus also showed significant association with the caudate nucleus in young adults (Table 1, Table S2). Most TWAS-COLOC genes from genome-wide–significant risk loci for subcortical volumes that also showed nominally significant TWAS association in young adults were described to have constant expression levels in subcortical regions throughout the life course, including in the prenatal period (Figure S2).
      Figure thumbnail gr4
      Figure 4Heatmaps of the TWASs of the caudate nucleus, putamen, and pallidum reaching transcriptome-wide significance and colocalized in older persons (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ,
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ) and reaching at least nominal significance in young adults (i-Share cohort). ∗TWAS significant (p < 3.30 × 10−6). ∗∗Conditionally significant (p < .05); ∗∗∗COLOC PP4 > 0.75. +Nominally significant in i-Share study; ++Significant in i-Share study (after multiple testing correction [accumbens: p < .05; amygdala: p < .05; caudate: p < 2.17 × 10−3; hippocampus: p < 1.67 × 10−2; pallidum: p < 2.94 × 10−3; putamen: p < 3.85 × 10−3; thalamus: p < .05]). BA, Brodmann area; CMC, CommonMind Consortium; COLOC, colocalization analysis; DLPFC, dorsolateral prefrontal cortex; GTEX, Genotype-Tissue Expression; i-Share, Internet-based Students HeAlth Research Enterprise; NTR, Netherlands Twin Registry; RNA-seq, RNA sequencing; TWAS, transcriptome-wide association study; YFS, Young Finns Study.

      Clinical Correlates

      When exploring the association of the 14 genetic variants (in 10 independent loci) associated with subcortical volumes in both young and older adults (Table 1) with AD, PD, and general cognitive function, we observed a genome-wide–significant association of the lead SNP for smaller caudate nucleus volume at chr16p11.2-12.1 with lower general cognitive function (Table 3). No association reached significance with AD and PD after multiple testing correction.
      Table 3Association of Genetic Variants Associated With Subcortical Volumes in Older and Young Adults With Alzheimer’s and Parkinson’s Diseases and General Cognitive Function
      SNPLocusA1Subcortical Volumes: Young Adults (n = 1777)Alzheimer’s Disease (n = 75,397/844,024)Parkinson’s Disease (n = 33,674/449,056)General Cognitive Function (n = 282,014)
      zpzpzpzp
      Amygdala
       rs1111129312q23.2C−2.844.50 × 10−3
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      −0.794.28 × 10−10.228.22 × 10−1−2.677.63 × 10−3
      Caudate Nucleus
       rs109099011p36.32C−4.468.03 × 10−6
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      −0.079.43 × 10−1−0.645.20 × 10−1−1.491.38 × 10−1
       rs1083089411q14.3T−3.435.98 × 10−4
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      0.208.43 × 10−1−2.123.40 × 10−2−0.377.10 × 10−1
       rs195335314q22.3A−3.731.92 × 10−4
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      1.461.44 × 10−1−2.609.23 × 10−3−1.601.10 × 10−1
       rs411566816p11.2A−3.801.45 × 10−4
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      −1.846.64 × 10−2−2.014.52 × 10−2−7.302.88 × 10−13
      Significant after correction for multiple testing (p < 1.67 × 10−3).
       rs106279420q11.21C−3.161.59 × 10−3−1.551.22 × 10−1−1.866.19 × 10−2−0.079.48 × 10−1
      Hippocampus
       rs1717800612q14.3G−4.781.72 × 10−6
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      −2.923.49 × 10−3−1.846.53 × 10−2−0.724.75 × 10−1
       rs11320521612q24.22C−3.811.41 × 10−4
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      −2.103.59 × 10−2−0.317.60 × 10−10.734.68 × 10−1
      Pallidum
       rs94527014q22.3G−3.682.29 × 10−4
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      1.571.17 × 10−1−2.451.47 × 10−2−2.044.16 × 10−2
       rs11381854620q11.21T−2.638.46 × 10−3−1.023.09 × 10−1−1.601.09 × 10−10.109.18 × 10−1
      Putamen
       rs74451695q14.3G−2.501.25 × 10−20.814.19 × 10−1−0.129.08 × 10−13.632.88 × 10−4
      Significant after correction for multiple testing (p < 1.67 × 10−3).
       rs1280026411q23.3A−4.361.30 × 10−5
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      −0.754.54 × 10−1−0.208.39 × 10−10.953.44 × 10−1
       rs801717214q22.3A−5.261.43 × 10−7
      Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      1.441.49 × 10−1−2.541.10 × 10−2−1.984.77 × 10−2
       rs606095420q11.21T−2.805.05 × 10−3−1.531.26 × 10−1−1.846.53 × 10−20.009.98 × 10−1
      Only loci significant in single-variant analysis or in transcriptome-wide association study in the i-Share study are presented here.
      i-Share, Internet-based Students HeAlth Research Enterprise; SNP, single nucleotide polymorphism.
      a Significant after multiple testing correction (accumbens: p < 3.13 × 10−3; amygdala: p < 1.25 × 10−2; caudate: p < 1.25 × 10−3; hippocampus: p < 2.08 × 10−3; pallidum: p < 2.08 × 10−3; putamen: p < 1.39 × 10−3; thalamus: p < 6.25 × 10−3).
      b Significant after correction for multiple testing (p < 1.67 × 10−3).
      Using generalized summary data–based Mendelian randomization, genetically predicted AD was significantly associated with smaller volumes of most subcortical structures except the pallidum in dementia-free middle-aged and older adults from the general population, and with larger putamen volume in young adults (Figure 5; Table S4). Genetically predicted PD was significantly associated with larger thalamus volume in young adults (Figure 5; Table S4).
      Figure thumbnail gr5
      Figure 5Association between genetically predicted Alzheimer’s and Parkinson’s diseases and subcortical volumes in young, middle-aged to older, and older adults using Mendelian randomization. “Young adults” refers to the i-Share cohort, “middle-aged to older adults” refers to subcortical volumes from the latest largest GWASs (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ,
      • Hibar D.P.
      • Adams H.H.H.
      • Jahanshad N.
      • Chauhan G.
      • Stein J.L.
      • Hofer E.
      • et al.
      Novel genetic loci associated with hippocampal volume.
      ), and “older adults” refers to the GWAS of subcortical volumes except hippocampal volume after removing the ENIGMA and CHARGE cohorts with some young participants and for the CHARGE GWAS of hippocampal volume (
      • Bis J.C.
      • DeCarli C.
      • Smith A.V.
      • van der Lijn F.
      • Crivello F.
      • Fornage M.
      • et al.
      Common variants at 12q14 and 12q24 are associated with hippocampal volume.
      ). CHARGE, Cohorts for Heart and Aging Research in Genomic Epidemiology; ENIGMA, Enhancing NeuroImaging Genetics through Meta-Analysis; GWAS, genome-wide association study; NS, not significant.

      Discussion

      In large population-based cohort studies across the adult lifespan, we identified a consistent trend toward decreasing heritability of subcortical volumes with increasing age. We observed significant genetic correlation of caudate nucleus and pallidum volumes between young and older adults. GRSs for smaller caudate nucleus, putamen, or hippocampus volumes in older adults were significantly associated with smaller volumes of the same structures in young adults. Individually, 10 of 38 independent loci associated with subcortical volumes in older adults also showed significant associations with the corresponding volumes in young adults. Using TWASs with colocalization analyses, we found evidence for expression levels of 16 genes to be significantly associated with caudate nucleus, putamen, or pallidum volume both in older and young adults, pointing to biological pathways underlying structural brain changes across the adult lifespan. One genome-wide–significant caudate nucleus locus in older and young adults (at TUFM) was associated with lower general cognitive function at genome-wide significance. We also observed an association of genetically determined AD with smaller volumes of most subcortical structures in middle-aged and older dementia-free adults.
      The observed heritability trends of subcortical volumes across the adult lifespan, both in an unrelated and in a family-based population-based setting, support and expand on a previous meta-analysis of brain volume heritability estimates in twin and family studies. The latter showed increasing heritability from childhood to early adulthood and decreasing heritability from young adulthood to old age (
      • Batouli S.A.H.
      • Trollor J.N.
      • Wen W.
      • Sachdev P.S.
      The heritability of volumes of brain structures and its relationship to age: A review of twin and family studies.
      ). Potential explanations include an increase of the environmental contribution to variation in subcortical volumes with increasing age, thus leading to a relative decrease of the genetic contribution, differential timing of gene expression during life, or age of onset of some disorders (
      • Giedd J.N.
      • Schmitt J.E.
      • Neale M.C.
      Structural brain magnetic resonance imaging of pediatric twins.
      ).
      Genetic variants associated with smaller subcortical volumes in older persons were associated with smaller volumes of the same structure already in early adulthood, both individually and aggregated in a GRS. These results suggest that biological pathways influencing subcortical volumes in older age already have an impact on the latter in young adulthood. Although our analyses only focused on adult age, one could speculate that at least some of the susceptibility variants for smaller subcortical structures already showing significant associations at 20 years of age may be involved in developmental processes. This is supported by a recent study showing that polygenic risk scores for subcortical volumes in middle-aged to older adults were already associated with these volumes during infancy and early childhood (
      • Lamballais S.
      • Jansen P.R.
      • Labrecque J.A.
      • Ikram M.A.
      • White T.
      Genetic scores for adult subcortical volumes associate with subcortical volumes during infancy and childhood.
      ). Several genes in loci associated with subcortical volumes in older persons were reported to be involved in neurodevelopmental processes in experimental work and are implicated in Mendelian disorders (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ). FAT3 at the chr11q14.3 locus that also showed association with caudate nucleus volume in the i-Share study was for instance shown to be involved in neuronal morphogenesis and cell migration (
      • Satizabal C.L.
      • Adams H.H.H.
      • Hibar D.P.
      • White C.C.
      • Knol M.J.
      • Stein J.L.
      • et al.
      Genetic architecture of subcortical brain structures in 38,851 individuals.
      ).
      Ten individual loci associated with smaller subcortical volumes in older adults were already associated with smaller volumes of the same structures in young adults. We also showed association of subcortical volumes with up- or downregulation of three genes in loci associated with caudate nucleus, putamen, and pallidum volumes both in older and young adults: MYLK2, FRG1B, and MLLT10P1. Among these, upregulation of MYLK2 expression in brain tissues was significantly associated with larger caudate nucleus and putamen volume, in both older and young adults (Figure 4; Figure S3). In brain single-cell RNA-sequencing analyses in mice, mylk2 appears to be expressed in arterial and arteriolar smooth muscle cells and in pericytes (
      • Vanlandewijck M.
      • He L.
      • Mäe M.A.
      • Andrae J.
      • Ando K.
      • Del Gaudio F.
      • et al.
      A molecular atlas of cell types and zonation in the brain vasculature.
      ,
      • He L.
      • Vanlandewijck M.
      • Mäe M.A.
      • Andrae J.
      • Ando K.
      • Del Gaudio F.
      • et al.
      Single-cell RNA sequencing of mouse brain and lung vascular and vessel-associated cell types.
      ). MYLK2 encodes a myosin light-chain kinase, a calcium/calmodulin-dependent enzyme, harboring rare mutations causing hypertrophic cardiomyopathy (
      • Davis J.S.
      • Hassanzadeh S.
      • Winitsky S.
      • Lin H.
      • Satorius C.
      • Vemuri R.
      • et al.
      The overall pattern of cardiac contraction depends on a spatial gradient of myosin regulatory light chain phosphorylation.
      ).
      Expression levels of four genes (CCDC101, NPIPB7, NPIPB9, and TUFM) in a caudate nucleus GWAS locus that also shows significant association in young adults (chr16p11.2-12.1) (Table 1) were associated with caudate nucleus volume at transcriptome-wide significance in older adults and nominal significance in young adults. Among these, increased TUFM expression was associated with smaller caudate nucleus volume, with evidence for colocalization in several tissues (Figure 4; Figure S3). TUFM is involved in combined oxidative phosphorylation deficiency 4, a disease causing developmental regression, microcephaly, and basal ganglia atrophy (
      • Kohda M.
      • Tokuzawa Y.
      • Kishita Y.
      • Nyuzuki H.
      • Moriyama Y.
      • Mizuno Y.
      • et al.
      A comprehensive genomic analysis reveals the genetic landscape of mitochondrial respiratory chain complex deficiencies.
      ,
      • Valente L.
      • Tiranti V.
      • Marsano R.M.
      • Malfatti E.
      • Fernandez-Vizarra E.
      • Donnini C.
      • et al.
      Infantile encephalopathy and defective mitochondrial DNA translation in patients with mutations of mitochondrial elongation factors EFG1 and EFTu.
      ). Interestingly, the lead variant associated with smaller caudate nucleus at chr16p11.2-12.1 was associated with lower general cognitive function at genome-wide significance. Additional studies are required to confirm these findings and explore the role of TUFM.
      The most significant association with subcortical volumes in young adults was observed for an intergenic variant at chr14q22.3, with KTN1 as the closest gene, reaching near to genome-wide significance in association with putamen volume despite a limited sample size. Genetic variants in this region had previously been shown to be associated with putamen shape in healthy adolescents (
      • Hibar D.P.
      • Stein J.L.
      • Renteria M.E.
      • Arias-Vasquez A.
      • Desrivières S.
      • Jahanshad N.
      • et al.
      Common genetic variants influence human subcortical brain structures.
      ), and our results thus strengthen the evidence for an important role of this locus in modulating putamen structure across the lifespan.
      It is noteworthy that genetic correlations between young and older adults were significant for striatal and pallidal volumes only, and most significant associations of GRSs and individual variants across age groups, as well as transcriptome-wide associations, were observed for these structures, and to a lesser extent the hippocampus. This could be at least partly explained by the recent observation that striatal and pallidal volumes peak in childhood and decline steadily thereafter, while volumes of the thalamus, amygdala, and hippocampus peak later and start declining from the sixth decade onward (
      • Dima D.
      • Modabbernia A.
      • Papachristou E.
      • Doucet G.E.
      • Agartz I.
      • Aghajani M.
      • et al.
      Subcortical volumes across the lifespan: Data from 18,605 healthy individuals aged 3-90 years.
      ). Moreover, interindividual variability of thalamus, amygdala, and hippocampus were found to increase with age, suggesting that these structures may be more susceptible to environmental factors or late-acting genes (
      • Dima D.
      • Modabbernia A.
      • Papachristou E.
      • Doucet G.E.
      • Agartz I.
      • Aghajani M.
      • et al.
      Subcortical volumes across the lifespan: Data from 18,605 healthy individuals aged 3-90 years.
      ).
      Using Mendelian randomization (generalized summary data–based Mendelian randomization), we identified significant associations of genetically determined AD, but not PD, with all subcortical volumes except the pallidum in older dementia-free adults, consistent with prior observations from studies using polygenic risk scores of AD on the hippocampus and amygdala (
      • Chauhan G.
      • Adams H.H.H.
      • Bis J.C.
      • Weinstein G.
      • Yu L.
      • Töglhofer A.M.
      • et al.
      Association of Alzheimer disease GWAS loci with MRI-markers of brain aging.
      ,
      • Lupton M.K.
      • Strike L.
      • Hansell N.K.
      • Wen W.
      • Mather K.A.
      • Armstrong N.J.
      • et al.
      The effect of increased genetic risk for Alzheimer’s disease on hippocampal and amygdala volume.
      ,
      • Stocker H.
      • Möllers T.
      • Perna L.
      • Brenner H.
      The genetic risk of Alzheimer’s disease beyond APOE ε4: Systematic review of Alzheimer’s genetic risk scores.
      ,
      • Mormino E.C.
      • Sperling R.A.
      • Holmes A.J.
      • Buckner R.L.
      • De Jager P.L.
      • Smoller J.W.
      • Sabuncu M.R.
      Polygenic risk of Alzheimer disease is associated with early- and late-life processes.
      ,
      • Harrison J.R.
      • Mistry S.
      • Muskett N.
      • Escott-Price V.
      From polygenic scores to precision medicine in Alzheimer’s disease: A systematic review.
      ). These associations were not observed in young adults, in contrast with other studies, which identified an association between polygenic risk scores of AD and hippocampal volume in adolescent and young adults (
      • Foley S.F.
      • Tansey K.E.
      • Caseras X.
      • Lancaster T.
      • Bracht T.
      • Parker G.
      • et al.
      Multimodal brain imaging reveals structural differences in Alzheimer’s disease polygenic risk carriers: A study in healthy young adults.
      ,
      • Axelrud L.K.
      • Santoro M.L.
      • Pine D.S.
      • Talarico F.
      • Gadelha A.
      • Manfro G.G.
      • et al.
      Polygenic risk score for Alzheimer’s disease: Implications for memory performance and hippocampal volumes in early life.
      ). These had used polygenic risk scores composed of SNPs selected at less stringent p-value thresholds (
      • Foley S.F.
      • Tansey K.E.
      • Caseras X.
      • Lancaster T.
      • Bracht T.
      • Parker G.
      • et al.
      Multimodal brain imaging reveals structural differences in Alzheimer’s disease polygenic risk carriers: A study in healthy young adults.
      ,
      • Axelrud L.K.
      • Santoro M.L.
      • Pine D.S.
      • Talarico F.
      • Gadelha A.
      • Manfro G.G.
      • et al.
      Polygenic risk score for Alzheimer’s disease: Implications for memory performance and hippocampal volumes in early life.
      ). In young adults, we found that genetically determined AD was associated with larger putamen volume and genetically determined PD with larger thalamus volume. If confirmed, these results could reflect complex mechanisms whereby biological pathways contributing to larger maximal volumes in certain brain regions early in life could also enhance late-life neurodegenerative processes, akin to the observation that rates of age-related maturation are significantly correlated with rates of decline of white matter tracts (
      • Kochunov P.
      • Williamson D.E.
      • Lancaster J.
      • Fox P.
      • Cornell J.
      • Blangero J.
      • Glahn D.C.
      Fractional anisotropy of water diffusion in cerebral white matter across the lifespan.
      ).
      To our knowledge, this is one of the first studies exploring associations of genetic variants with subcortical volumes across the adult lifespan. Our analyses were based on high-quality MRI measurements and genome-wide genotype data in several cohorts, including a unique cohort of young students and leverage large scale meta-analyses conducted within the CHARGE (Cohorts for Heart and Aging Research in Genomic Epidemiology) Consortium.
      We acknowledge limitations. While we describe trends in heritability estimates, we could not formally compare whether these differed significantly between age groups, therefore decreasing heritability estimates with increasing age need to be interpreted with caution. The sample size of our young adult cohort was limited, particularly compared with the large-scale meta-analysis of GWASs in older adults, and included a majority of women. Our study was restricted to the adult lifespan and should be complemented by cohorts of children and adolescents to capture the full spectrum of genetic determinants across the entire life course. Whereas a longitudinal design would be most appropriate to explore changes in heritability estimates and genetic determinants across the lifespan in the same individuals, repeated MRIs and genetic analyses in large population-based samples are limited and have been available for the past 20 years only. While we show compelling association of genetic risk variants for AD with smaller subcortical volumes in middle and older dementia-free community persons, we did not observe such an association in young adults. Moreover, while one of the loci associated with lower caudate nucleus volume showed a genome-wide–significant association with lower general cognitive function, most of the loci associated with subcortical volumes across the adult lifespan were not associated with neurodegenerative diseases after multiple testing correction. This may suggest that these loci are not necessarily reflecting early neurodegenerative processes, but rather may point to developmental or nonpathological processes related to healthy aging. The fact that genetic loci associated with subcortical volumes in middle-aged to older adults (aggregated in polygenic risk scores) were recently found to be associated with these volumes already during infancy and early childhood could be a potential argument for a stronger role of developmental processes (
      • Lamballais S.
      • Jansen P.R.
      • Labrecque J.A.
      • Ikram M.A.
      • White T.
      Genetic scores for adult subcortical volumes associate with subcortical volumes during infancy and childhood.
      ). We may have been underpowered to measure the modifying effects (deleterious or protective) of loci modulating subcortical volumes on the occurrence of neurodegenerative diseases. Finally, we cannot exclude selective survival bias for some variants when exploring their relationship with late-life neurodegenerative diseases (if variants associated with subcortical volumes also influence survival).
      In conclusion, our findings provide novel insight into the genetic determinants of subcortical volumes across the adult lifespan, with some evidence suggesting that specific genes such as MYLK2 and TUFM may have a causal role in determining subcortical volumes already in young adulthood. Further research is warranted to decipher the underlying biological mechanisms and inform prevention strategies for common late-life neurodegenerative diseases, for which pathological processes are known to start long before their clinical diagnosis.

      Acknowledgments and Disclosures

      The i-Share study is conducted by the Universities of Bordeaux and Versailles Saint-Quentin-en-Yvelines (France). The i-Share study has received funding by the French National Agency ( Agence Nationale de la Recherche [ANR]), via the “Investissements d’Avenir” programme (Grant No. ANR-10-COHO-05) and from the University of Bordeaux Initiative of Exellence (IdEX). This project has also received funding from the European Research Council under the European Union (EU) Horizon 2020 research and innovation programme under Grant No. 640643 . The 3C-Dijon study is conducted under a partnership agreement among the Institut National de la Santé et de la Recherche Médicale, University of Bordeaux, and Sanofi-Aventis. The Fondation pour la Recherche Médicale funded the preparation and initiation of the study. The 3C-Dijon study is also supported by the Caisse Nationale Maladie des Travailleurs Salariés , Direction Générale de la Santé , Mutuelle Générale de l'Education Nationale , Institut de la Longévité , Conseils Régionaux of Aquitaine and Bourgogne, Fondation de France , and Ministry of Research–Institut National de la Santé et de la Recherche Médicale Programme “Cohortes et collections de données biologiques.” CT and SD have received investigator-initiated research funding from the ANR and from the Fondation Leducq . This work was supported by the National Foundation for Alzheimer’s Disease and Related Disorders , the Institut Pasteur de Lille , the laboratory of excellence DISTALZ, and the Centre National de Génotypage. This project is an EU Joint Programme–Neurodegenerative Disease Research project. The project is supported through the following funding organizations under the aegis of the EU Joint Programme – Neurodegenerative Disease Research (www.jpnd.eu): Australia, National Health and Medical Research Council ; Austria, Federal Ministry of Science , Research and Economy; Canada, Canadian Institutes of Health Research ; France, French National Research Agency ; Germany, Federal Ministry of Education and Research; the Netherlands, the Netherlands Organisation for Health Research and Development; United Kingdom, Medical Research Council . This project has received funding from the EU Horizon 2020 research and innovation programme under Grant No. 643417 . This project has also received funding from the EU Horizon 2020 research and innovation programme under Grant Nos. 667375 and 754517 . This work was also supported by a grant overseen by the ANR as part of the “Investment for the Future Programme” ANR-18-RHUS-0002 . The Framingham Heart Study was supported by the National Heart, Lung, and Blood Institute (Grant Nos. N01-HC-25195 , HHSN268201500001I , and 75N92019D00031 ). This study was also supported by grants from the National Institute on Aging (Grant Nos. R01AG031287 , R01AG054076 , R01AG049607 , R01AG059421 , R01AG059725 , U01AG049505 , U01AG052409 ) and the National Institute of Neurological Disorders and Stroke (Grant No. R01 NS017950 ). The generation of FreeSurfer data was supported by the Alzheimer's Association (Grant No. AARGD-16-443384 ). Funding for SHARe Affymetrix genotyping was provided by National Heart, Lung, and Blood Institute Grant No. N02-HL64278 . Infrastructure for the CHARGE Consortium is supported in part by the National Heart, Lung, and Blood Institute Grant No. R01HL105756 , and NeuroCHARGE is supported in part by the National Institute on Aging Grant No. R01AG033193 . AL and AT benefited from a grant from the Fondation pour la Recherche Médicale (Grant No. DIC202161236446 ). QLG benefited from the Digital Public Health Graduate Program, a PhD program supported by the French Investment for the Future Programme (Grant No. 17-EURE-0019 ). CLS receives support from National Institute of Health (Grant No. R01AG059727). CD received funding from National Institute of Aging (Grant No. P30 AG0101029).
      We thank Dr. Mathieu Bourgey, Dr. Hans Markus Munter, Alexandre Belisle, and Rui Li at the McGill Genome Center for processing the i-Share study genome-wide genotype data. We thank Professor Philippe Amouyel for supporting the funding of the 3C-Dijon study genome-wide genotyping. Part of the computations were performed at the Bordeaux Bioinformatics Center, University of Bordeaux, and the Centre de Ressource et Développement en Informatique Médicale at the University of Bordeaux, on a server infrastructure supported by the Fondation Claude Pompidou. The computational work reported in this article was performed on the Shared Computing Cluster, which is administered by Boston University’s Research Computing Services. We also thank all the Framingham Heart Study participants.
      The authors report no biomedical financial interests or potential conflicts of interest.

      Supplementary Material

      References

        • Yin H.H.
        • Knowlton B.J.
        The role of the basal ganglia in habit formation.
        Nat Rev Neurosci. 2006; 7: 464-476
        • McDonald A.J.
        • Mott D.D.
        Functional neuroanatomy of amygdalohippocampal interconnections and their role in learning and memory: Amygdalohippocampal Interconnections.
        J Neurosci Res. 2017; 95: 797-820
        • Hikosaka O.
        • Kim H.F.
        • Yasuda M.
        • Yamamoto S.
        Basal ganglia circuits for reward value–guided behavior.
        Annu Rev Neurosci. 2014; 37: 289-306
        • Salzman C.D.
        • Fusi S.
        Emotion, cognition, and mental state representation in amygdala and prefrontal cortex.
        Annu Rev Neurosci. 2010; 33: 173-202
        • Floresco S.B.
        The nucleus accumbens: An interface between cognition, emotion, and action.
        Annu Rev Psychol. 2015; 66: 25-52
        • Dima D.
        • Modabbernia A.
        • Papachristou E.
        • Doucet G.E.
        • Agartz I.
        • Aghajani M.
        • et al.
        Subcortical volumes across the lifespan: Data from 18,605 healthy individuals aged 3-90 years.
        Hum Brain Mapp. 2022; 43: 452-469
        • Coupé P.
        • Manjón J.V.
        • Lanuza E.
        • Catheline G.
        Lifespan changes of the human brain in Alzheimer’s disease.
        Sci Rep. 2019; 9: 3998
        • de Jong L.W.
        • van der Hiele K.
        • Veer I.M.
        • Houwing J.J.
        • Westendorp R.G.J.
        • Bollen E.L.E.M.
        • et al.
        Strongly reduced volumes of putamen and thalamus in Alzheimer’s disease: An MRI study.
        Brain. 2008; 131: 3277-3285
        • Mak E.
        • Su L.
        • Williams G.B.
        • Firbank M.J.
        • Lawson R.A.
        • Yarnall A.J.
        • et al.
        Baseline and longitudinal grey matter changes in newly diagnosed Parkinson’s disease: ICICLE-PD study.
        Brain. 2015; 138: 2974-2986
        • Carriere N.
        • Besson P.
        • Dujardin K.
        • Duhamel A.
        • Defebvre L.
        • Delmaire C.
        • Devos D.
        Apathy in Parkinson’s disease is associated with nucleus accumbens atrophy: A magnetic resonance imaging shape analysis.
        Mov Disord. 2014; 29: 897-903
        • Möller C.
        • Dieleman N.
        • van der Flier W.M.
        • Versteeg A.
        • Pijnenburg Y.
        • Scheltens P.
        • et al.
        More atrophy of deep gray matter structures in frontotemporal dementia compared to Alzheimer’s disease.
        J Alzheimers Dis. 2015; 44: 635-647
        • Graff-Radford J.
        • Williams L.
        • Jones D.T.
        • Benarroch E.E.
        Caudate nucleus as a component of networks controlling behavior.
        Neurology. 2017; 89: 2192-2197
        • Pasquini J.
        • Durcan R.
        • Wiblin L.
        • Gersel Stokholm M.
        • Rochester L.
        • Brooks D.J.
        • et al.
        Clinical implications of early caudate dysfunction in Parkinson’s disease.
        J Neurol Neurosurg Psychiatry. 2019; 90: 1098-1104
        • Luo X.
        • Mao Q.
        • Shi J.
        • Wang X.
        • Li C.-S.R.
        Putamen gray matter volumes in neuropsychiatric and neurodegenerative disorders.
        World J Psychiatry Ment Health Res. 2019; 3: 1020
        • Satizabal C.L.
        • Adams H.H.H.
        • Hibar D.P.
        • White C.C.
        • Knol M.J.
        • Stein J.L.
        • et al.
        Genetic architecture of subcortical brain structures in 38,851 individuals.
        Nat Genet. 2019; 51: 1624-1636
        • Glahn D.C.
        • Kent J.W.
        • Sprooten E.
        • Diego V.P.
        • Winkler A.M.
        • Curran J.E.
        • et al.
        Genetic basis of neurocognitive decline and reduced white-matter integrity in normal human brain aging.
        Proc Natl Acad Sci U S A. 2013; 110: 19006-19011
        • Hunter D.J.
        Gene–environment interactions in human diseases.
        Nat Rev Genet. 2005; 6: 287-298
        • Jack C.R.
        • Holtzman D.M.
        Biomarker modeling of Alzheimer’s disease.
        Neuron. 2013; 80: 1347-1358
        • Hibar D.P.
        • Stein J.L.
        • Renteria M.E.
        • Arias-Vasquez A.
        • Desrivières S.
        • Jahanshad N.
        • et al.
        Common genetic variants influence human subcortical brain structures.
        Nature. 2015; 520: 224-229
        • Hibar D.P.
        • Adams H.H.H.
        • Jahanshad N.
        • Chauhan G.
        • Stein J.L.
        • Hofer E.
        • et al.
        Novel genetic loci associated with hippocampal volume.
        Nat Commun. 2017; 8: 13624
        • Bis J.C.
        • DeCarli C.
        • Smith A.V.
        • van der Lijn F.
        • Crivello F.
        • Fornage M.
        • et al.
        Common variants at 12q14 and 12q24 are associated with hippocampal volume.
        Nat Genet. 2012; 44: 545-551
        • Foley S.F.
        • Tansey K.E.
        • Caseras X.
        • Lancaster T.
        • Bracht T.
        • Parker G.
        • et al.
        Multimodal brain imaging reveals structural differences in Alzheimer’s disease polygenic risk carriers: A study in healthy young adults.
        Biol Psychiatry. 2017; 81: 154-161
        • Brouwer R.M.
        • Panizzon M.S.
        • Glahn D.C.
        • Hibar D.P.
        • Hua X.
        • Jahanshad N.
        • et al.
        Genetic influences on individual differences in longitudinal changes in global and subcortical brain volumes: Results of the ENIGMA plasticity working group: Heritability Estimates of Brain Changes.
        Hum Brain Mapp. 2017; 38: 4444-4458
        • Chouinard-Decorte F.
        • McKay D.R.
        • Reid A.
        • Khundrakpam B.
        • Zhao L.
        • Karama S.
        • et al.
        Heritable changes in regional cortical thickness with age.
        Brain Imaging Behav. 2014; 8: 208-216
        • Pomponio R.
        • Erus G.
        • Habes M.
        • Doshi J.
        • Srinivasan D.
        • Mamourian E.
        • et al.
        Harmonization of large MRI datasets for the analysis of brain imaging patterns throughout the lifespan.
        NeuroImage. 2020; 208: 116450
        • Montagni I.
        • Guichard E.
        • Kurth T.
        Association of screen time with self-perceived attention problems and hyperactivity levels in French students: A cross-sectional study.
        BMJ Open. 2016; 6e009089
        • 3C Study Group
        Vascular factors and risk of dementia: Design of the three-city study and baseline characteristics of the study population.
        Neuroepidemiology. 2003; 22: 316-325
        • Maillard P.
        • Delcroix N.
        • Crivello F.
        • Dufouil C.
        • Gicquel S.
        • Joliot M.
        • et al.
        An automated procedure for the assessment of white matter hyperintensities by multispectral (T1, T2, PD) MRI and an evaluation of its between-centre reproducibility based on two large community databases.
        Neuroradiology. 2008; 50: 31-42
        • Yang J.
        • Lee S.H.
        • Goddard M.E.
        • Visscher P.M.
        GCTA: A tool for genome-wide complex trait analysis.
        Am J Hum Genet. 2011; 88: 76-82
        • Almasy L.
        • Blangero J.
        Multipoint quantitative-trait linkage analysis in general pedigrees.
        Am J Hum Genet. 1998; 62: 1198-1211
        • Bulik-Sullivan B.K.
        • Loh P.-R.
        • Finucane H.K.
        • Ripke S.
        • Yang J.
        • Patterson N.
        • et al.
        LD Score regression distinguishes confounding from polygenicity in genome-wide association studies.
        Nat Genet. 2015; 47: 291-295
        • Bulik-Sullivan B.
        • Finucane H.K.
        • Anttila V.
        • Gusev A.
        • Day F.R.
        • Loh P.-R.
        • et al.
        An atlas of genetic correlations across human diseases and traits.
        Nat Genet. 2015; 47: 1236-1241
        • Yang J.
        • Zaitlen N.A.
        • Goddard M.E.
        • Visscher P.M.
        • Price A.L.
        Advantages and pitfalls in the application of mixed-model association methods.
        Nat Genet. 2014; 46: 100-106
        • Li J.
        • Ji L.
        Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix.
        Heredity. 2005; 95: 221-227
        • Gusev A.
        • Ko A.
        • Shi H.
        • Bhatia G.
        • Chung W.
        • Penninx B.W.J.H.
        • et al.
        Integrative approaches for large-scale transcriptome-wide association studies.
        Nat Genet. 2016; 48: 245-252
        • Giambartolomei C.
        • Vukcevic D.
        • Schadt E.E.
        • Franke L.
        • Hingorani A.D.
        • Wallace C.
        • Plagnol V.
        Bayesian test for colocalisation between pairs of genetic association studies using summary statistics.
        PLoS Genet. 2014; 10e1004383
        • Kang H.J.
        • Kawasawa Y.I.
        • Cheng F.
        • Zhu Y.
        • Xu X.
        • Li M.
        • et al.
        Spatio-temporal transcriptome of the human brain.
        Nature. 2011; 478: 483-489
        • Schwartzentruber J.
        • Cooper S.
        • Liu J.Z.
        • Barrio-Hernandez I.
        • Bello E.
        • Kumasaka N.
        • et al.
        Genome-wide meta-analysis, fine-mapping and integrative prioritization implicate new Alzheimer’s disease risk genes.
        Nat Genet. 2021; 53: 392-402
        • Nalls M.A.
        • Blauwendraat C.
        • Vallerga C.L.
        • Heilbron K.
        • Bandres-Ciga S.
        • Chang D.
        • et al.
        Identification of novel risk loci, causal insights, and heritable risk for Parkinson’s disease: A meta-analysis of genome-wide association studies.
        Lancet Neurol. 2019; 18: 1091-1102
        • Davies G.
        • Lam M.
        • Harris S.E.
        • Trampush J.W.
        • Luciano M.
        • Hill W.D.
        • et al.
        Study of 300,486 individuals identifies 148 independent genetic loci influencing general cognitive function.
        Nat Commun. 2018; 9: 2098
        • Zhu Z.
        • Zheng Z.
        • Zhang F.
        • Wu Y.
        • Trzaskowski M.
        • Maier R.
        • et al.
        Causal associations between risk factors and common diseases inferred from GWAS summary data.
        Nat Commun. 2018; 9: 224
        • Batouli S.A.H.
        • Trollor J.N.
        • Wen W.
        • Sachdev P.S.
        The heritability of volumes of brain structures and its relationship to age: A review of twin and family studies.
        Ageing Res Rev. 2014; 13: 1-9
        • Giedd J.N.
        • Schmitt J.E.
        • Neale M.C.
        Structural brain magnetic resonance imaging of pediatric twins.
        Hum Brain Mapp. 2007; 28: 474-481
        • Lamballais S.
        • Jansen P.R.
        • Labrecque J.A.
        • Ikram M.A.
        • White T.
        Genetic scores for adult subcortical volumes associate with subcortical volumes during infancy and childhood.
        Hum Brain Mapp. 2021; 42: 1583-1593
        • Vanlandewijck M.
        • He L.
        • Mäe M.A.
        • Andrae J.
        • Ando K.
        • Del Gaudio F.
        • et al.
        A molecular atlas of cell types and zonation in the brain vasculature.
        Nature. 2018; 554: 475-480
        • He L.
        • Vanlandewijck M.
        • Mäe M.A.
        • Andrae J.
        • Ando K.
        • Del Gaudio F.
        • et al.
        Single-cell RNA sequencing of mouse brain and lung vascular and vessel-associated cell types.
        Sci Data. 2018; 5: 180160
        • Davis J.S.
        • Hassanzadeh S.
        • Winitsky S.
        • Lin H.
        • Satorius C.
        • Vemuri R.
        • et al.
        The overall pattern of cardiac contraction depends on a spatial gradient of myosin regulatory light chain phosphorylation.
        Cell. 2001; 107: 631-641
        • Kohda M.
        • Tokuzawa Y.
        • Kishita Y.
        • Nyuzuki H.
        • Moriyama Y.
        • Mizuno Y.
        • et al.
        A comprehensive genomic analysis reveals the genetic landscape of mitochondrial respiratory chain complex deficiencies.
        PLoS Genet. 2016; 12e1005679
        • Valente L.
        • Tiranti V.
        • Marsano R.M.
        • Malfatti E.
        • Fernandez-Vizarra E.
        • Donnini C.
        • et al.
        Infantile encephalopathy and defective mitochondrial DNA translation in patients with mutations of mitochondrial elongation factors EFG1 and EFTu.
        Am J Hum Genet. 2007; 80: 44-58
        • Chauhan G.
        • Adams H.H.H.
        • Bis J.C.
        • Weinstein G.
        • Yu L.
        • Töglhofer A.M.
        • et al.
        Association of Alzheimer disease GWAS loci with MRI-markers of brain aging.
        Neurobiol Aging. 2015; 36: 1765.e7-1765.e16
        • Lupton M.K.
        • Strike L.
        • Hansell N.K.
        • Wen W.
        • Mather K.A.
        • Armstrong N.J.
        • et al.
        The effect of increased genetic risk for Alzheimer’s disease on hippocampal and amygdala volume.
        Neurobiol Aging. 2016; 40: 68-77
        • Stocker H.
        • Möllers T.
        • Perna L.
        • Brenner H.
        The genetic risk of Alzheimer’s disease beyond APOE ε4: Systematic review of Alzheimer’s genetic risk scores.
        Transl Psychiatry. 2018; 8: 166
        • Mormino E.C.
        • Sperling R.A.
        • Holmes A.J.
        • Buckner R.L.
        • De Jager P.L.
        • Smoller J.W.
        • Sabuncu M.R.
        Polygenic risk of Alzheimer disease is associated with early- and late-life processes.
        Neurology. 2016; 87: 481-488
        • Harrison J.R.
        • Mistry S.
        • Muskett N.
        • Escott-Price V.
        From polygenic scores to precision medicine in Alzheimer’s disease: A systematic review.
        J Alzheimers Dis. 2020; 74: 1271-1283
        • Axelrud L.K.
        • Santoro M.L.
        • Pine D.S.
        • Talarico F.
        • Gadelha A.
        • Manfro G.G.
        • et al.
        Polygenic risk score for Alzheimer’s disease: Implications for memory performance and hippocampal volumes in early life.
        Am J Psychiatry. 2018; 175: 555-563
        • Kochunov P.
        • Williamson D.E.
        • Lancaster J.
        • Fox P.
        • Cornell J.
        • Blangero J.
        • Glahn D.C.
        Fractional anisotropy of water diffusion in cerebral white matter across the lifespan.
        Neurobiol Aging. 2012; 33: 9-20