LETTER doi:10.1038/nature10781 Genetic contributions to stability and change in intelligence from childhood to old age Ian J.Deary1.2*,Jian Yang3*,Gail Davies2,Sarah E.Harris2.4,Albert Tenesa4.5,David Liewald2,Michelle Luciano.2, Lorna M.Lopez2,Alan J.Gow.2,Janie Corley,Paul Redmond',Helen C.Fox,Suzanne J.Rowe,Paul Haggarty', Geraldine McNeill,Michael E.Goddards,David J.Porteous2.4,Lawrence J.Whalley6,John M.Starr2.9&Peter M.Visscher2.3.10.+ Understanding the determinants of healthy mental ageing is a life course are largely unreplicated2.Therefore,an important novel priority for society today.So far,we know that intelligence dif- contribution would be to partition the covariance between intelligence ferences show high stability from childhood to old ageand there scores at either end of the human life course into genetic and environ- are estimates of the genetic contribution to intelligence at different mental causes.To address this,the present study applies a new ana- agess6.However,attempts to discover whether genetic causes con- lytical method'3-7 to genome-wide association data from human tribute to differences in cognitive ageing have been relatively un- participants with general cognitive ability test scores in childhood informative-1.Here we provide an estimate of the genetic and and again in old age. environmental contributions to stability and change in intelligence Participants were members of the Aberdeen Birth Cohort 1936 across most of the human lifetime.We used genome-wide single (ABC1936)and the Lothian Birth Cohorts of 1921 and 1936 nucleotide polymorphism (SNP)data from 1,940 unrelated indi- (LBC1921,LBC1936)112.They are community-dwelling,surviving viduals whose intelligence was measured in childhood (age 11 members of the Scottish Mental Surveys of 1932(the 1921-born indi- years)and again in old age (age 65,70 or 79 years)12.We use a viduals)and 1947(the 1936-born individuals),in which they took a statistical method that allows genetic(co)variance to be estimated well-validated test of general intelligence (Moray House Test)at a from SNP data on unrelated individuals'3-17.We estimate that mean age of 11 years.They were traced and re-tested again in old causal genetic variants in linkage disequilibrium with common age on a large number of medical and psychosocial factors for studies SNPs account for 0.24 of the variation in cognitive ability change of healthy mental and physical ageing.Here,we use cognitive ability from childhood to old age.Using bivariate analysis,we estimate a test data from childhood and from the first occasion of testing in old genetic correlation between intelligence at age 11 years and in old age for each subject.For all three cohorts,cognitive ability in old age age of 0.62.These estimates,derived from rarely available data on was measured using the first unrotated principal component from a lifetime cognitive measures,warrant the search for genetic causes number of diverse cognitive tests.Additionally,the LBC1921 and of cognitive stability and change. LBC1936 cohorts re-took the Moray House Test in old age.Thus, General cognitive ability (also known as general intelligence,org) the present study partitions into genetic and environmental causes is an important human trait.It shows consistent and strong associa- the variance in stability and change in general intelligence over a tions with important life outcomes such as educational and occu- period of between 54 and 68 years.Testing for 599,011 SNPs was pational success,social mobility,health,illness and survival's. performed on the Illumina610-Quadvl chip(Illumina);the genotyp- Maintaining good general cognitive ability in old age is associated with ing of the samples in this study was described previously7 and quality better physical health and the ability to carry out everyday tasks0 control is described in Methods Summary. Intelligence differences are highly heritable from adolescence,and To estimate additive genetic and environmental contributions to through adulthood to old age36.Long-term follow-up studies have variation in cognitive ageing we used genotype information from shown that about half ofthe phenotypic variance in general intelligence 536,295 genome-wide autosomal SNPs.The method used here is a in old age is accounted for by its measure in childhood4.The corollary multivariate extension ofour recently developed method,which allows of this is that there are systematic changes through the life course in the the estimation of distant relationships between conventionally un- rank order of intelligence between people;that is,some people's related individuals from the SNP data and correlates genome-wide intelligence ages better than others.The determinants of stability SNP similarity with phenotypic similarity.A detailed description and change in intelligence across the human life course are being of the overall approach and statistical methods is given in Supplemen- sought,and candidate determinants include a wide range of genetic tary Fig.I and the Supplementary Note.We used a linear mixed model and environmental factors'.3.3192122.There have been longitudinal to estimate variance components.The methodology for the estimation studies within childhood/adolescence,middle adulthood and old age, of genetic variation from population samples was described previously but none that stretches from childhood to old age with the same indi- and has been applied to continuous traits,including height,body- viduals(to our knowledge).Until now,the proportion of the variance mass index and cognitive ability,and to disease The method in lifetime cognitive stability and change explained by genetic and is analogous to a pedigree analysis,with the important difference that environmental causes has been almost unknown.Apart from a small we estimate distant relatedness from SNP markers.Because the rela- contribution from variation in the APOE gene,suggested individual tionships are estimated from common SNP markers,phenotypic vari- genetic contributions to stability and change in intelligence across the ance explained by such estimated relationships is due to linkage Department of Psychology,University of Edinburgh,7 George Square,Edinburgh EH8 9JZ UK.Centre for Cognitive Ageing and Cognitive Epidemiology.University of Edinburgh,7 George Square, Edinburgh EH8 9JZ,UK.Queensland Institute of Medical Research,300 Herston Road,Brisbane,Queensland 4006,Australia.Medical Genetics Section,Molecular Medicine Centre,Institute of Genetics and Molecular Medicine,Wester General Hospital,Edinburgh EH4 2XU,UK 5The Roslin Institute,Royal(Dick)School of Veterinary Studies,University of Edinburgh,Roslin,Edinburgh EH25 9RG,UK. Institute of Applied Health Sciences,Foresterhill,Aberdeen AB25 2ZD,UK.Nutrition and Epigenetics Group,Rowett Institute of Nutrition and Health,University of Aberdeen,Greenburn Road,Bucksburn Aberdeen AB21 9SB,UK Department of Food and Agricultural Systems,University of Melbourne,Parkville,Victoria 3011,Australia and Victorian Department of Primary Industries,Bundoora,Victoria 3083.Australia.Alzheimer Scotland Dementia Research Centre,Universityof Edinburgh,7George Square,Edinburgh EH89JZ,UKUniversiyof Queensland Diamantina Institute,Universiy of Queensland,Princess Alexandra Hospital,Brisbane,Queensland 4102,Australia 1The Queensland Brain Institute,The University of Queensland,Brisbane,Queensland 4072,Australia *These authors contributed equally to this work 212 NATUREI VOL 4829 FEBRUARY 2012 2012 Macmillan Publishers Limited.All rights reserved
LETTER doi:10.1038/nature10781 Genetic contributions to stability and change in intelligence from childhood to old age Ian J. Deary1,2*, Jian Yang3 *, Gail Davies1,2, Sarah E. Harris2,4, Albert Tenesa4,5, David Liewald1,2, Michelle Luciano1,2, Lorna M. Lopez1,2, Alan J. Gow1,2, Janie Corley1 , Paul Redmond1 , Helen C. Fox6 , Suzanne J. Rowe5 , Paul Haggarty7 , Geraldine McNeill6 , Michael E. Goddard8 , David J. Porteous2,4, Lawrence J. Whalley6 , John M. Starr2,9 & Peter M. Visscher2,3,10,11* Understanding the determinants of healthy mental ageing is a priority for society today1,2. So far, we know that intelligence differences show high stability from childhood to old age3,4 and there are estimates of the genetic contribution to intelligence at different ages5,6. However, attempts to discover whether genetic causes contribute to differences in cognitive ageing have been relatively uninformative7–10. Here we provide an estimate of the genetic and environmental contributions to stability and change in intelligence across most of the human lifetime. We used genome-wide single nucleotide polymorphism (SNP) data from 1,940 unrelated individuals whose intelligence was measured in childhood (age 11 years) and again in old age (age 65, 70 or 79 years)11,12. We use a statistical method that allows genetic (co)variance to be estimated from SNP data on unrelated individuals13–17. We estimate that causal genetic variants in linkage disequilibrium with common SNPs account for 0.24 of the variation in cognitive ability change from childhood to old age. Using bivariate analysis, we estimate a genetic correlation between intelligence at age 11 years and in old age of 0.62. These estimates, derived from rarely available data on lifetime cognitive measures, warrant the search for genetic causes of cognitive stability and change. General cognitive ability (also known as general intelligence, or g18) is an important human trait. It shows consistent and strong associations with important life outcomes such as educational and occupational success, social mobility, health, illness and survival18. Maintaining good general cognitive ability in old age is associated with better physical health and the ability to carry out everyday tasks19,20. Intelligence differences are highly heritable from adolescence, and through adulthood to old age5,6. Long-term follow-up studies have shown that about half of the phenotypic variance in general intelligence in old age is accounted for by its measure in childhood3,4. The corollary of this is that there are systematic changes through the life course in the rank order of intelligence between people; that is, some people’s intelligence ages better than others. The determinants of stability and change in intelligence across the human life course are being sought, and candidate determinants include a wide range of genetic and environmental factors1,5,7,19,21,22. There have been longitudinal studies within childhood/adolescence, middle adulthood and old age, but none that stretches from childhood to old age with the same individuals (to our knowledge). Until now, the proportion of the variance in lifetime cognitive stability and change explained by genetic and environmental causes has been almost unknown. Apart from a small contribution from variation in the APOE gene, suggested individual genetic contributions to stability and change in intelligence across the life course are largely unreplicated22. Therefore, an important novel contribution would be to partition the covariance between intelligence scores at either end of the human life course into genetic and environmental causes. To address this, the present study applies a new analytical method13–17 to genome-wide association data from human participants with general cognitive ability test scores in childhood and again in old age. Participants were members of the Aberdeen Birth Cohort 1936 (ABC1936) and the Lothian Birth Cohorts of 1921 and 1936 (LBC1921, LBC1936)11,12,17. They are community-dwelling, surviving members of the Scottish Mental Surveys of 1932 (the 1921-born individuals) and 1947 (the 1936-born individuals), in which they took a well-validated test of general intelligence (Moray House Test) at a mean age of 11 years. They were traced and re-tested again in old age on a large number of medical and psychosocial factors for studies of healthy mental and physical ageing. Here, we use cognitive ability test data from childhood and from the first occasion of testing in old age for each subject. For all three cohorts, cognitive ability in old age was measured using the first unrotated principal component from a number of diverse cognitive tests. Additionally, the LBC1921 and LBC1936 cohorts re-took the Moray House Test in old age. Thus, the present study partitions into genetic and environmental causes the variance in stability and change in general intelligence over a period of between 54 and 68 years. Testing for 599,011 SNPs was performed on the Illumina610-Quadv1 chip (Illumina); the genotyping of the samples in this study was described previously17 and quality control is described in Methods Summary. To estimate additive genetic and environmental contributions to variation in cognitive ageing we used genotype information from 536,295 genome-wide autosomal SNPs. The method used here is a multivariate extension of our recently developed method, which allows the estimation of distant relationships between conventionally unrelated individuals from the SNP data and correlates genome-wide SNP similarity with phenotypic similarity13,15. A detailed description of the overall approach and statistical methods is given in Supplementary Fig. 1 and the Supplementary Note. We used a linear mixed model to estimate variance components. The methodology for the estimation of genetic variation from population samples was described previously and has been applied to continuous traits, including height, bodymass index and cognitive ability13,15–17, and to disease23. The method is analogous to a pedigree analysis, with the important difference that we estimate distant relatedness from SNP markers. Because the relationships are estimated from common SNP markers, phenotypic variance explained by such estimated relationships is due to linkage 1 Department of Psychology, University of Edinburgh, 7 George Square, Edinburgh EH8 9JZ, UK. 2 Centre for Cognitive Ageing and Cognitive Epidemiology, University of Edinburgh, 7 George Square, Edinburgh EH8 9JZ, UK. 3 Queensland Institute of Medical Research, 300 Herston Road, Brisbane, Queensland 4006, Australia. 4 Medical Genetics Section, Molecular Medicine Centre, Institute of Genetics and Molecular Medicine, Western General Hospital, Edinburgh EH4 2XU, UK. 5 The Roslin Institute, Royal (Dick) School of Veterinary Studies, University of Edinburgh, Roslin, Edinburgh EH25 9RG, UK. 6 Institute of Applied Health Sciences, Foresterhill, Aberdeen AB25 2ZD, UK. 7 Nutrition and Epigenetics Group, Rowett Institute of Nutrition and Health, University of Aberdeen, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, UK. 8 Department of Food and Agricultural Systems, University of Melbourne, Parkville, Victoria 3011, Australia and Victorian Department of Primary Industries, Bundoora, Victoria 3083, Australia. 9 Alzheimer Scotland Dementia Research Centre, University of Edinburgh, 7 George Square, Edinburgh EH8 9JZ, UK. 10University of Queensland Diamantina Institute, University of Queensland, Princess Alexandra Hospital, Brisbane, Queensland 4102, Australia. 11The Queensland Brain Institute, The University of Queensland, Brisbane, Queensland 4072, Australia. *These authors contributed equally to this work. 212 | NATURE | VOL 482 | 9 FEBRUARY 2012 ©2012 Macmillan Publishers Limited. All rights reserved
ETTER RESEARCH disequilibrium between the genotyped markers and unknown causal derived from a first-order Taylor series of the logarithm of the likelihood variants4.The method estimates genetic variation from SNPs that about the parameter estimates?s and these can be biased for modest are in linkage disequilibrium with unknown causal variants,and so sample sizes.A more appropriate procedure is to use the likelihood- provides a lower limit of the total narrow sense heritability because ratio test statistic to test the hypotheses that the genetic correlation additive variation due to variants that are not in linkage disequilibrium coefficient is zero(no genetic correlation)or 1 (perfect genetic correla- with the genotyped SNPs is not captured. tion).When using a likelihood-ratio test,the estimated genetic correla- We first performed a univariate analysis of cognitive ageing(Sup- tion coefficient of0.62 has a borderline significant difference from zero plementary Note),which we had defined previously as intelligence (likelihood-ratio test statistic =2.56,P=0.055,one-sided test)(Sup- scores in old age phenotypically adjusted for intelligence at childhood, plementary Fig.2),and does not differ significantly from 1.This was by fitting the Moray House Test of intelligence at age 11 as a linear tested by fitting a repeatability model (which implies a genetic correla- covariate24.We estimated that 0.24(standarderror 0.20)ofphenotypic tion of 1.0 and the same heritability of repeat observations)that has variance in cognitive ageing was accounted for by the SNP-based three fewer parameters than the full bivariate model.It resulted in a similarity matrix.We next conducted a bivariate genetic analysis of very similar value of the maximum log-likelihood value;the likelihood- intelligence scores early and later in life,to partition the observed ratio test statistic was 5.6(P=0.133,3 degrees of freedom)(Sup- phenotypic covariance in intelligence measured in childhood and plementary Table 3). old age into genetic and environmental sources of variation.Informa- LBC1921 and LBC1936had the same Moray House Test administered tion on the environmental correlation comes from the comparison of at age 11 and again in old age.The bivariate analyses were repeated, the two phenotypes within individuals whereas the genetic correlation therefore,using the same test ofintelligence in childhood and old age in is inferred from between-individual comparisons of the two pheno- this subsample of the cohorts.The phenotypic correlation between types(Supplementary Note).That is,the analysis can inform us about Moray House Test intelligence at age 11 and in old age was 0.68 genetic and environmental contributions to stability and change in (standard error 0.01)(Table 1).The bivariate analysis resulted in intelligence across the life course.The phenotypic correlation between estimates of the proportion of phenotypic variation explained by all Moray House Test intelligence at age 11 and the general intelligence SNPs for the Moray House Test,as follows:0.30(standard error 0.23) component in old age was 0.63 (standard error 0.02)(Table 1).The at age 11;and 0.29(standard error 0.22)at age 70-79.The genetic bivariate analysis resulted in estimates of the proportion ofphenotypic correlation between these two traits was 0.80 (standard error 0.27) variation explained by all SNPs for cognition,as follows:0.48(standard When using a likelihood-ratio test,the estimated genetic correlation error 0.18)at age 11;and 0.28(standard error 0.18)at age 65,70 or 79 coefficient of 0.80 is not significantly different from zero (likelihood- (referred to hereafter as 65-79).The genetic correlation between these ratio test statistic 1.51,P=0.11).The environmental correlation two traits was 0.62 (standard error 0.22),and the environmental cor- between these two traits was 0.63 (standard error 0.13).From the relation was 0.65(standard error 0.12).From the results of the bivariate results of the bivariate analyses we can make a prediction of the pro- analyses we can make a prediction about the proportion of phenotypic portion of phenotypic variance explained by the SNPs for the Moray variance explained by the SNPs for cognition at 65-79 years given the House Test at 70-79 years conditional on the phenotype at age phenotype at age 11 years.This provided a prediction of0.21(standard 11 years.This results in an estimate of0.074(standard error 0.24)(Sup- error 0.20),which is consistent with the actual estimate of 0.24 plementary Table 4).Although the standard errors of the estimates are (standard error 0.20)from the univariate analysis (Supplementary larger because a smaller data set was used,the results are similar to Table 1),suggesting that the bivariate normal distribution assumption those using the full data and it appears that the choice of phenotype at old age (Moray House Test or a linear combination of a number of underlying the bivariate analysis is reasonable.Hence,the results from the bivariate analysis contain the full description of the genetic and tests)has not led to a bias in inference.The estimates suggest that environmental relationships between cognition at childhood,cog- cognition early and late in life are similar traits,with possibly some nition at old age,and cognitive change.We re-ran this model with genetic variation for cognitive change. different cut-offs for relatedness(Supplementary Table 2).The esti- Using population-based genetic analyses,we have quantified,for the mates are very similar but with,as expected,larger standard errors for first time,the genetic and environmental contribution to stability and more stringent cut-offs,which result in a smaller sample size.This change in intelligence differences for most of the human lifespan. shows that the results are not driven by unusually high correlations Genetic factors seem to contribute much to the stability of intelligence differences across the majority of the human lifespan.We provide a for a few close relatives. In the present analyses we did not adopt the usual procedure of lower limit of the narrow sense heritability of lifetime cognitive ageing. dividing the parameter estimates by the standard errors to obtain test The point estimate using a general cognitive ability component in old statistics and accompanying P values,because the standard errors were age is 0.24,albeit with a large standard error(0.20).We describe the estimate as a lower limit because the methods used in the present study allow us only to estimate the proportion of the genetic variation con- Table 1 Bivariate analysis of intelligence at age 11 and at age 65-79 tributing to cognitive ageing that is captured by genetic variants in linkage disequilibrium with common SNPs;this will be lower than the Using general intelligence Using Moray House total narrow sense heritability.We do not have a good estimate of the component in old age Test in old age total amount of additive genetic variation for cognitive ageing,and so Estimate Standard error* Estimate Standard error* we cannot easily quantify any heritability that is missing from our h12 0.478 0.177 029R 022g estimate.Some of the possible genetic contribution we have found to n2 0.280 0.177 0.289 0.221 cognitive change might be attributable to developmental change G 0.623 0.218 0.798 0.266 0.652 0.125 0.630 0.132 between age 11 and young adulthood.However,the large phenotypic e 0.627 0.015 0.680 0.014 correlation between age 11 and old-age intelligence,and the fact that haare variance explained by all SNPs tor intelligence at age 11 and old age. heritability estimates ofgeneral intelligence by age 11 are at about adult respectively:re is genetic correlation:r is residual correlation:re is phenotypic correlation.A total of levels3,lead us to posit that most of the genetic variation we have found 1.940 unrelated individuals were included with the general intelligence component phenotype data at is a contribution to ageing-related cognitive changes.The estimate of childhood (1830)or old age(1839)(1,729 individuals had both phenotypes).Of the 1,515 LBC1921 and LBC1936 individuals,there were 1,391 with geneticinformation and Moray House Test scores both the genetic contribution to lifetime cognitive change was lower when, at age 11 and in old age. for a subsample,the same test was used in childhood and old age. .The standard errors are estimated from a first-order Taylo nsion about the estimated maximumlikelihood values and may be biased downwards For testing hypotheses we have used the The bivariate analysis conducted here quantifies how differences likelihood-ratio test statistic.which is more accurate. in intelligence early and late in life are attributable to environmental 9 FEBRUARY 2012 VOL 482 NATURE 213 2012 Macmillan Publishers Limited.All rights reserved
disequilibrium between the genotyped markers and unknown causal variants13,14,21. The method estimates genetic variation from SNPs that are in linkage disequilibrium with unknown causal variants, and so provides a lower limit of the total narrow sense heritability because additive variation due to variants that are not in linkage disequilibrium with the genotyped SNPs is not captured. We first performed a univariate analysis of cognitive ageing (Supplementary Note), which we had defined previously as intelligence scores in old age phenotypically adjusted for intelligence at childhood, by fitting the Moray House Test of intelligence at age 11 as a linear covariate24. We estimated that 0.24 (standard error 0.20) of phenotypic variance in cognitive ageing was accounted for by the SNP-based similarity matrix. We next conducted a bivariate genetic analysis of intelligence scores early and later in life, to partition the observed phenotypic covariance in intelligence measured in childhood and old age into genetic and environmental sources of variation. Information on the environmental correlation comes from the comparison of the two phenotypes within individuals whereas the genetic correlation is inferred from between-individual comparisons of the two phenotypes (Supplementary Note). That is, the analysis can inform us about genetic and environmental contributions to stability and change in intelligence across the life course. The phenotypic correlation between Moray House Test intelligence at age 11 and the general intelligence component in old age was 0.63 (standard error 0.02) (Table 1). The bivariate analysis resulted in estimates of the proportion of phenotypic variation explained by all SNPs for cognition, as follows: 0.48 (standard error 0.18) at age 11; and 0.28 (standard error 0.18) at age 65, 70 or 79 (referred to hereafter as 65–79). The genetic correlation between these two traits was 0.62 (standard error 0.22), and the environmental correlation was 0.65 (standard error 0.12). From the results of the bivariate analyses we can make a prediction about the proportion of phenotypic variance explained by the SNPs for cognition at 65–79 years given the phenotype at age 11 years. This provided a prediction of 0.21 (standard error 0.20), which is consistent with the actual estimate of 0.24 (standard error 0.20) from the univariate analysis (Supplementary Table 1), suggesting that the bivariate normal distribution assumption underlying the bivariate analysis is reasonable. Hence, the results from the bivariate analysis contain the full description of the genetic and environmental relationships between cognition at childhood, cognition at old age, and cognitive change. We re-ran this model with different cut-offs for relatedness (Supplementary Table 2). The estimates are very similar but with, as expected, larger standard errors for more stringent cut-offs, which result in a smaller sample size. This shows that the results are not driven by unusually high correlations for a few close relatives. In the present analyses we did not adopt the usual procedure of dividing the parameter estimates by the standard errors to obtain test statistics and accompanying P values, because the standard errors were derived from a first-order Taylor series of the logarithm of the likelihood about the parameter estimates25 and these can be biased for modest sample sizes. A more appropriate procedure is to use the likelihoodratio test statistic to test the hypotheses that the genetic correlation coefficient is zero (no genetic correlation) or 1 (perfect genetic correlation).When using a likelihood-ratio test, the estimated genetic correlation coefficient of 0.62 has a borderline significant difference from zero (likelihood-ratio test statistic 5 2.56, P 5 0.055, one-sided test) (Supplementary Fig. 2), and does not differ significantly from 1. This was tested by fitting a repeatability model (which implies a genetic correlation of 1.0 and the same heritability of repeat observations) that has three fewer parameters than the full bivariate model. It resulted in a very similar value of the maximum log-likelihood value; the likelihoodratio test statistic was 5.6 (P 5 0.133, 3 degrees of freedom) (Supplementary Table 3). LBC1921 and LBC1936 had the sameMoray House Test administered at age 11 and again in old age. The bivariate analyses were repeated, therefore, using the same test of intelligence in childhood and old age in this subsample of the cohorts. The phenotypic correlation between Moray House Test intelligence at age 11 and in old age was 0.68 (standard error 0.01) (Table 1). The bivariate analysis resulted in estimates of the proportion of phenotypic variation explained by all SNPs for the Moray House Test, as follows: 0.30 (standard error 0.23) at age 11; and 0.29 (standard error 0.22) at age 70–79. The genetic correlation between these two traits was 0.80 (standard error 0.27). When using a likelihood-ratio test, the estimated genetic correlation coefficient of 0.80 is not significantly different from zero (likelihoodratio test statistic 5 1.51, P 5 0.11). The environmental correlation between these two traits was 0.63 (standard error 0.13). From the results of the bivariate analyses we can make a prediction of the proportion of phenotypic variance explained by the SNPs for the Moray House Test at 70–79 years conditional on the phenotype at age 11 years. This results in an estimate of 0.074 (standard error 0.24) (Supplementary Table 4). Although the standard errors of the estimates are larger because a smaller data set was used, the results are similar to those using the full data and it appears that the choice of phenotype at old age (Moray House Test or a linear combination of a number of tests) has not led to a bias in inference. The estimates suggest that cognition early and late in life are similar traits, with possibly some genetic variation for cognitive change. Using population-based genetic analyses, we have quantified, for the first time, the genetic and environmental contribution to stability and change in intelligence differences for most of the human lifespan. Genetic factors seem to contribute much to the stability of intelligence differences across the majority of the human lifespan. We provide a lower limit of the narrow sense heritability of lifetime cognitive ageing. The point estimate using a general cognitive ability component in old age is 0.24, albeit with a large standard error (0.20). We describe the estimate as a lower limit because the methods used in the present study allow us only to estimate the proportion of the genetic variation contributing to cognitive ageing that is captured by genetic variants in linkage disequilibrium with common SNPs; this will be lower than the total narrow sense heritability. We do not have a good estimate of the total amount of additive genetic variation for cognitive ageing, and so we cannot easily quantify any heritability that is missing from our estimate. Some of the possible genetic contribution we have found to cognitive change might be attributable to developmental change between age 11 and young adulthood. However, the large phenotypic correlation between age 11 and old-age intelligence, and the fact that heritability estimates of general intelligence by age 11 are at about adult levels5 , lead us to posit that most of the genetic variation we have found is a contribution to ageing-related cognitive changes. The estimate of the genetic contribution to lifetime cognitive change was lower when, for a subsample, the same test was used in childhood and old age. The bivariate analysis conducted here quantifies how differences in intelligence early and late in life are attributable to environmental Table 1 | Bivariate analysis of intelligence at age 11 and at age 65–79 Using general intelligence component in old age Using Moray House Test in old age Estimate Standard error* Estimate Standard error* h1 2 0.478 0.177 0.298 0.229 h2 2 0.280 0.177 0.289 0.221 rG 0.623 0.218 0.798 0.266 re 0.652 0.125 0.630 0.132 rP 0.627 0.015 0.680 0.014 Where h1 2 and h2 2 are variance explained by all SNPs for intelligence at age 11 and old age, respectively; rG is genetic correlation; re is residual correlation; rP is phenotypic correlation. A total of 1,940 unrelated individuals were included with the general intelligence component phenotype data at childhood (1,830) or old age (1,839) (1,729 individuals had both phenotypes). Of the 1,515 LBC1921 and LBC1936 individuals, there were 1,391 with genetic information and Moray House Test scores both at age 11 and in old age. * The standard errors are estimated from a first-order Taylor series expansion about the estimated maximum likelihood values and may be biased downwards25. For testing hypotheses we have used the likelihood-ratio test statistic, which is more accurate. LETTER RESEARCH 9 FEBRUARY 2012 | VOL 482 | NATURE | 213 ©2012 Macmillan Publishers Limited. All rights reserved
RESEARCH LETTER or genetic factors.A genetic correlation of zero would imply that METHODS SUMMARY intelligence early and late in life are entirely separate traits genetically, Subjects.Recruitment,phenotyping and genotyping of the samples were and that variation in the change in intelligence from childhood to old described previously The mental test at age 11 was a Moray House age is partly genetic and a function of the heritability of intelligence Test In old age,general intelligence was derived using principal components early and late in life.At the other extreme,a genetic correlation of one analysis of a number of mental tests and saving scores on the first unrotated implies that the two traits have the same genetic determinants,so that principal component(Supplementary Note).In old age,the assessments ofgeneral any variation in the change in intelligence between the two stages in life intelligence were made at ages as follows:ABC1936,64.6 years(standard devi- ation 0.9);LBC1936,69.5(standard deviation 0.8);LBC1921,79.1 (standard devi- is purely environmental.At conventional levels of significance we ation 0.6).The LBC1921 and LBC1936 samples,but not the ABC1936,had repeat could not rule out either a genetic correlation of zero or one;however, testing of the Moray House Test(already taken at age 11 years)at 79.1 and 69.5 our estimates suggest that genetics and environment could each con- years,respectively.After applying the genome-wide complex trait analysis tribute substantially to the covariance between intelligence at age 11 method 5,the distribution of inferred relationships in the samples was as shown and old age,and that genetic factors might have a role in cognitive in Supplementary Fig.3.We removed one of each pair of individuals whose change between the two stages of the life course. estimated genetic relatedness was >0.2.We retained 1,940 individuals with child- The samples studied here comprise the birth cohorts'survivors, hood or old-age phenotype data (1,729 individuals had both):ABC1936,425; those healthy enough to take part in the studies,and people with less LBC1921,512;and LBC1936,1,003.Of the 1,515 LBC1921 and LBC1936 indivi- cognitive decline.Therefore,we considered whether our estimate of duals,there were 1,391 with genetic information and Moray House Test scores at age 11 and in old age genetic variation at older ages may be biased downwards because of Genotyping quality control.Quality control procedures were performedper SNP censoring.From life tables officially published by the Scottish and per sample.Individuals were excluded from further analysis if genetic and Government based on census data,we estimate that the individuals reported gender did not agree.Samples with a call rate 0.95,and those showing in our oldest sample who were born in 1921 and alive at age 11 are evidence of non-European descent by multidimensional scaling,were removed" among the~50%that were still alive at the time of sample collection. SNPs were included in the analyses if they met the following conditions:call We know that lower childhood cognitive ability per se is associated rate 20.98,minor allele frequency 20.01,and Hardy-Weinberg equilibrium test with premature mortality26,which,of course,our analyses adjust for,as with P20.001.After these quality control stages,1,948 samples remained specified in the models.However,because there is a paucity of data (ABC1936,N=426;LBC1921,N=517;LBC1936,N=1,005),and536,295 autosomal SNPs were included in the analysis about genetic influences on lifetime cognitive change,we have limited information with regard to how these might affect life expectancy.The Received 5 September:accepted 12 December 2011. only way to know across the lifespan would have been if all children Published online 18 January 2012 (that is,the ones who survived to older ages-whom we know about- and the ones who did not)had been genotyped in 1947.For non- Plassman,B.L,Williams,J.W.Burke,J.R.Holsinger,T.Benjamin,S.Systematic normative (that is,pathological)cognitive change,there are genetic review:factors associated with risk for and possible prevention of decline in later life.Ann.Inter Med.153,182-193(2010). risk factors associated with younger-onset Alzheimer's disease that 2 Brayne.C.The elephant in the room-healthy brains in later life,epidemiology and result in premature mortality,but such strongly heritable disease is public health.Nature Rev.Neurosci.8,233-239(2007). rare and the genes do not seem to affect normative cognitive ageing in Deary.I.J.Whalley,LJ,Lemmon,H.,Crawford,J.R.Starr,J.M.The stability of those aged 70 years and over22.Hence,this is not a concern with regard individual differences in mental ability from childhood to old age:follow-up of the 1932 Scottish Mental Survey.Intelligence 28,49-55(2000). to our analyses.APOE 84 is a well-known risk factor for non-normative 4 Gow,A.J.etal.Stability and change in intelligence from age 11 to ages 70,79,and cognitive decline,but any differential effect on survival occurs later in 87:the Lothian Birth Cohorts of 1921 and 1936.Psychol Aging 26,232-240 life,and is thus unlikely to have resulted in attrition in our cohort. (2011) Deary.I.J..Johnson.W.Houlihan,L M.Genetic foundations of human Moreover,APOE is in Hardy-Weinbergequilibrium in even our oldest intelligence.Hum.Genet 126,215-232(2009). samples24,supporting this inference.Other known genetic risk factors 6. Deary,I.J.,Penke,L Johnson,W.The neuroscience of human intelligence for Alzheimer's disease have a very small effect on the risk of disease27 differences.Nature Rev.Neurosci.11,201-211(2010). Hence,a priori,we have nothing to suggest anything but a largely 1 Lee,T.,Henry,J.D.,Trollor,J.N.Sachdev,P.S.Genetic influences on cognitive functions in the elderly:a selective review of twin studies.Brain Res Rev.64,1-13 neutral effect of genes that influence cognitive ageing on survival. (2010). However,if there is an effect,the example of cognition2(by contrast Reynolds,C.A.etal.Quantitative genetic analysis of latent growth curve models of with cognitive change)would suggest that this would be negative, cognitive abilities in adulthood.Dev.Psychol.41,3-16(2005). 9. Finkel,D.,Reynolds,C.A.McArdle,J.J.Hamagami,F.Pedersen,N.L Genetic which would somewhat reduce genetic variation in cognitive change eed drives variation in aging of spatial and memory across the lifespan among the survivors. Until now,studies aimed at finding genetic contributions to cognitive 10.McGue,M.&Christensen,K.Social activity and healthy aging:a study of aging Danish twins.Twin Res.Hum.Genet 10,255-265 (2007). ageing have offered little information.They use too-short follow-up 11.Deary,I Whiteman,M.C.Starr,J.M.,Whalley,LJ.&Fox,H.C.The impact of periods,thereby providing too small an amount of cognitive change?. childhood intelligence in later life:following up the Scottish Mental Surveys of Cognitive assessments tend to be made only within old age,even though 1932and1947.J.Pes.Soc.Psychol..86.130-147(2004). cognitive ageing occurs from young adulthood onwards.They are 12.Deary,I.J.et al.The Lothian Birth Cohort 1936:a study to examine influences on cognitive ageing from age 11 to age 70 and beyond.BMC Geriatr.7,28(2007). largely based on behavioural data in twin samples rather than informa- 13.Yang.J.et al.Common SNPs explain a large proportion of the heritability for tion on DNA variation.The present study is unusual and valuable in human height.Nature Genet 42,565-569(2010) capturing over half a century of cognitive stability and change and 14.Visscher,P.M,Yang.J.Goddard,M.E.Acommentary on 'common SNPs explain a large proportion of the heritability for human height'by Yang et al.(2010).Twin examining its causes.The results here provide estimates for the genetic Res.Hum.Genet13,517-524(2010). and environmental contributions to cognitive stability and change 15.Yang,J.Lee,H.,Goddard,M.E Visscher,P.M.GCTA:a tool for genome-wide across most of the human lifespan.Even with almost 2,000 individuals, complex trait analysis.Am.J.Hum.Genet 88,76-82(2011). the study's power was insufficient to achieve conventional levels of 16. Yang,J.et al.Genome partitioning of genetic variation for complex traits using common SNPs.Nature Genet 43,519-525(2011). significance for the estimates.Our emphasis here has not been on 17.Davies.G.et al Genome-wide association studies establish that human the traditional significance thresholds for P values per se,but in trying intelligence is highly heritable and polygenic.Mol.Psychiatry 16,996-1005 to partition variance in cognitive ability into environmental and 2011 18. Deary,I.J.Intelligence.Annu.Rev.Psychol.63,453-482 (2012). genetic causes.The phenotypes available here are rare,and so these 19.Deary,I.J.etal.Age-associated cognitive decline.Br.Med.Bull 92,135-152 point estimates are useful to guide future research.The present find- (2009) ings render attractive a search for genetic mechanisms of cognitive 20.Tucker-Drob,E.M.Neurocognitive functions a nd eve yday functions change change across the life course.They also suggest the importance of together in old age.Neuropsychology 5,368-377(2011) 21.Powell,J.E Visscher,P.M.Goddard,M.E.Reconciling the analysis of IBD and environmental contributions to lifetime cognitive change. IBS in complex trait studies.Nature Rev.Genet 11,800-805(2010). 214 NATURE VOL 4829 FEBRUARY 2012 2012 Macmillan Publishers Limited.All rights reserved
or genetic factors. A genetic correlation of zero would imply that intelligence early and late in life are entirely separate traits genetically, and that variation in the change in intelligence from childhood to old age is partly genetic and a function of the heritability of intelligence early and late in life. At the other extreme, a genetic correlation of one implies that the two traits have the same genetic determinants, so that any variation in the change in intelligence between the two stages in life is purely environmental. At conventional levels of significance we could not rule out either a genetic correlation of zero or one; however, our estimates suggest that genetics and environment could each contribute substantially to the covariance between intelligence at age 11 and old age, and that genetic factors might have a role in cognitive change between the two stages of the life course. The samples studied here comprise the birth cohorts’ survivors, those healthy enough to take part in the studies, and people with less cognitive decline. Therefore, we considered whether our estimate of genetic variation at older ages may be biased downwards because of censoring. From life tables officially published by the Scottish Government based on census data, we estimate that the individuals in our oldest sample who were born in 1921 and alive at age 11 are among the ,50% that were still alive at the time of sample collection. We know that lower childhood cognitive ability per se is associated with premature mortality26, which, of course, our analyses adjust for, as specified in the models. However, because there is a paucity of data about genetic influences on lifetime cognitive change, we have limited information with regard to how these might affect life expectancy. The only way to know across the lifespan would have been if all children (that is, the ones who survived to older ages—whom we know about— and the ones who did not) had been genotyped in 1947. For nonnormative (that is, pathological) cognitive change, there are genetic risk factors associated with younger-onset Alzheimer’s disease that result in premature mortality, but such strongly heritable disease is rare and the genes do not seem to affect normative cognitive ageing in those aged 70 years and over22. Hence, this is not a concern with regard to our analyses.APOE e4 is a well-known risk factor for non-normative cognitive decline, but any differential effect on survival occurs later in life, and is thus unlikely to have resulted in attrition in our cohort. Moreover,APOE is in Hardy–Weinberg equilibrium in even our oldest samples24, supporting this inference. Other known genetic risk factors for Alzheimer’s disease have a very small effect on the risk of disease27. Hence, a priori, we have nothing to suggest anything but a largely neutral effect of genes that influence cognitive ageing on survival. However, if there is an effect, the example of cognition26 (by contrast with cognitive change) would suggest that this would be negative, which would somewhat reduce genetic variation in cognitive change across the lifespan among the survivors. Until now, studies aimed atfinding genetic contributions to cognitive ageing have offered little information. They use too-short follow-up periods, thereby providing too small an amount of cognitive change7,22. Cognitive assessments tend to be made only within old age, even though cognitive ageing occurs from young adulthood onwards. They are largely based on behavioural data in twin samples rather than information on DNA variation. The present study is unusual and valuable in capturing over half a century of cognitive stability and change and examining its causes. The results here provide estimates for the genetic and environmental contributions to cognitive stability and change across most of the human lifespan. Even with almost 2,000 individuals, the study’s power was insufficient to achieve conventional levels of significance for the estimates. Our emphasis here has not been on the traditional significance thresholds for P values per se, but in trying to partition variance in cognitive ability into environmental and genetic causes. The phenotypes available here are rare, and so these point estimates are useful to guide future research. The present findings render attractive a search for genetic mechanisms of cognitive change across the life course. They also suggest the importance of environmental contributions to lifetime cognitive change. METHODS SUMMARY Subjects. Recruitment, phenotyping and genotyping of the samples were described previously11,12,17. The mental test at age 11 was a Moray House Test11,12. In old age, general intelligence was derived using principal components analysis of a number of mental tests and saving scores on the first unrotated principal component (Supplementary Note). In old age, the assessments of general intelligence were made at ages as follows: ABC1936, 64.6 years (standard deviation 0.9); LBC1936, 69.5 (standard deviation 0.8); LBC1921, 79.1 (standard deviation 0.6). The LBC1921 and LBC1936 samples, but not the ABC1936, had repeat testing of the Moray House Test (already taken at age 11 years) at 79.1 and 69.5 years, respectively. After applying the genome-wide complex trait analysis method13,15, the distribution of inferred relationships in the samples was as shown in Supplementary Fig. 3. We removed one of each pair of individuals whose estimated genetic relatedness was .0.2. We retained 1,940 individuals with childhood or old-age phenotype data (1,729 individuals had both): ABC1936, 425; LBC1921, 512; and LBC1936, 1,003. Of the 1,515 LBC1921 and LBC1936 individuals, there were 1,391 with genetic information and Moray House Test scores at age 11 and in old age. Genotyping quality control. Quality control procedures were performed per SNP and per sample. Individuals were excluded from further analysis if genetic and reported gender did not agree. Samples with a call rate # 0.95, and those showing evidence of non-European descent by multidimensional scaling, were removed17. SNPs were included in the analyses if they met the following conditions: call rate $ 0.98, minor allele frequency $ 0.01, and Hardy–Weinberg equilibrium test with P $ 0.001. After these quality control stages, 1,948 samples remained (ABC1936, N 5 426; LBC1921, N 5 517; LBC1936, N 5 1,005), and 536,295 autosomal SNPs were included in the analysis. Received 5 September; accepted 12 December 2011. Published online 18 January 2012. 1. Plassman, B. L., Williams, J. W., Burke, J. R., Holsinger, T. & Benjamin, S. Systematic review: factors associated with risk for and possible prevention of decline in later life. Ann. Intern. Med. 153, 182–193 (2010). 2. Brayne, C. The elephant in the room—healthy brains in later life, epidemiology and public health. Nature Rev. Neurosci. 8, 233–239 (2007). 3. Deary, I. J., Whalley, L. J., Lemmon, H., Crawford, J. R. & Starr, J. M. The stability of individual differences in mental ability from childhood to old age: follow-up of the 1932 Scottish Mental Survey. Intelligence 28, 49–55 (2000). 4. Gow, A. J. et al. Stability and change in intelligence from age 11 to ages 70, 79, and 87: the Lothian Birth Cohorts of 1921 and 1936. Psychol. Aging 26, 232–240 (2011). 5. Deary, I. J., Johnson, W. & Houlihan, L. M. Genetic foundations of human intelligence. Hum. Genet. 126, 215–232 (2009). 6. Deary, I. J., Penke, L. & Johnson, W. The neuroscience of human intelligence differences. Nature Rev. Neurosci. 11, 201–211 (2010). 7. Lee, T., Henry, J. D., Trollor, J. N. & Sachdev, P. S. Genetic influences on cognitive functions in the elderly: a selective review of twin studies. Brain Res. Rev. 64, 1–13 (2010). 8. Reynolds, C. A. et al. Quantitative genetic analysis of latent growth curve models of cognitive abilities in adulthood. Dev. Psychol. 41, 3–16 (2005). 9. Finkel, D., Reynolds, C. A., McArdle, J. J., Hamagami, F. & Pedersen, N. L. Genetic variance in processing speed drives variation in aging of spatial and memory abilities. Dev. Psychol. 45, 820–834 (2009). 10. McGue, M. & Christensen, K. Social activity and healthy aging: a study of aging Danish twins. Twin Res. Hum. Genet. 10, 255–265 (2007). 11. Deary, I. J., Whiteman, M. C., Starr, J. M., Whalley, L. J. & Fox, H. C. The impact of childhood intelligence in later life: following up the Scottish Mental Surveys of 1932 and 1947. J. Pers. Soc. Psychol. 86, 130–147 (2004). 12. Deary, I. J. et al. The Lothian Birth Cohort 1936: a study to examine influences on cognitive ageing from age 11 to age 70 and beyond. BMC Geriatr. 7, 28 (2007). 13. Yang, J. et al. Common SNPs explain a large proportion of the heritability for human height. Nature Genet. 42, 565–569 (2010). 14. Visscher, P. M., Yang, J. & Goddard, M. E. A commentary on ‘common SNPs explain a large proportion of the heritability for human height’ by Yang et al. (2010). Twin Res. Hum. Genet. 13, 517–524 (2010). 15. Yang, J., Lee, H., Goddard, M. E. & Visscher, P. M. GCTA: a tool for genome-wide complex trait analysis. Am. J. Hum. Genet. 88, 76–82 (2011). 16. Yang, J. et al. Genome partitioning of genetic variation for complex traits using common SNPs. Nature Genet. 43, 519–525 (2011). 17. Davies, G. et al. Genome-wide association studies establish that human intelligence is highly heritable and polygenic. Mol. Psychiatry 16, 996–1005 (2011). 18. Deary, I. J. Intelligence. Annu. Rev. Psychol. 63, 453–482 (2012). 19. Deary, I. J. et al. Age-associated cognitive decline. Br. Med. Bull. 92, 135–152 (2009). 20. Tucker-Drob, E. M. Neurocognitive functions and everyday functions change together in old age. Neuropsychology 25, 368–377 (2011). 21. Powell, J. E., Visscher, P. M. & Goddard, M. E. Reconciling the analysis of IBD and IBS in complex trait studies. Nature Rev. Genet. 11, 800–805 (2010). RESEARCH LETTER 214 | NATURE | VOL 482 | 9 FEBRUARY 2012 ©2012 Macmillan Publishers Limited. All rights reserved
LETTER RESEARCH 22.Harris,S.E.Deary,I.J.The genetics of cognitive ability and cognitive ageing in Research Into Ageing(continues as part of Age UK's The Disconnected Mind project). healthy older people.Trends Cogn.Sci.15,388-394(2011). Phenotype colle ction in the ABC1936 was supported by the BBSRC,the Wellcome 23.Lee,S.H.Wray,N.R.Goddard,M.E.Visscher,P.M.Estimating missing Trust and the Alzheimer's Research Trust The Australian-based researchers heritability for disease from genome-wide complex trait analysis.Am.J.Hum acknowledge support from the Australian Research Council and the National Health Genet88,294-305(2011). and Medical Research Council.M.L is a Royal Society of Edinburgh/Lloyds TSB 24.Deary,I.J.etal Cognitive change and the APOE4 allele.Nature418,932(2002) Foundation for Scotland Personal Research Fellow.The work was undertaken in The 25.Gilmour,A.R.,Thompson,R.Cullis,B.R.Average information REML:an efficient University of Edinburgh Centre for Cognitive Ageing and Cognitive Epidemiology, algorithm for variance parameter estimation in linear mixed models.Biometrics part of the cross council Lifelong Health and Wellbeing Initiative(GO700704/ 51,1440-1450(1995). 84698),for which funding from the BBSRC,EPSRC,ESRC and MRC is gratefully 26.Calvin,C.M.et al.Intelligence in youth and all-cause mortality:systematic review acknowledged. with meta-analysis.Int.J.Epidemiol.40,626-644(2011). 27.Hollingworth,P.et al.Common variants at ABCA7,MS4A6A/MS4A4E,EPHAI. Author Contributions I.D.and P.M.V.designed the study.J.Y.and P.M.V.performed CD33 and CD2AP are associated with Alzheimer's disease.Nature Genet 43, statistical analyses,with I.J.D.,M.E.G.,A.T.and S.R.contributing to discussions 429-435(2011) regarding analyses.G.D.S.E.H.,D.L A.T.M.L.and L.M.L.performed quality control analysesandprepareddata.S.E.H.M.L,LM.L.AJ.G.J.C.P.R.H.C.F.,SJ.R.P.H.,LJ.W. Supplementary Information is linked to the online version of the paper at G.M.,DJ.P.,J.M.S.and IJ.D.contributed genotype and phenotype data.LJ.D.,P.M.V.and www.nature.com/nature. J.Y.contributed to writing the paper and Supplementary Information.All authors contributed to revising the paper and Supplementary Information. Acknowledgements We thank the cohort participants who contributed to these studies.Genotyping of the ABC1936,LBC1921 and LBC1936 cohorts and the Author Information Reprints and pemmissions information is available at analyses conducted here were supported by the UK's Biotechnology and Biological www.nature.com/reprints.The authors declare no competing financial interests. Sciences Research Council (BBSRC).Phenotype collection in the LBC1921 was Readers are welcome to comment on the online version of this article at supported by the BBSRC,The Royal Society and The Chief Scientist Office of the www.nature.com/nature.Correspondence and requests for materials should be Scottish Goverment.Phenotype collection in the LBC1936 was supported by addressed to I.D.(i.deary@ed.ac.uk)or P.M.V.(peter.visscher@ug.edu.au). 9 FEBRUARY 2012 VOL 482 NATURE 215 2012 Macmillan Publishers Limited.All rights reserved
22. Harris, S. E. & Deary, I. J. The genetics of cognitive ability and cognitive ageing in healthy older people. Trends Cogn. Sci. 15, 388–394 (2011). 23. Lee, S. H., Wray, N. R., Goddard, M. E. & Visscher, P. M. Estimating missing heritability for disease from genome-wide complex trait analysis. Am. J. Hum. Genet. 88, 294–305 (2011). 24. Deary, I. J. et al. Cognitive change and the APOE e4 allele. Nature 418, 932 (2002). 25. Gilmour, A. R., Thompson, R. & Cullis, B. R. Average information REML: an efficient algorithm for variance parameter estimation in linear mixed models. Biometrics 51, 1440–1450 (1995). 26. Calvin, C. M. et al. Intelligence in youth and all-cause mortality: systematic review with meta-analysis. Int. J. Epidemiol. 40, 626–644 (2011). 27. Hollingworth, P. et al. Common variants at ABCA7, MS4A6A/MS4A4E, EPHA1, CD33, and CD2AP are associated with Alzheimer’s disease. Nature Genet. 43, 429–435 (2011). Supplementary Information is linked to the online version of the paper at www.nature.com/nature. Acknowledgements We thank the cohort participants who contributed to these studies. Genotyping of the ABC1936, LBC1921 and LBC1936 cohorts and the analyses conducted here were supported by the UK’s Biotechnology and Biological Sciences Research Council (BBSRC). Phenotype collection in the LBC1921 was supported by the BBSRC, The Royal Society and The Chief Scientist Office of the Scottish Government. Phenotype collection in the LBC1936 was supported by Research Into Ageing (continues as part of Age UK’s The Disconnected Mind project). Phenotype collection in the ABC1936 was supported by the BBSRC, the Wellcome Trust and the Alzheimer’s Research Trust. The Australian-based researchers acknowledge support from the Australian Research Council and the National Health and Medical Research Council. M.L. is a Royal Society of Edinburgh/Lloyds TSB Foundation for Scotland Personal Research Fellow. The work was undertaken in The University of Edinburgh Centre for Cognitive Ageing and Cognitive Epidemiology, part of the cross council Lifelong Health and Wellbeing Initiative (G0700704/ 84698), for which funding from the BBSRC, EPSRC, ESRC and MRC is gratefully acknowledged. Author Contributions I.J.D. and P.M.V. designed the study. J.Y. and P.M.V. performed statistical analyses, with I.J.D., M.E.G., A.T. and S.J.R. contributing to discussions regarding analyses. G.D., S.E.H., D.L., A.T., M.L. and L.M.L. performed quality control analyses and prepared data. S.E.H., M.L., L.M.L., A.J.G., J.C., P.R., H.C.F., S.J.R., P.H., L.J.W., G.M., D.J.P., J.M.S. and I.J.D. contributed genotype and phenotype data. I.J.D., P.M.V. and J.Y. contributed to writing the paper and Supplementary Information. All authors contributed to revising the paper and Supplementary Information. Author Information Reprints and permissions information is available at www.nature.com/reprints. The authors declare no competing financial interests. Readers are welcome to comment on the online version of this article at www.nature.com/nature. Correspondence and requests for materials should be addressed to I.J.D. (i.deary@ed.ac.uk) or P.M.V. (peter.visscher@uq.edu.au). LETTER RESEARCH 9 FEBRUARY 2012 | VOL 482 | NATURE | 215 ©2012 Macmillan Publishers Limited. All rights reserved
