Genetic vulnerability

Research Review:Genetic vulnerability or differential susceptibility in child development:

the case of attachment

Marian J.Bakermans-Kranenburg and Marinus H.van IJzendoorn

Centre for Child and Family Studies,Leiden University,The Netherlands

Gene–environment interactions interpreted in terms of differential susceptibility may play a large part in the explanation of individual differences in human development.Reviewing studies on the behavioral and molecular genetics of attachment,we present evidence for interactions between genetic and envir-onmental factors explaining individual differences in attachment security and disorganization.In particular,the DRD47-repeat polymorphism seems associated with an increased risk for disorganized attachment,but only when combined with environmental risk.Gene–environment(G·E)interactions may be interpreted as genetic vulnerability or differential susceptibility.We found support for the differential susceptibility hypothesis predicting not only more negative outcomes for susceptible children in unfavorable environments,but also positive outcomes for susceptible children in favorable environments.Keywords:Attachment,gene–environment interaction,differential susceptibility, disorganized attachment,dopamine,behavior problems,genetics.

Attachment theory is an important theoretical framework for studying infant–parent relationships and individual differences in emotion regulation. Attachment is the evolutionary-based,innate predisposition of a child to seek proximity to and contact with a speci?c?gure,notably when the child is frightened,tired or ill(Bowlby,1969/1982).In the current review we take the development of attach-ment as an example to demonstrate the crucial role of gene–environment interactions in explaining individual developmental differences.We argue that gene–environment interactions should not be restricted to negative environmental in?uences, emphasizing genetic vulnerabilities,but should be interpreted in terms of differential susceptibility of children to environmental in?uences,‘for better or for worse’(Belsky,1997,2005).

The idea that the environment plays a major part in explaining individual differences in quality of the infant–parent attachment relationship is central to attachment theory(Ainsworth,Blehar,Waters,& Wall,1978;Main,1990).Parental sensitivity is generally regarded as the single most important determinant of infant attachment security,particu-larly for the three organized attachment strategies: insecure-avoidant,secure,and insecure-ambivalent attachment(Ainsworth et al.,1978;Bretherton, 1990;Pederson&Moran,1995).Moreover,meta-analytic results(Van IJzendoorn,1995)demon-strated that parents’sensitivity to their children’s attachment signals is partly determined by parents’own secure or insecure mental representation of attachment as assessed with the Adult Attachment Interview(AAI;Main,Kaplan,&Cassidy,1985;Hesse,1999).Maternal sensitivity,however, accounts for only a third of the association between parental attachment representation and infant attachment security,leaving a transmission gap of unexplained variance in infant attachment security (Van IJzendoorn,1995;Belsky,2005).Genetic factors affecting both security of attachment in the parent and in the offspring might bridge this gap.

Whereas secure as well as insecure avoidant or insecure ambivalent attachment relationships can be considered organized strategies,adaptive to the child’s environment(Main,1990)and speculated to enhance the child’s reproductive?tness(Belsky, 1999),some attachment relationships are charac-terized by the(momentary)absence or breakdown of an otherwise organized strategy,hence de?ned as disorganized(Main&Solomon,1990;Hesse&Main, 2006).Indices of disorganized attachment behavior are sequential or simultaneous display of contra-dictory behaviors,such as distress and avoidance, undirected or misdirected movements and expres-sions,stereotypical and anomalous movements or postures,freezing or stilling behaviors,or expres-sions of fear or apprehension regarding the parent (Main&Solomon,1990),which are unrelated to neurocognitive impairments in the child(Pipp-Siegel,Siegel,&Dean,1999).

The prevalence of attachment disorganization ranges from about15%in normal,non-clinical samples to80%in maltreated samples(Carlson, Cicchetti,Barnett,&Braunwald,1989).Disorgan-ized attachment and sensitivity have been found to be only weakly associated,with a combined meta-analytic effect size of r?.10(Van IJzendoorn, Schuengel,&Bakermans-Kranenburg,1999).But associations between speci?c environmental risk

Con?ict of interest statement:No con?icts declared.

Journal of Child Psychology and Psychiatry48:12(2007),pp1160–1173doi:10.1111/j.1469-7610.2007.01801.x

ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

Published by Blackwell Publishing,9600Garsington Road,Oxford OX42DQ,UK and350Main Street,Malden,MA02148,USA

factors,including parental unresolved loss or trauma,and later infant disorganization have been established in a number of studies(for a narrative review,see Lyons-Ruth&Jacobvitz,1999;for a meta-analysis,see Van IJzendoorn et al.,1999). Main and Hesse’s(1990)hypothesis of a link between frightening or frightened parental behavior and disorganized attachment has been tested and demonstrated(Abrams,Rifkin,&Hesse,2006; Grienenberger,Kelly,&Slade,2005;Jacobvitz, Leon,&Hazen,2006;Schuengel,Bakermans-Kranenburg,&Van IJzendoorn,1999;True,Pisani, &Oumar,2001).Lyons-Ruth,Bronfman,and Par-sons(1999)added extremely insensitive or disrupt-ive parental behaviors as potentially eliciting disorganization(see also Goldberg,Benoit,Blokland, &Madigan,2003;Madigan,Moran,&Pederson, 2006b).

In a recent meta-analysis on the determinants of infant attachment disorganization,Madigan et al. (2006a)found robust and moderate associations between maternal unresolved loss or trauma, maternal frightening or anomalous behavior,and disorganized attachment in12studies including more than800families in total.However,less than half of the association between unresolved loss/ trauma and disorganized attachment was explained by mediation of frightening parental behavior,and a second transmission gap has become apparent, pointing to the need for further exploration of infant, parental,ecological,and genetic moderating and mediating factors explaining disorganized attach-ment.Genes seem plausible candidates for bridging the transmission gap because parents and infants share50%of their genes,and intergenerational transmission of(disorganized)attachment may (partly)be based on transmission of genes from one generation to the next.

Behavioral genetics

Behavioral geneticists have stressed two major points:First,traditional research designs with one child per family,focusing on between-family com-parisons,may confound genetic similarities between parents and children with supposedly shared envir-onmental in?uences(Rowe,1994;Plomin,1994). Second,talking about the in?uence of‘the environ-ment’on child development confounds shared and unique environment.Indeed,behavioral genetic research on twins,siblings,and unrelated(adoptive or step-)children have changed our views on child development drastically,and it has been argued that we have overestimated the role of parents in child development(Harris,1998;but see Collins,Mac-coby,Steinberg,Hetherington,&Bornstein,2000; Rose,1995).

Much of the behavioral genetic research involves large samples but meager measurements;paper-and-pencil methods have been amazingly popular. The statistical need for relatively large samples has probably hampered the application of behavioral genetic methodologies in the?eld of attachment where detailed,labor-intensive observational assessments prevail.Only a few studies on attach-ment using genetically informative designs have been conducted.Nevertheless,behavior genetic methodologies can be viewed as useful natural experiments(Rutter,2003)to test developmental theories.

Behavioral genetics of attachment

Attachment of siblings and unrelated pairs. Genetically informative studies include pairs of children with divergent degrees of relatedness growing up in the same environment(or identical twins growing up in different families).Attachment theory would suggest that children growing up in the same family are likely to relate in similar ways to their parents,at least when parental attach-ment representations and interactive styles remain stable across time.A study with biological siblings (N?138pairs)who were observed with their mother in the Strange Situation Procedure(SSP)when each sibling was between12and14months of age indeed showed signi?cant concordance of attachment(62%; 68%for siblings of the same gender;Van IJzendoorn et al.,2000).Maternal insensitivity to both siblings (shared environment)was associated with concor-dance of sibling non-security.

In adoptive and step-families biologically unre-lated siblings share the same caregiver,and any concordance in attachment quality is thus accounted for by shared environmental in?uences. Unfortunately,no data on attachment for such families has become available yet.However,Sagi et al.(1995)reported on attachment of biologically unrelated‘siblings’in Israeli kibbutzim,where children typically spend much of their days with six infants and two(biologically unrelated)caregivers in an‘infant house’from their third month onwards. The classi?cations assigned to each pair of children observed with the same caregiver showed a concord-ance of70%(n?27pairs of children),suggesting that a substantial shared environmental component accounted for the variance in attachment security.

The designs of these kibbutz and sibling studies did not allow for distinguishing the role of(shared and non-shared)environment from the role of genes, because of the absence of contrast groups with varying degrees of genetic relatedness.The question of heritability of(disorganized)attachment could thus not be addressed.

Attachment of twins.The?rst study of attachment of twins goes back to the early seventies(Brooks& Lewis,1974).Attachment behavior of17pairs of

Research Review:Genetic vulnerability or differential susceptibility in child development1161ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

opposite-sex twins was observed during15min of free play of each of the twins with their mother.Proximity seeking,looking,touching,and pleasant vocal-izations directed toward the mother were scored. Because the study focused on differences in attach-ment behaviors between boys and girls no associa-tions between pairs were reported.Moreover,since only dizygotic(DZ)twins were included,any twin concordance in attachment behavior could not be compared to pairs with other degrees of relatedness.

In general,however,twin studies(including both MZ and DZ twins)provide a unique opportunity to test the similarity of siblings’attachment relation-ships under conditions of similar age and child-rearing contexts.Several studies on MZ and DZ twins’attachment have been conducted,usually involving relatively small samples.Ricciuti(1992) combined three small sample studies of twins (Goldberg,Perrotta,Minde,&Corter,1986,see also Minde,Corter,Goldberg,&Jeffers,1990;Goldsmith &Campos,1990;Vandell,Owen,Wilson,&Hen-derson,1988).In this combined sample of27DZ twin pairs Ricciuti(1992)found78%concordance of attachment security;concordance in the29MZ twin pairs was66%,and she therefore concluded that attachment security did not show genetic in?uence.

Finkel,Wille,and Matheny(1998)found67% concordance of security in34MZ twins but only38% concordance in26DZ twins,indicating only‘modest rearing environment effects’(Finkel et al.,1998, p.7).It should be noted that the percentage of DZ twin concordance is below chance(see Van IJzendoorn et al.,2000).Finkel and Matheny(2000) reported on the complete Louisville Twin Study of attachment(including the60twin pairs from the previous report).In108DZ twin pairs they found only48%concordance of attachment security, whereas in99MZ twin pairs66%concordance was found.Finkel and her colleagues(1998,2000)estim-ated the heritability of attachment at25%;the remaining variance was attributable to non-shared environmental in?uences,without any role for shared environmental in?uences.Finkel et al.(1998, 2000),however,used a separation–reunion proced-ure originally designed for assessing temperament, and it should be noted that this adapted attachment measure was only moderately associated with the Strange Situation procedure.

O’Connor and Croft(2001)observed110MZ and DZ twin pairs at43months in the Strange Situation procedure.Attachment was coded using the pre-school coding manual developed by Cassidy and Marvin in cooperation with the MacArthur Working Group on Attachment(1992).Their?ndings sug-gested a modest role for genetic in?uence(14%)and substantial in?uences of shared(32%)and non-shared(53%)environment.They did not?nd a mediating role for temperamental differences in explaining similarities or dissimilarities in attach-ment between the siblings.

Participants from two sites,London(United King-dom,62twin pairs)and Leiden(The Netherlands,76 twin pairs),were combined in Bokhorst et al.’s (2003)study on attachment in twins.The two stud-ies used similar designs and measures.For secure versus non-secure attachment as observed in the Strange Situation procedure52%of the variance in attachment security was explained by shared envir-onment,and48%of the variance was explained by unique environmental factors and measurement error.The role of genetic factors was negligible.In the same study,genetic factors did explain77%of the variance in temperamental reactivity(and non-shared environmental factors and measurement error23%),but differences in temperamental reac-tivity were not associated with attachment concor-dance.Similar results were shown for infant–father attachment and temperamental dependency,both assessed with the Attachment Q-Sort(Vaughn& Waters,1990;Waters,1995)in the Leiden part of this sample(Bakermans-Kranenburg,Van IJzen-doorn,Bokhorst,&Schuengel,2004).Attachment security was largely explained by shared environ-mental(59%)and unique environmental(41%)fac-tors.For temperamental dependency,genetic factors explained66%of the variance,and non-shared environmental factors34%.

Interestingly,disorganized attachment appears to behave differently when subjected to behavior-genetic analysis.So far only one study(Bokhorst et al.,2003)has applied behavioral genetics to dis-organized attachment.Genetic modeling pointed to only unique environmental factors explaining the variance in disorganization.

It should be noted that twin studies of adult attachment representations and parental sensitivity are lacking.Whether environmental in?uences are evident in measures of attachment in later childhood and adulthood remains an important question to be addressed in future research.So far,genetic studies have only covered self-reported adult attachment styles(Brussoni,Jang,Livesley,&MacBeth,2000; Neyer,2002)or self-reported parenting(Spinath& O’Connor,2003)–with diverse(and sometimes mixed)?ndings.Recently,the?rst study with AAIs conducted on twins has been reported(Constantino et al.,2006),but genetic analyses were impossible since the design included only MZ twins(with and without conduct problems)reared together.Inter-estingly,in a sample of genetically unrelated(adop-ted)sibling pairs(mean age39years),61% concordance of AAI attachment security was found, suggesting that shared environment in?uences sibling similarities in attachment in adulthood (Caspers,Yucuis,Troutman,Arndt,&Langbehn,in press).

The internal working models view of attachment would predict that continuity over time in patterns of attachment would be driven by shared environ-mental factors(Fraley,2002)–a prediction that

1162Marian J.Bakermans-Kranenburg and Marinus H.van IJzendoorn

ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

could be tested in longitudinal genetic studies of attachment.Several behavior geneticists have sug-gested that shared environmental factors are likely to be more signi?cant in infancy and early childhood than later in life,and indeed genetic studies of individual differences in mental development and temperament con?rm this view(e.g.,Plomin,1994; Plomin&Daniels,1987).Longitudinal genetic research would allow for the estimation of genetic and environmental in?uences at different develop-mental stages and of their in?uence on age-to-age changes in attachment security.

Shared and non-shared environment in attachment

What are these shared and non-shared environ-mental factors affecting infant attachment security, and what does the prominent role for unique environment in the case of disorganized attachment mean?For attachment security the importance of shared environmental factors con?rms one of the basic notions of attachment theory,and it may point to the vital importance of parental sensitivity.The question is whether parental sensitivity is shared or non-shared across twins or siblings and,critically, whether sensitivity is associated with the shared or non-shared aspects of attachment security.An important weakness of univariate behavioral genetic studies such as described above is that they quantify only the cumulative effects of a potentially large number of unspeci?ed processes(genetic or envir-onmental).Since proposed mechanisms or estab-lished correlates are not measured,it is dif?cult to evaluate how behavior-genetic?ndings?t with empirical data on associations,and whether causal interpretations of one kind or another are plausible. The quanti?cation of variance components is not informative about what speci?c factors actually play a part(Deater-Deckard,2000).In recent years there has been increasing recognition of the limitations of univariate methods and the bene?ts of multivariate designs(e.g.,Rutter,2003).

Using multivariate genetic analyses,Fearon et al. (2006)included both maternal sensitivity and infant attachment of the Bokhorst et al.(2003)twin sam-ple,and examined the extent to which genetic and environmental aspects of maternal sensitivity accounted for the pattern of similarity and dissimil-arity in twins’attachments to their mother.In theory,three factors might be responsible for the association between sensitivity and attachment. First,the association between sensitivity and attachment may be attributed to shared environ-mental factors.If maternal sensitivity is in?uenced by her attachment representation and this in turn causes variations in attachment security(Van IJz-endoorn,1995),then consistencies in parenting behavior between twins lead to consistencies in their attachment classi?cations(regardless of their

zygosity).In that case the attachment classi?cation

of each twin is equally well predicted from the

sensitivity shown towards either twin,regardless of

genetic similarity.In contrast,when genetic factors

underlie the association between maternal sensit-

ivity and attachment(sensitivity as experienced by

the child and attachment in?uenced by the same

genes),sensitivity expressed to one twin is a better

predictor of the other twin’s attachment classi?ca-

tion in MZ twins than in DZ twins.Finally,non-

shared environmental factors are present when a

parent’s sensitivity to one twin does not predict the

other twin’s attachment security;in such cases

strong within-twin correlations between sensitivity

and attachment are found but low cross-twin corre-

lations.Bivariate behavior genetic models use the

differences in within-twin and cross-twin correla-

tions to estimate genetic,shared environmental and

non-shared environmental correlations between two

measures(Neale&Cardon,1992;Plomin,DeFries,

McClearn,&McGuf?n,2001;Purcell,2001).

Fearon et al.(2006)found no evidence for a con-

tribution of genetic factors(residing in the infants)to

differences in maternal sensitivity.The correlations

for sensitivity as rated from 1.5hours of home

observations by independent coders for twin1and

twin2were high in both MZ and DZ twins(between.64

and.69).Genetic modeling indicated that shared

environmental factors explained66%of the variance

in sensitivity as experienced by the child,and non-

shared environmental factors34%.Using bivariate

behavior-genetic analyses it was then tested whether

shared environmental and/or non-shared environ-

mental processes were responsible for the link

between sensitivity and attachment(since there was

no genetic contribution to maternal sensitivity).

Shared environmental effects on sensitivity correlated

with shared environmental effects on attachment

(r C?.58),indicating that shared environmental effects underlie the association between maternal

sensitivity and attachment security,and that the

shared environmental component of maternal sensit-

ivity accounted for approximately33%of the corre-

lation between twins in attachment security.

Non-shared environmental effects created an

inverse relationship between sensitivity and attach-

ment.This negative non-shared effect was small in

size,accounting for only2%of the non-shared

variance in attachment.What this implies is that

non-shared environmental effects had the effect of

suppressing the phenotypic association resulting

from the shared environmental pathway between

sensitivity and attachment.Greater sensitivity to one

twin that was not expressed towards the other was

associated with lower likelihood of security in the

?rst twin.Or,alternatively,sensitivity towards twin

2(that was not shown to twin1)impacted positively

on the twin1’s attachment security.In short,shared

variance in maternal sensitivity accounted for part of

Research Review:Genetic vulnerability or differential susceptibility in child development1163ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

the similarity between twins in attachment security, but weak non-shared associations between sensit-ivity and attachment suppressed the magnitude of the correlation between attachment and sensitivity in twin children.The attachment security of one twin appears to depend on the relationship the parent has with the other twin.These?ndings underscore the importance of effects of relationships on relation-ships within a family system(Hinde&Stevenson-Hinde,1988).The identi?cation of such non-shared in?uences is essential for the understanding of attachment processes.

For disorganized attachment,at least three factors may contribute to the large effect of non-shared environment.First,in behavioral genetic modeling non-shared environment cannot be distinguished from measurement error.The coding of attachment disorganization might be liable to intercoder differ-ences and low reliability.It seems,however,unlikely that measurement error can account for a large part of the non-shared environmental factor,since the intercoder reliabilities in the pertinent studies were adequate,and both statistically and theoretically signi?cant associations have been reported across a substantial number of studies concerning the determinants and sequelae of disorganized attach-ment(Van IJzendoorn et al.,1999).

Second,unique characteristics of the children that are not shared by their(twin)siblings might be responsible for the effect of non-shared environment. These unique child characteristics might either make the children differentially susceptible to parenting in?uences(Belsky,1997,2005),or they may evoke differential parenting.Bokhorst et al. (2003)reported that differences in observed fearful-ness,parent-reported temperament,and peri-natal conditions and experiences between the twins were not associated with differences in attachment dis-organization.Nevertheless,aspects of parenting that are experienced by one but not the other twin may have consequences for each of the twins’attachment as was shown in the Fearon et al.(2006)study.In a similar vein,attachment disorganization of one twin may depend on the parenting behavior shown towards the other twin.Bivariate genetic studies of anomalous parenting and disorganized attachment would be highly informative in this regard.

Third,attachment disorganization may have vari-ous determinants,some of which may reside in the unique environment(e.g.,one twin sibling but not the other may have a near resemblance with a deceased relative,with unique consequences for parental caregiving),others in the shared environ-ment(e.g.,maternal depression,con?icts in the marital relationship)or in genetic factors(see below).

A large non-shared environmental effect may re?ect heterogeneity of the mechanisms leading to dis-organized attachment,and maybe also heterogeneity in the phenotype of disorganization.

Taken together,the in?uence of non-shared envir-onment is evident in organized and particularly in disorganized attachment behavior.New ways of conceptualizing and investigating these non-shared in?uences are essential for the understanding of attachment processes.Their identi?cation requires valid measures of environmental factors,together with designs that can separate genetic from envir-onmental in?uences.For disorganized attachment, the absence of any genetic effect might be somewhat surprising given the molecular genetic?ndings of a Hungarian research group(see below),and given the idea that dissociation might be the underlying mechanism of disorganization(Hesse&Main,2006). For dissociation some studies have reported a role for genetic factors(Becker-Blease et al.,2004;Jang et al.,1998),although others did not(Waller&Ross, 1997;see also Harari,Van IJzendoorn,&Baker-mans-Kranenburg,2007).

It should be noted that behavioral genetic studies of attachment used small samples and the con?dence boundaries of the reported values for both genetic and environmental factors are broad;in the case of dis-organized attachment they are even based on only one study.Furthermore,?ndings of behavioral genetic studies do not provide information about the relation between actual genes and behavior.Gene expression is not a straightforward process,and may be more dependent on environmental in?uences than once thought(e.g.,Meaney,2001;Meaney&Szyf,2005; Rutter,2006).Molecular genetic studies are a com-plement–and sometimes an alternative–to behav-ioral genetics.

Molecular genetics:the dopamine system DRD4and-521C/T.In a?rst molecular genetic study on attachment Lakatos and colleagues(Laka-tos et al.,2000)found an association between the dopamine D4receptor(DRD4)7-repeat allele and attachment disorganization.In previous studies the DRD4gene polymorphism had shown associations with pathological impulsive behavior and substance abuse in adults and attention de?cit hyperactivity disorder(ADHD)in children.The dopaminergic system is engaged in attentional,motivational,and reward mechanisms(Robbins&Everitt,1999),and the7-repeat allele has been linked to lower dopa-mine reception ef?ciency.In Lakatos’(2000)low-risk Hungary sample of95infants the estimated relative risk for disorganized attachment among children carrying the7-repeat allele was4.15.There was no association between the7-repeat allele and the three organized attachment classi?cations.

A second study with the same group of infants involved additional analyses including the functional -521C/T single nucleotide polymorphism(Lakatos et al.,2002).The-521T variant has been shown to reduce transcriptional ef?ciency by approximately

1164Marian J.Bakermans-Kranenburg and Marinus H.van IJzendoorn

ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

40%compared to the-521C allele(Okuyama, Ishiguro,Toru,&Arinami,1999),and may thus be relevant for the dopaminergic neurotransmission. Indeed the association between disorganized attachment and the7-repeat DRD4allele was stronger in the presence of the-521T variant.In the presence of both the DRD47-repeat and the-521T allele in the upstream regulatory region of the DRD4 gene the risk for disorganized attachment was sug-gested to increase tenfold(Lakatos et al.,2002).The -521C/T genotype itself had no effect on attachment classi?cation.

A third study followed(Gervai et al.,2005).Using genetic data of the parents of the same sample of infants as used in the previous two reports,the focus was now on the haplotype.Whereas in the Lakatos et al.(2002)study the mere presence of both a DRD4 7-repeat allele and a-521T variant in the child’s DNA had been associated with disorganized attach-ment,parental DNA now provided the possibility to determine whether the presence of the T.7haplotype (containing both DRD47-repeat allele and the-521T variant)was responsible for attachment https://www.wendangku.net/doc/107190598.html,ing transmission disequilibrium tests,it was tested whether heterozygous parents transmit-ted the T.7haplotype at higher-than-chance rate to disorganized infants.The frequencies of the haplo-types of the two polymorphisms(DRD47-repeat and -521T)did not differ between parents of disorga-nized and parents of secure children,and a prefer-ential transmission of the7-repeat allele from parents to disorganized children was not signi?cant (p?.07).Parental genetic data,therefore,did not con?rm the link between the two polymorphisms and disorganized attachment.However,the absence of the T.7haplotype was related to secure attach-ment(p?.01).

Need for replication.Since multiple gene effects are based on an increasingly large number of statistical tests,and are thus liable to false positives(Ebstein, Benjamin,&Belmaker,2002),and the number of participants with particular genotype combinations and the predicted phenotype becomes increasingly small(Van Gestel&Van Broekhoven,2003),there was a clear need for independent replication of the study.Replication and–in a later stage–meta-analyses are strongly recommended as strategies for rigorously testing the validity of reported gene to behavior associations(e.g.,Ioannidis,2003). Whereas a modest genetic effect would not be incompatible with an environmental explanation of disorganized attachment(e.g.,through differential susceptibility;Belsky,1997,2005),a tenfold increase of the risk to become disorganized as the result of two polymorphisms(Lakatos et al.,2002)is stronger than the developmental model of disorgan-ized attachment would suggest.

In the Leiden twin sample(Bokhorst et al.,2003) cheek cells were collected from132children at 50months of age.As described above,they had been observed with their mothers in the SSP at 13months.The Leiden twins did not show an asso-ciation of disorganized attachment with the7-repeat DRD4allele(p?.38),nor with the-521C/T geno-type(p?.20).The interaction was not signi?cant either(p?.12).In the presence of the-521T variant, the7-repeat DRD4allele did not increase the risk for attachment disorganization.Including one child of each twin pair did not change the outcomes;the Hungarian?ndings were not replicated.

No parental DNA was available for the Leiden twins,and thus any disequilibrium in the trans-mission of the T.7haplotype could not be tested.It was,however,determined which children were carry-ing a T.7haplotype and which children certainly had no T.7haplotype.Contrasting these children (n?115)for attachment security and disorganiza-tion no differences were found:secure attachment was evident in58%of the children with the T.7haplotype and in57%of the children without the T.7haplotype,and disorganized attachment was observed in23%of the children carrying the T.7haplotype and in19%of the children without the T.7haplotype(Bakermans-Kranenburg&Van IJzendoorn,2004).A second attempt to replicate Gervai et al.’s(2005)?nding of an association between the presence or absence of the T.7haplo-type and either attachment security or disorganiza-tion in a different sample also failed(Van IJzendoorn &Bakermans-Kranenburg,2006).

Further attempts followed.The appealing link between DRD47-repeat and disorganized attachment has now been tested in six samples that we are aware of,some of them with unpublished data on this issue (the Hungarian sample,Lakatos et al.,2000,2002; Gervai et al.,2005;the Leiden and London twin samples,Bakermans-Kranenburg&Van IJzendoorn, 2004;Bokhorst et al.,2003;Fearon et al.,2006;the Bielefeld study,Spangler&Zimmermann,2007;the Leiden study on intergenerational transmission of disorganized attachment,Schuengel et al.,1999;Van IJzendoorn&Bakermans-Kranenburg,2006;and the Leiden Longitudinal Study,Van der Mark,Van IJzendoorn,&Bakermans-Kranenburg,2002).The combined sample size of these studies amounts to 542infant–mother dyads.Examining the association between the presence of the DRD47-repeat allele and disorganized attachment in this combined group shows no signi?cant effect,v2(df?1,N?542)?.29, p?.66.

Summarizing,we may say that the pioneering study of the Hungarian group(Gervai et al.,2005; Lakatos et al.,2000,2002)reported a potentially important role of two polymorphisms related to the dopamine system,but independent replication of the study did not con?rm the role of the DRD47-repeat allele and the-521C/T promoter gene in disorgan-ized attachment–although the power to?nd the tenfold increase in disorganized attachments in the

Research Review:Genetic vulnerability or differential susceptibility in child development1165ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

attempted replication was more than95% (Bakermans-Kranenburg&Van IJzendoorn,2004).

The combined effect size for the association between DRD4and disorganized attachment is d?.05.It should be noted that this is much smaller than the combined effect size for the association between anomalous parenting and disorganized attachment(Madigan et al.,2006b),which is d?.72(N?644).In previous studies the dopamine system has been associated with attention,state regulation,and orienting responses in children,but also with alcohol and drug addiction in adults (Benjamin,Ebstein,&Belmaker,2002;Ebstein et al.,2002).Its in?uence on human functioning seems to be broad-band and rather non-speci?c. State regulation and orientation to stimuli in the outside world may concern cognitive processes involved in the emergence of attachments–as well as in other aspects of socio-emotional development.If further research were to con?rm that the dopamine system is important for state regulation in infants, infants with less adequate state regulatory abilities may have more problems with processing potentially inconsistent and paradoxical communication from parents suffering from unresolved loss.However,the scarce behavioural and genetic data are not unequivocal.Spangler et al.(Spangler,Fremmer-Bombik&Grossmann,1996;Spangler&Gross-mann,1999)reported an association between attachment disorganization and lower newborn orienting ability and regulation.Somewhat para-doxically,Ebstein et al.(1998)found the DRD4 7-repeat allele to be associated with better infant orientation and state regulation.The latter results do not seem to support the idea that infants with the 7-repeat DRD4polymorphism were overrepresented in Spangler et al.’s low-oriented newborns who were later classi?ed as disorganized.Now that genetic data of Spangler et al.’s(1996)sample are available (Spangler&Zimmermann,2007),it appears that the DRD47-repeat allele indeed did not play a role in explaining disorganized attachment in this sample. Gene–environment interaction

In view of the behavioral genetic and molecular studies on attachment,it must be concluded that the empirical evidence to date falls short in explaining the presence or absence of secure and disorganized attachment behavior on the basis of mere genetic factors.An alternative or complementary explana-tion including genetic contributions to disorganized attachment pertains to the possibility of gene–envir-onment interaction,the interplay between‘nature’and‘nurture’deriving from genetic effects on liability to risk exposure and susceptibility to environmental risks(e.g.,Bennett et al.,2002;Caspi et al.,2002, 2003;Fox et al.,2005;Kaufman et al.,2004).As demonstrated by Suomi(1999)for primates and by Caspi et al.(2002,2003)for humans,an individual’s response to environmental in?uences may be mod-erated by his or her genetic make-up(gene–environment interaction).Studies examining the in?uence of(measured)gene by(measured)envir-onment(G·E)interactions illuminate how inherit-ance contributes to both the dynamics and the outcome of development(Michel&Moore,1995; Mof?tt,2005;Mof?tt,Caspi,&Rutter,2005;Rutter, 2006;Rutter,Mof?tt,&Caspi,2006).A failure to include G·E interaction may lead to an overestima-tion of the genetic component in the case of an interaction between genes and shared environ-ment(Rutter,Pickles,Murray,&Eaves,2001),and to an overestimation of the non-shared environ-mental effects when actual interaction effects between genes and non-shared environment are not taken into account(Eaves,Silberg,&Erkanli,2003). The?nding that in the only genetically informative study of attachment disorganization(Bokhorst et al., 2003)all variance was attributed to the non-shared environmental component might re?ect this omis-sion of potential G·E effects.

For disorganized attachment,the crucial question may be whether infants with the7-repeat DRD4 allele are more susceptible to parental unresolved loss and anomalous parenting behavior than infants without this allele.Recently,the?rst study com-bining molecular genetics with measurement of the pertinent environmental in?uences on disorganized attachment has been conducted(Van IJzendoorn& Bakermans-Kranenburg,2006).Unresolved loss or trauma as a distal index of a risk environment and frightening maternal behavior as a proximal risk factor in the emergence of attachment disorganiza-tion were included.In a sample of63mothers who experienced at least one important loss(Schuengel et al.,1999),unresolved loss was assessed with the AAI.In the AAI(George,Kaplan,&Main,1985;Main et al.,1985)the interviewee is asked about childhood experiences and the in?uence of these experiences,but also about any loss of loved ones and other traumatic experiences.Independent from secure or insecure mental representation of attach-ment,an unresolved classi?cation is assigned if lapses in the monitoring of reasoning or discourse appear in the discussion of loss or trauma experi-ences or if the participant describes extreme behav-ioral reactions in connection with these experiences (Main&Goldwyn,1984/1998).Frightening/fright-ened behavior(Main&Hesse,1992/1995)of the mothers was coded on the basis of3to4hours of home observations when the infants were10to 11months old.The Strange Situation procedure (Ainsworth et al.,1978)was administered between 14and15months.

A moderating role of the DRD4gene was found: Maternal unresolved loss or trauma was associated with infant disorganization,but only in the presence of the DRD47-repeat allele in the child.Children

1166Marian J.Bakermans-Kranenburg and Marinus H.van IJzendoorn

ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

with the short DRD4allele did not show higher scores for disorganized attachment when their mother was unresolved.The increase in risk for disorganization in children with the7-repeat allele exposed to maternal unresolved loss/trauma com-pared to children without these combined risks was 18.8fold.The T.7haplotype showed a similar interaction effect:an elevated risk for infant dis-organization in the case of maternal unresolved loss (odds ratio 3.24).However,because children with long DRD4combined with-521CC were absent in the sample the effects of the haplotype containing both the long DRD4and the-521T alleles could not be distinguished from the effects of DRD4alone.For maternal frightening behaviour no moderating effects were found.

What do these?ndings imply?First,all particip-ating mothers experienced the loss of a relative.The absence of an association between DRD47-repeat and infant disorganization in this group of mothers means that the environmental risk of being exposed to a parent who suffered an important loss is insuf-?cient for the child to become disorganized.As attachment theory would predict(Ainsworth& Eichberg,1991;Hesse,1999;Hesse&Main,2006), only parents who show signs of unresolved loss or trauma in their cognitive representation of past experiences might unwillingly stimulate their infants to become disorganized.Second,frightening paren-tal behavior showed no moderating effect on infant disorganized attachment.The distal risk factor ‘unresolved loss’must be transmitted to the child through some kind of parental behavior,but fright-ening behavior might be just one of the possible pathways(and thus one of the relevant proximal risk factors),with others being important as well. Although the speci?c parental behavior that constit-utes the active ingredient of the environmental component has yet to be revealed,the?ndings nicely illustrate gene–environment interaction;they sup-port the idea that genetic effects may be contingent upon gene–environment co-action(Rutter,2006).

Research into gene–environment interactions, complementary to the traditional environment-focused explanations of attachment,may be a fruitful approach to the understanding of the developmental pathways to(in-)secure and disorganized attach-ment.Extension of studies combining molecular genetics with careful measurement of environmental factors may provide the pathways to unraveling the interplay between genes and environment in the development of attachment relationships.

Towards a model for gene·environment interaction as differential susceptibility

The search for gene–environment interactions may be inspired by the hypothesis of‘differential sus-ceptibility’(Belsky,1997,2005;Boyce&Ellis,2005),or by the idea of‘genetic vulnerability’(Paris,2000; Rutter,2006).The former idea implies that some individuals,perhaps for genetic reasons,are more susceptible than others to positive as well as negat-ive rearing experiences that in?uence both positive and negative outcomes,whereas the latter model views some children as having a(genetic)suscept-ibility exclusively to a negative rearing experience resulting in a negative outcome(see Figure1).These two models are of course not mutually exclusive,but the differential susceptibility hypothesis predicts that not only will children who are more susceptible to environmental in?uences show more negative outcomes in unfavorable environments,but also that these children will show more positive outcomes when they grow up in favorable environments. Although much of the biological processes involved in these models are still unknown,changes in gene expression,known as environmental programming, may play a role(Meaney,2001;Meaney&Szyf, 2005).

Applied to the study of parental unresolved loss and infant disorganized attachment,the differential susceptibility hypothesis would predict higher levels of disorganized attachment for infants with the long DRD4variants when their mothers struggle with unresolved loss or other trauma,and lower levels of disorganization in children with the same genotype when their mothers do not show evidence of unre-solved loss.We tested this hypothesis empirically by contrasting children with the long DRD4variant of mothers without unresolved loss with the three other groups.The contrast was signi?cant;t(20.76, unequal variances)?4.70(p<.01).Children with long DRD4who had mothers without unresolved loss showed signi?cantly less attachment disorgan-ization than children with the short DRD4variant (regardless of their mothers’unresolved loss status)or children with long DRD4and mothers with unresolved loss.The contrast with the latter subgroup was also signi?cant t(4.37,unequal vari-ances)?4.53(p<.01),as was the contrast with the group of children with short DRD4(regardless of their mothers’unresolved loss status),t

(26.34,

Research Review:Genetic vulnerability or differential susceptibility in child development1167ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

unequal variances)?3.19(p<.01).These?ndings support the notion that the DRD47-repeat allele constitutes not only a genetic risk but may be a genetic basis for differential susceptibility over the whole range,with most positive outcomes for children with the long DRD4variant without the negative environmental factor.

Studies examining a negative outcome(disorgan-ized attachment)and negative rearing environ-ments(i.e.,unresolved loss,frightening behavior) may be considered only partially testing the differ-ential susceptibility hypothesis,because the absence of a negative outcome and a negative rearing environment is not necessary equal to the presence of a positive rearing environment and positive developmental outcomes.In this respect a crucial question that remains to be addressed is whether the genotype that appears to make a child vulnerable to unresolved loss vis-a`-vis disorganization also makes the organism more susceptible to certain positive rearing experiences(e.g.,high quality of sensitive parenting)when it comes to a positive outcome(e.g., higher levels of attachment security).

In that regard one other study on differential susceptibility and the role of DRD4polymorphisms may be of interest.The study addressed not attach-ment security or disorganization,but externalizing problem behaviors(Bakermans-Kranenburg&Van IJzendoorn,2006).The7-repeat DRD4allele has been associated with several forms of(mal-)adjust-ment in both childhood and adulthood,such as aggression and other externalizing problems in children(Benjamin et al.,2002;Schmidt,Fox, Rubin,Hu,&Hamer,2002).The role of harsh and insensitive parenting in the development of children’s aggressive and antisocial behaviors has also been amply documented(see Campbell,2002), but?ndings were not always convergent.The inter-play between DRD4and observed parental insensit-ivity for externalizing problem behaviors(Child Behavior Checklist(CBCL);Achenbach,1992)was tested in a sample of47twins,with a random selection of one child per twin pair included in the analyses(Bakermans-Kranenburg&Van IJzendo-orn,2006).A gene–environment interaction effect was found:Children with the7-repeat DRD4allele and insensitive mothers displayed signi?cantly more externalizing behaviors than children without the DRD47-repeat allele(irrespective of maternal sensitivity).Maternal insensitivity was associated with externalizing(oppositional,aggressive)behav-iors,but only in the presence of the DRD47-repeat polymorphism.The increase in externalizing behav-iors in children with the7-repeat allele exposed to insensitive care compared to children without these combined risks was six fold.The results were repli-cated in the second half of the twin pair sample,with similar results(Bakermans-Kranenburg&Van IJzendoorn,2006).It should be noted that DRD4 was not related to maternal sensitivity,excluding active gene–environment correlation as an alternat-ive explanation for the?ndings(see Kim-Cohen et al.,2006).

In order to test the‘favorable side’of the differen-tial susceptibility hypothesis,we now contrasted children with the long DRD4polymorphism and sensitive mothers with the other groups.Indeed, children with the DRD47-repeat polymorphism and sensitive mothers tended to show the lowest levels of externalizing problem behavior,t(43)?1.84,p?.07.A similar tendency was found for oppositional behavior,t(43)?1.82,p?.07;for aggressive behavior the contrast was signi?cant,t(9.99, unequal variances)?2.31,p?.04,see Figure2. Thus,children were differentially susceptible to both sensitive and insensitive parenting dependent on the presence of the7-repeat DRD4allele.

The next step in this line of research is testing whether children with the DRD47-repeat allele are differentially susceptible to experimental manipula-tion of the environment.The?ndings of the?rst experimental tests of(measured)gene by(observed) environment interaction in human development indicate that children are indeed differentially sus-ceptible to intervention effects depending on genetic differences.Children with the DRD47-repeat allele were more susceptible to experimentally induced changes in maternal discipline with respect to externalizing behavior outcomes(Bakermans-Kran-enburg,Van IJzendoorn,Pijlman,Mesman,&Juffer, in press)and they showed lower levels of daily cor-tisol(Bakermans-Kranenburg,Van IJzendoorn, Mesman,Alink,&Juffer,in press)after an inter-vention aimed at enhancing maternal sensitivity and positive discipline strategies(Van Zeijl et al.,2006).

These?rst studies testing the applicability of the differential susceptibility hypothesis for positive environmental effects on the more susceptible groups involve relatively small samples and certainly

1168Marian J.Bakermans-Kranenburg and Marinus H.van IJzendoorn

ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

need replication.Furthermore,future research may address the question whether such differential susceptibility to favorable and unfavorable environ-mental in?uences is also applicable to adult attachment and parenting behaviors.Parents with a DRD47-repeat allele may,for example,be more vulnerable to unresolved loss when they suffer the loss of a loved one or experience other traumatic events,or their parenting may be more affected by the presence or absence of,e.g.,a supportive partner relationship.

When these questions are addressed,it should be realized that genes steering the dopamine produc-tion are only one set of candidates to be investigated. As an example of a polymorphism that affects the serotonin rather than the dopamine system,Caspi et al.(2003)found that negative life events predicted depression among carriers of a short variant of the 5-HTTLPR serotonin transporter,whereas no direct association between the5-HTT gene and depression was observed.Exploratory investigations of adult attachment suggest that attachment representations of adult adoptees and the biological offspring of adoptive parents with a short5-HTT allele are less often secure when assessed with the AAI than their counterparts with the L-L variant(Riggins-Caspers, Cadoret,&Philibert,2003).

Both serotonin and dopamine systems are highly relevant for attachment-related studies of the inter-play of genes and environment,because various investigations have documented the signi?cance of these systems for psychological functioning(e.g., Bennett et al.,2002;Caspi et al.,2002,2003;Fox et al.,2005;Jaffee et al.,2005;Kaufman et al., 2004).In a similar vein,genes affecting the oxytocin and vasopressin neuropeptide systems may play a part in the establishment of social bonds and the regulation of emotions(Fries,Ziegler,Kurian,Jac-oris,&Pollak,2005).

We therefore propose a model(see Figure1)in which genes steering these systems are included as genotypes indicating genetic vulnerability or differ-ential susceptibility(with the DRD4receptor poly-morphism and the5-HTTLPR serotonin transporter as examples that have proven to be useful in past research).The selection of risk environments in order to examine gene–environment interaction (Mof?tt,2005;Mof?tt et al.,2005)may be optimal to uncover genetic vulnerabilities,but the additional inclusion of more positive environments may be necessary to examine the bright side that is implied in the differential susceptibility hypothesis. Acknowledgements

The authors were supported by research grants from the Netherlands Organization for Scienti?c Research (MJBK:VIDI grant no.452-04-306;MHvIJ:NWO SPINOZA prize).Correspondence to

Marian J.Bakermans-Kranenburg and Marinus H. van IJzendoorn,Centre for Child and Family Studies,Leiden University,P.O.Box9555,2300RB Leiden,The Netherlands;Tel+31715273434; Fax+31715273945;Email:bakermans@fsw. leidenuniv.nl;vanijzen@fsw.leidenuniv.nl References

Abrams,K.,Rifkin,A.,&Hesse,E.(2006).Examining the role of parental frightened/frightening subtypes in predicting disorganized attachment within a brief observation procedure.Development and Psycho-pathology,18,345–361.

Achenbach,T.M.(1992).Manual for the Child Behavior Checklist/2-3and1992pro?le.Burlington,VT:Uni-versity of Vermont,Department of Psychiatry. Ainsworth,M.D.S.,Blehar,M.C.,Waters,E.,&Wall,S. (1978).Patterns of attachment:A psychological study of the Strange Situation.Hillsdale,NJ:Erlbaum. Ainsworth,M.D.S.,&Eichberg,C.G.(1991).Effects on infant–mother attachment of mother’s unresolved loss of an attachment?gure or other traumatic experience.In C.M.Parkes,J.Stevenson-Hinde,& P.Marris(Eds.),Attachment across the life cycle(pp. 160–183).London:Routledge.

Bakermans-Kranenburg,M.J.,&Van IJzendoorn,M.H. (2004).No association of the dopamine d4receptor (drd4)and-521c/t promoter polymorphisms with infant attachment disorganization.Attachment and Human Development,6,211–218.

Bakermans-Kranenburg,M.J.,&Van IJzendoorn,M.H. (2006).Gene–environment interaction of the dopa-mine d4receptor(DRD4)and observed maternal insensitivity predicting externalizing behavior in pre-schoolers.Developmental Psychobiology,48,406–409.

Bakermans-Kranenburg,M.J.,Van IJzendoorn,M.H., Bokhorst,C.L.,&Schuengel,C.(2004).The impor-tance of shared environment in infant–father attach-ment:A behavioral genetic study of the Attachment Q-Sort.Journal of Family Psychology,18,545–549. Bakermans-Kranenburg,M.J.,Van IJzendoorn,M.H., Mesman,J.,Alink,L.R.A.,&Juffer, F.(in press). Effects of an attachment-based intervention on daily cortisol moderated by Drd4:A randomized control trial on1–3-year-olds screened for externalizing behavior.Development and Psychopathology. Bakermans-Kranenburg,M.J.,Van IJzendoorn,M.H., Pijlman,F.T.A.,Mesman,J.,&Juffer,F.(in press). Differential susceptibility to intervention:Dopamine D4Receptor Polymorphism(DRD4VNTR)moderates effects on toddlers’externalizing behavior in a ran-domized control trial.Developmental Psychology. Becker-Blease,K.A.,Deater-Deckard,K.,Eley,T.,et al. (2004).A genetic analysis of individual difference in dissociative behaviors in childhood and adolescence. Journal of Child Psychology and Psychiatry,45, 522–532.

Belsky,J.(1997).Theory testing,effect-size evaluation, and differential susceptibility to rearing in?uences:

Research Review:Genetic vulnerability or differential susceptibility in child development1169ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

The case of mothering and attachment.Child Develop-ment,68,598–600.

Belsky,J.(1999).Modern evolutionary theory and patterns of attachment.In J.Cassidy&P.R.Shaver (Eds.),Handbook of attachment:Theory,research, and clinical implications(pp.141–161).New York:The Guilford Press.

Belsky,J.(2005).Differential susceptibility to rearing in?uence:An evolutionary hypothesis and some evid-ence.In B.Ellis&D.Bjorklund(Eds.),Origins of the social mind:Evolutionary psychology and child development(pp.139–163).New York:The Guilford Press.

Belsky,J.,Bakermans-Kranenburg,M.J.&Van IJzendoorn,M.H.(in press).For better and for worse: Differential susceptibility to environmental in?u-ences.Current Directions in Psychology.

Benjamin,J.,Ebstein,R.P.,&Belmaker,R.H.(2002). Personality genetics,2002.Israel Journal of Psychi-atry and Related Sciences,39,271–279.

Bennett,A.J.,Lesch,K.P.,Heils,A.,Long,J.G.,Lorenz, J.G.,Shoaf,S.E.,et al.(2002).Early experience and serotonin transporter gene variation interact to in?u-ence primate CNS function.Molecular Psychiatry,7, 118–122.

Bokhorst,C.L.,Bakermans-Kranenburg,M.J.,Fearon, R.M.P.,van IJzendoorn,M.H.,Fonagy,P.,&Schuen-gel,C.(2003).The importance of shared environment in mother–infant attachment security:A behavioral genetic study.Child Development,74,1769–1782. Boyce,W.T.,&Ellis,B.(2005).Biological sensitivity to context:I.An evolutionary-developmental theory of the origins and functions of stress reactivity.Develop-ment and Psychopathology,17,271–301.

Bowlby,J.(1969/1982).Attachment and loss(vol.1). Attachment(2nd edn in1982).New York:Basic Books. Bretherton,I.(1990).Communication patterns,inter-nal working models,and the intergenerational trans-mission of attachment relationships.Infant Mental Health Journal,11,237–252.

Brooks,J.,&Lewis,M.(1974).Attachment behavior in thirteen-month-old,opposite-sex twins.Child Development,45,243–247.

Brussoni,M.J.,Jang,K.L.,Livesley,W.J.,&MacBeth, T.M.(2000).Genetic and environmental in?uences on adult attachment styles.Personal Relationships,7, 283–289.

Campbell,S.B.(2002).Behavior problems in preschool children.New York:The Guilford Press.

Carlson,V.,Cicchetti,D.,Barnett,D.,&Braunwald,K. (1989).Disorganized/disoriented attachment rela-tionships in maltreated infants.Developmental Psy-chology,25,525–531.

Caspers,K.,Yucuis,R.,Troutman, B.,Arndt,S.,& Langbehn,D.(in press).A sibling adoption study of adult attachment:The in?uence of shared environ-ment on attachment states of mind.Attachment and Human Development..

Caspi,A.,McClay,J.,Mof?tt,T.E.,Mill,J.,Martin,J., Craig,I.W.,et al.(2002).Role of genotype in the cycle of violence in maltreated children.Science,297,851–854.

Caspi,A.,Sugden,K.,Mof?tt,T.E.,Taylor,A.,Craig, I.W.,Harrington,H.,et al.(2003).In?uence of life

stress on depression:Moderation by a polymorphism in the5-HTT gene.Science,30,386–389.

Cassidy,J.,&Marvin,R.S.,with the MacArthur Working Group on Attachment(1992).Attachment organization in21/2to41/2year olds:Coding manual.Unpublished manuscript,University of Virginia.

Collins,W.A.,Maccoby,E.E.,Steinberg,L.,Hethering-ton,E.M.,&Bornstein,M.H.(2000).Contemporary research on parenting:The case of nature and nurture.American Psychologist,55,218–232. Constantino,J.N.,Chackes,L.M.,Wartner,U.G., Gross,M.,Brophy,S.L.,Vitale,J.,&Heath, A.C. (2006).Mental representations of attachment in identical female twins with and without conduct problems.Child Psychiatry and Human Development, 37,65–72.

Deater-Deckard,K.(2000).Parenting and child behav-ioral adjustment in early childhood:A quantitative genetic approach to studying family processes.Child Development,71,468–484.

Eaves,L.J.,Silberg,J.L.,&Erkanli,A.(2003).Resolv-ing multiple epigenetic pathways to adolescent depression.Journal of Child Psychology and Psychi-atry,44,1006–1014.

Ebstein,R.P.,Benjamin,J.,&Belmaker,R.H.(2002). Behavioral genetics,genomics,and personality.In R. Plomin,J.C.DeFries,I.W.Craig,&P.McGuf?n (Eds.),Behavioral genetics in the postgenomic era (pp.365–388).Washington,DC:American Psycho-logical Association.

Ebstein,R.P.,Levine,J.,Geller,V.,Auerbach,J., Gritsenko,I.,&Belmaker,R.H.(1998).Dopamine D4receptor and serotonin transporter promoter in the determination of neonatal temperament.Molecu-lar Psychiatry,3,238–246.

Fearon,R.M.,Van IJzendoorn,M.H.,Fonagy,P., Bakermans-Kranenburg,M.J.,Schuengel, C.,& Bokhorst, C.L.(2006).In search of shared and non-shared environmental factors in security of attachment:A behavior-genetic study of the associ-ation between sensitivity and attachment security. Developmental Psychology,42,1026–1040.

Finkel, D.,&Matheney, A.P.(2000).Genetic and environmental in?uences on a measure of infant attachment security.Twin Research,3,242–250. Finkel, D.,Wille, D.E.,&Matheny, A.P.(1998). Preliminary results from a twin study of infant–caregiver attachment.Behavior Genetics,28,1–8. Fox,N.A.,Nichols,K.E.,Henderson,H.A.,Rubin,K., Schmidt,L.,Hamer, D.,Ernst,M.,&Pine, D.S. (2005).Evidence for a gene–environment interaction in predicting behavioral inhibition in middle child-hood.Psychological Science,16,921–926.

Fraley,R.C.(2002).Attachment stability from infancy to adulthood:Meta-analysis and dynamic modeling of developmental mechanisms.Personality and Social Psychology Review,6,123–151.

Fries,A.B.W.,Ziegler,T.E.,Kurian,J.R.,Jacoris,S.,& Pollak,S.D.(2005).Early experience in humans is associated with changes in neuropeptides critical for regulating social behavior.Proceedings of the National Academy of Sciences of the United States of America, 102,17237–17240.

1170Marian J.Bakermans-Kranenburg and Marinus H.van IJzendoorn

ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

George, C.,Kaplan,N.,&Main,M.(1985).Adult Attachment Interview.Berkeley:University of Califor-nia.

Gervai,J.,Nemoda,Z.,Lakatos,K.,Ronai,Z.,Toth,I., Ney,K.,&Sasvari-Szekely,M.(2005).Transmission disequilibrium tests con?rm the link between DRD4 gene polymorphism and infant attachment.American Journal of Medical Genetics B.Neuropsychiatric Genetics,132B,126–130.

Goldberg,S.,Benoit,D.,Blokland,K.,&Madigan,S. (2003).Atypical maternal behavior,maternal repre-sentations and infant disorganized attachment. Development and Psychopathology,15,239–257. Goldberg,S.,Perrotta,M.,Minde,K.,&Corter, C. (1986).Maternal behavior and attachment in low-birth-weight twins and singletons.Child Develop-ment,57,34–46.

Goldsmith,H.H.,&Campos,J.J.(1990).The structure of temperamental fear and pleasure in infants:A psychometric perspective.Child Development,61, 1944–1964.

Grienenberger,J.,Kelly,K.,&Slade,A.(2005).Mater-nal re?ective functioning,mother–infant affective communication,and infant attachment:Exploring the link between mental states of observed caregiving behavior in the intergenerational transmission of attachment.Attachment and Human Development, 7,299–311.

Harari,D.,Van IJzendoorn,M.H.,&Bakermans-Kran-enburg,M.J.(2007).Attachment disorganization and dissociation.In E.Vermetten,M.J.Dorahy,& D. Spiegel(Eds.),Traumatic dissociation:Neurobiology and treatment(pp.31–54).Washington,DC:Ameri-can Psychiatric Press,Inc.

Harris,J.R.(1998).The nurture assumption:Why children turn out the way they do.Parents matter less than you think and peers matter more.New York:Free Press.

Hesse, E.(1999).The Adult Attachment Interview: Historical and current perspectives.In J.Cassidy& P.R.Shaver(Eds.),Handbook of attachment:Theory, research,and clinical implications(pp.395–433).New York:The Guilford Press.

Hesse,E.,&Main,M.(2006).Frightened,threatening, and dissociative parental behavior in low-risk sam-ples:Description,discussion,and interpretations. Development and Psychopathology,18,309–343. Hinde,R.A.,&Stevenson-Hinde,J.(Eds.).(1988). Relationships within families.Mutual in?uences. Oxford:Clarendon Press.

Ioannidis,I.(2003).Genetic associations:True or false? Trends in Molecular Medicine,9,135–138. Jacobvitz, D.,Leon,K.,&Hazen,N.(2006).Does expectant mothers’unresolved trauma predict frightening maternal behavior?Risk and protective factors.Development and Psychopathology,18,363–379.

Jaffee,S.R.,Caspi,A.,Mof?tt,T.E.,Dodge,K.A.,Rutter, M.,Taylor,A.,&Tully,L.(2005).Nature·Nurture: Genetic vulnerabilities interact with child maltreat-ment to promote behaviour problems.Development and Psychopathology,17,67–84.

Jang,K.L.,Paris,J.,Zweig-Frank,H.,et al.(1998). Twin study of dissociative experience.Journal of Nervous and Mental Disease,186,345–351.Kaufman,J.,Yang, B.Z.,Douglas-Palumberi,H., Houshyar,S.,Lipschitz,D.,Krystal,J.H.,&Gelern-ter,J.(2004).Social supports and serotonine trans-porter gene moderate depression in maltreated children.Proceedings of the National Academy of Sciences,101,17316–17421.

Kim-Cohen,J.,Caspi, A.,Taylor, A.,Williams, B., Newcombe,R.,Craig,I.W.,&Mof?tt,T.E.(2006). MAOA,maltreatment,and gene–environment inter-action predicting children’s mental health:New evidence and a meta-analysis.Molecular Psychiatry, 11,903–913.

Lakatos,K.,Nemoda,Z.,Toth,I.,Ronai,Z.,Ney,K., Sasvari-Szekely,M.,et al.(2002).Further evidence for the role of the dopamine D4receptor(DRD4)gene in attachment disorganization:Interaction of the exon III48-bp repeat and the521C/T promoter poly-morphisms.Molecular Psychiatry,7,27–31. Lakatos,K.,Toth,I.,Nemoda,Z.,Ney,K.,Sasvari-Szekely,M.,&Gervai,J.(2000).Dopamine D4 receptor(DRD4)gene polymorphism is associated with attachment disorganization in infants.Molecular Psychiatry,5,633–637.

Lyons-Ruth,K.,Bronfman,E.,&Parsons,E.(1999). Maternal frightened,frightening,or atypical behavior and disorganized infant attachment patterns.Mono-graphs of the Society for Research in Child Develop-ment,64,67–96.

Lyons-Ruth,K.,&Jacobvitz, D.(1999).Attachment disorganization:Unresolved loss,relational violence, and lapses in behavioral and attentional strategies.In J.Cassidy&P.R.Shaver(Eds.),Handbook of attach-ment:Theory,research,and clinical applications(pp. 520–554).New York:The Guilford Press.

Madigan,S.,Bakermans-Kranenburg,M.J.,Van IJzen-doorn,M.H.,Moran,G.,Pederson,D.R.,&Benoit,D. (2006a).Unresolved states of mind,anomalous parental behavior,and disorganized attachment:A review and meta-analysis of a transmission gap. Attachment and Human Development,8,89–111. Madigan,S.,Moran,G.,&Pederson, D.R.(2006b). Unresolved states of mind,disorganized attachment relationships,and disrupted mother–infant interac-tions of adolescent mothers and their infants. Developmental Psychology,42,293–304.

Main,M.(1990).Cross-cultural studies of attachment organization:Recent studies,changing methodo-logies,and the concept of conditional strategies. Human Development,33,48–61.

Main,M.,&Goldwyn,R.(1984–1998).Adult attachment scoring and classi?cation system.Berkeley:Unpub-lished scoring manual,Department of Psychology, University of California.

Main,M.,&Hesse, E.(1990).Parents’unresolved traumatic experiences are related to infant disorgan-ized attachment status:Is frightened and/or fright-ening parental behavior the linking mechanism?In M.T.Greenberg, D.Cicchetti,& E.M.Cummings (Eds.),Attachment in the preschool years:Theory, research,and intervention.The John D.and Catherine T.MacArthur Foundation series on mental health and development(pp.161–182).Chicago,IL:The Univer-sity of Chicago Press.

Main,M.,&Hesse,E.(1992/1995).Frightening,fright-ened,timid/deferential,dissociated,or disorganized

Research Review:Genetic vulnerability or differential susceptibility in child development1171ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

behavior on the part of the parent:Coding system. Unpublished manual,University of California at Berkeley..

Main,M.,Kaplan,N.,&Cassidy,J.(1985).Security in infancy,childhood,and adulthood:A move to the level of representation.Monographs of the Society for Research in Child Development,50,66–104.

Main,M.,&Solomon,J.(1990).Procedures for identi-fying infants as disorganized/disoriented during the Ainsworth Strange Situation.In M.T.Greenberg,D. Cicchetti,&E.M.Cummings(Eds.),Attachment in the preschool years:Theory,research,and intervention. The John D.and Catherine T.MacArthur Foundation series on mental health and development(pp. 121–160).Chicago,IL:The University of Chicago Press.

Meaney,M.J.(2001).Maternal care,gene expression, and the transmission of individual differences in stress reactivity across generations.Annual Review of Neuroscience,24,1161–1192.

Meaney,M.J.,&Szyf,M.(2005).Maternal care as a model for experience-dependent chromatin plasti-city?Trends in Neurosciences,28,456–463. Michel,G.F.&Moore, C.L.(1995).Developmental psychobiology.An interdisciplinary science.Cam-bridge,MA:MIT.

Minde,K.,Corter,C.,Goldberg,S.,&Jeffers,D.(1990). Maternal preference between premature twins up to age four.Journal of American Academy for Child and Adolescent Psychiatry,29,367–374.

Mof?tt,T.E.(2005).The new look of behavioral genetics in developmental psychopathology:Gene–environ-ment interplay in antisocial behaviors.Psychological Bulletin,131,533–554.

Mof?tt,T.E.,Caspi,A.,&Rutter,M.(2005).Strategy for investigating interactions between measured genes and measured environments.Archives of General Psychiatry,62,473–481.

Neale,M.C.,&Cardon,L.R.(1992).Methodology for genetic studies of twins and families.New York: Kluwer Academic/Plenum Publishers.

Neyer, F.J.(2002).The dyadic interdependence of attachment security and dependency:A conceptual replication across older twin pairs and younger couples.Journal of Social and Personal Relationships, 19,483–503.

O’Connor,T.G,&Croft,C.M.(2001).A twin study of attachment in preschool children.Child Development, 72,1501–1511.

Okuyama,Y.,Ishiguro,H.,Toru,M.,&Arinami,T. (1999).A genetic polymorphism in the promoter region of DRD4associated with expression and schizophrenia.Biochemical and Biophysical Research Communications,258,292–295.

Paris,J.(2000).Predispositions,personality traits,and posttraumatic stress disorder.Harvard Review of Psychiatry,8,175–183.

Pederson, D.R.,&Moran,G.(1995).A categorical description of infant–mother relationships in the home and its relation to Q-sort measures of infant–mother interaction.Monographs of the Society for Research in Child Development,60,111–132.

Pipp-Siegel,S.,Siegel,C.H.,&Dean,J.(1999).Neuro-logical aspects of the disorganized-disoriented attach-ment classi?cation system:Differentiating quality of

the attachment relationship from neurological impairment.Monographs of the Society for Research in Child Development,64,25–44.

Plomin,R.(1994).The Emanuel Miller Memorial Lecture1993:Genetic research and identi?cation of environmental in?uences.Journal of Child Psycho-logy and Psychiatry,35,817–834.

Plomin,R.,&Daniels,D.(1987).Why are children in the same family so different from one another? Behavioral and Brain Sciences,10,1–16.

Plomin,R.,DeFries,J.C.,McClearn,G.E.,&McGuf?n, P.(2001).Behavior genetics.New York:Worth Pub-lishers.

Purcell,S.(2001).Statistical methods in behavioral genetics.In R.Plomin,J C.DeFries,G.E.McLearn,& P.McGuf?n(Eds.),Behavioral genetics(4th327–371).New York:Worth Publishers.

Ricciuti, A.E.(1992).Child–mother attachment:A twin study.(University Micro?lms No.9324873). Dissertation Abstracts International,54,3364–3364.

Riggins-Caspers,K.,Cadoret,R.J.,&Philibert,R. (2003,April).Overcoming adversity:An adoption study of attachment.Paper presented at the SRCD Biennial Conference,Tampa,USA.

Robbins,T.W.,&Everitt,B.J.(1999).Motivation and reward.In M.J.Zigmond et al.(Eds.),Fundamental neuroscience(pp.1246–1260).San Diego:Academic Press.

Rose,R.J.(1995).Genes and human behavior.Annual Review of Psychology,46,625–654.

Rowe,D.C.(1994).The limits of family in?uence.Genes, experience,and behavior.New York:The Guilford Press.

Rutter,M.(2003).Commentary:Nature–nurture inter-play in emotional disorders.Journal of Child Psycho-logy and Psychiatry,44,934–944.

Rutter,M.(2006).Genes and behavior.Nature–nurture interplay explained.Oxford:Blackwell.

Rutter,M.,Mof?tt,T.E.,&Caspi, A.(2006).Gene–environment interplay and psychopathology:Multiple varieties but real effects.Journal of Child Psychology and Psychiatry,47,226–261.

Rutter,M.,Pickles,A.,Murray,R.,&Eaves,L.(2001). Testing hypotheses on speci?c environmental causal effects on behavior.Psychological Bulletin,127,291–324.

Sagi,A.,Van IJzendoorn,M.H.,Aviezer,O.,Donnell,F., Koren-Karie,N.,Joels,T,&Harel,Y.(1995).Attach-ments in a multiple-caregiver and multiple-infant environment:The case of the Israeli kibbutzim.In E.Waters, B.E.Vaughn,G.Posada,&K.Kondo-Ikemura(Eds.),Caregiving,cultural,and cognitive perspectives on secure-base behavior and working models:New growing points ofattachment theory and research(pp.71–91).Monographs of the Society for Research on Child Development,60.(Serial#244No. 2–3).

Schmidt,L.A.,Fox,N.A.,Rubin,K.H.,Hu,S.,&Hamer, D.H.(2002).Molecular genetics of shyness and aggression in preschoolers.Personality and Indi-vidual Differences,33,227–238.

Schuengel, C.,Bakermans-Kranenburg,M.J.,&Van IJzendoorn,M.H.(1999).Frightening maternal behavior linking unresolved loss and disorganized

1172Marian J.Bakermans-Kranenburg and Marinus H.van IJzendoorn

ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.

infant attachment.Journal of Consulting and Clinical Psychology,67,54–63.

Spangler,G.,Fremmer-Bombik,E.,&Grossmann,K. (1996).Social and individual determinants of infant attachment security and disorganization.Infant Mental Health Journal,17,127–139.

Spangler,G.,&Grossmann,K.(1999).Individual and physiological correlates of attachment disorganiza-tion in infancy.In J.Solomon&C.George(Eds.), Attachment disorganization(pp.95–124).New York: The Guilford Press.

Spangler G.,&Zimmermann,P.(2007,April).Genetic contribution to attachment and temperament.Paper presented at the biennial meeting of the Society for Research in Child Development,March29–April1, Boston,MA,USA.

Spinath,F.M.,&O’Connor,T.G.(2003).A behavioral genetic study of the overlap between personality and parenting.Journal of Personality,71,785–808. Suomi,S.J.(1999).Attachment in rhesus monkeys.In J.Cassidy&P.R.Shaver(Eds.),Handbook of attach-ment:Theory,research,and clinical applications(pp. 181–197).New York:The Guilford Press.

True,M.,Pisani,L.,&Oumar,F.(2001).Infant–mother attachment among the Dogon of Mali.Child Develop-ment,72,1451–1466.

Vandell,D.L.,Owen,M.T.,Wilson,K.S.,&Henderson, V.K.(1988).Social development in infant twins:Peer and mother–child relationships.Child Development, 59,168–177.

Van der Mark,I.L.,Van IJzendoorn,M.H.,&Baker-mans-Kranenburg,M.J.(2002).Development of empathy in girls during the second year of life: Associations with parenting,attachment,and tem-perament.Social Development,11,451–468.

Van Gestel,S.,&Van Broekhoven,C.(2003).Genetics of personality:Are we making progress?Molecular Psychiatry,8,840–852.

Van IJzendoorn,M.H.(1995).Adult attachment repre-sentations,parental responsiveness,and infant

attachment:A meta-analysis on the predictive validity of the Adult Attachment Interview.Psychological Bulletin,117,387–403.

Van IJzendoorn,M.H.,&Bakermans-Kranenburg,M.J. (2006).DRD47-repeat polymorphism moderates the association between maternal unresolved loss or trauma and infant disorganization.Attachment and Human Development,8,291–307.

Van IJzendoorn,M.H.,Moran,G.,Belsky,J.,Pederson, D.,Bakermans-Kranenburg,M.J.,&Kneppers,K. (2000).The similarity of siblings’attachments to their mothers.Child Development,71,1086–1098.

Van IJzendoorn,M.H.,Schuengel,C.,&Bakermans-Kranenburg,M.J.(1999).Disorganized attachment in early childhood:Meta-analysis of precursors,con-comitants,and sequelae.Development and Psycho-pathology,11,225–249.

Van Zeijl,J.,Mesman,J.,Van IJzendoorn,M.H., Bakermans-Kranenburg,M.J.,Juffer, F.,Stolk, M.N.,Koot,H.M.,&Alink,L.R.A.(2006).Attach-ment-based intervention for enhancing sensitive discipline in mothers of1-to3-year-old children at risk for externalizing behavior problems:A random-ized controlled trial.Journal of Consulting and Clinical Psychology,74,994–1005.

Vaughn, B.E.,&Waters, E.(1990).Attachment behavior at home and in the laboratory:Q-Sort observations and strange situation classi?cations of one-year-olds.Child Development,61,1965–1973. Waller,N.G.,&Ross,C.A.(1997).The prevalence and biometric structure of pathological dissociation in the general population:Taxometric and behavior genetic ?ndings.Journal of Abnormal Psychology,106, 499–510.

Waters,E.(1995).Appendix A:The Attachment Q-Set (Version3.0).Monographs of the Society for Research in Child Development,60,234–246.

Manuscript accepted30May2007

Research Review:Genetic vulnerability or differential susceptibility in child development1173ó2007The Authors

Journal compilationó2007Association for Child and Adolescent Mental Health.