Content uploaded by Christian Kandler
Author content
All content in this area was uploaded by Christian Kandler on Oct 07, 2017
Content may be subject to copyright.
1
Genetic Basis of Personality Traits
Christian Kandler1, Julia Richter2, & Alexandra Zapko-Willmes1,2
1 Medical School Berlin
2 Bielefeld University
Contribution to: V. Zeigler-Hill, T.K. Shackelford (eds.),
Encyclopedia of Personality and Individual Differences. Springer International Publishing.
http://dx.doi.org/10.1007/978-3-319-28099-8_1473-1
Acknowledgment
The authors received support from the Deutsche Forschungsgemeinschaft KA 4088/2-1.
2
Synonyms
Heritability; genes; individual differences; personality development; gene–environment
correlation; gene × environment interaction
Abstract
It is no longer controversial that genetic differences account for individual differences in all
human traits, such as personality traits: About 40% to 50% of variation are attributable to
genetic sources. Heritability estimates are even larger for more accurate measures of
personality characteristics. The combined consideration of molecular genetic study results
and findings from different genetically informative family studies yields that the genetic
basis of personality traits reflects many gene variants of small effects, which interact in
complex ways among each other and with environmental factors. Moreover, individual
differences in the genetic makeup drive individual differences in experiences and thus
influence the course of individual trait development within the opportunities provided by
the environment.
3
Introduction
In their meta-analysis of the genetic contribution to individual differences in 17,804
human traits based on studies of twins, Polderman and colleagues (2015) estimated the
average heritability across all traits at 49% (i.e., 49% of observable individual differences are
attributable to genetic differences among humans). They also reported variation in the size
of heritability estimates, depending on the specific traits studied. Whereas they found high
heritability levels for anthropometric traits (e.g., height and weight), they reported lower
levels of heritability for psychological traits, such as mental disorders or social
characteristics. For personality traits, Polderman et al.’s meta-analytic review and two other
meta-analyses (Johnson, Vernon, & Feiler, 2008; Vukasović & Bratko, 2015) yielded
comparable average heritability estimates: About 40% to 50% of individual differences in
personality traits are attributable to genetic differences.
It is undisputed that genetic differences do account for personality differences.
However, it has yet to be thoroughly understood how genetic factors contribute to
individual differences in the development of personality traits. In this review article, we aim
to introduce the reader into behavioral genetic research on the nature of personality trait
variation and development. For that purpose, we first provide a general definition of
personality traits and then explain a common strategy to estimate the heritability of a trait.
Following up on this, we give a short overview of behavioral genetic findings on the
heritability of personality traits. Subsequently, we describe different biological and
behavioral pathways through which genes can influence personality traits and thus
contribute to the individual development. Finally, we outline implications for research on the
genetic basis of personality differences and development and provide an outlook for future
trends.
4
Conceptions of Personality Traits
Personality traits are commonly defined as individual consistencies in feelings,
thoughts, and actions across situations, contexts, and occasions. Although different trait
models disagree about the number of traits (or trait dimensions) that is essential to
characterize personality differences, they generally agree in the idea that a finite number of
basic traits exist that show a genetic basis (often labelled as dispositional traits, core traits,
or basic tendencies). Among many other models, the Big Five trait taxonomy is the most
influential conceptual model in this regard because it captures personality traits along five
dimensions. These five personality dimensions are portrayed in many different languages,
cultures, and species; they display relative stability over time and across situations, can be
observed early in life, are biologically anchored, and predict several specific behaviors as
well as life outcomes (McAdams, 2015; McCrae, 2009). The Big Five trait dimensions have
been differently labelled, but the most common terms are: Neuroticism (versus emotional
stability), extraversion (versus introversion), openness to experience, agreeableness, and
conscientiousness.
The Big Five trait model integrates various other models of temperament and
personality traits within a common framework (John, Naumann, & Soto, 2008). Thus, each of
the five trait dimensions hierarchically subsumes or can be assigned to a broader set of more
specific traits, called facets (e.g., anxiety, impulsiveness, sociability, activity,
straightforwardness, or dutifulness). Moreover, some researchers proposed that there is
also a meaningful level of the trait hierarchy below facets, labelled as nuances (Mõttus,
Kandler, Bleidorn, Riemann, & McCrae, 2016). Others argue in favor of substance in more
abstract levels of a personality trait hierarchy based on systematic correlations between all
or specific Big Five personality traits (DeYoung, 2006). However, the Big Five trait dimensions
5
have often served as most economic units (i.e., the best compromise between bandwidth
and parsimony) of the analysis for the description and theory of personality trait differences
and development (Kandler, Zimmermann, & McAdams, 2014; McAdams, 2015; Specht et al.,
2014). Along this line, we primarily review genetically informative research on the genetic
basis of the Big Five trait dimensions as fundamental anchor for the description and genetic
etiology of trait differences.
Twin Studies and Heritability Estimates
Evidence for a genetic basis of individual differences in the Big Five personality traits
and related traits has primarily come from the classical design of twins reared together
(Johnson et al., 2008). Heritability estimates derived from this twin study design rely on the
comparison of the similarity of monozygotic (MZ) twins with the similarity of dizygotic (DZ)
twins. MZ twins are genetically identical siblings who share all of their genetic makeup
including additive and nonadditive genetic influences that can vary among humans. Additive
genetic factors (GA) reflect the mean effects of substituting one gene variant for another at
any given gene locus. They are correlated among family members as a function of their
genetic relatedness. Nonadditive genetic influences due to genetic dominance effects can
only be shared by siblings with a specific probability. Genetic dominance effects describe
cases in which the expression of one gene variant on a specific trait depends on the
presence of one or more other modifying gene variants. This can occur within the same gene
loci (GD), referred to as allelic gene interaction, or between gene loci due to multiple
interactions between different genes, called epistatic gene × gene interactions (GI). Since
genetically identical MZ twins share all of their genetic makeup, all potential genetic
influences on a specific trait contribute to the MZ twins’ trait similarity and only those
6
environmental influences that are not shared (ENS) and measurement error (ε) can
contribute to their dissimilarity in trait measures (see Figure 1).
[Figure 1 about here]
DZ twins are same-aged siblings who have – like other biological first-degree siblings
– a 50% probability of sharing additive genetic factors (GA) and a 25% probability of sharing
within-gene-loci dominance effects (GD), because they get one of two gene variants at any
given gene locus from each parent resulting in four possible combinations of gene variants
within gene loci for first-degree siblings. However, since the number of potential
combinations between gene variants among gene loci is infinite, DZ twins and other
biological relatives (except genetically identical MZ twins) have a probability of sharing
between-gene-loci gene × gene interaction effects (GI) close to zero. As a consequence, all
potential genetic influences on a specific trait only partially contribute to the DZ twins’ trait
similarity (see Figure 1).
Since MZ and DZ twins may also share or have experienced age-related and common
environmental influences (given they are or were reared together), environmental factors
can also account for their resemblance (ES). Assuming that shared environmental factors act
to increase DZ twins’ resemblance to the same degree as they contribute to the similarity of
MZ twins (equal-environment assumption) and assuming independence of genetic and
environmental factors (additivity assumption)
1
, it follows from the logic of the twin design
that the difference between the MZ and the DZ twin similarities informs about the
heritability of a trait: The larger both the twin correlations and the difference between the
MZ and the DZ twin correlations, the larger the genetic contribution to individual trait
1
Even though the additivity assumption is implausible, because genetic and environmental influences interact
and transact in multiple ways as we will discuss later in this chapter, this model assumption helps to estimate
the net contributions of genetic and environmental sources to individual differences in traits.
7
differences. As a rule of thumb, heritability can be broadly estimated as twice the difference
between the MZ and the DZ twin correlations: h² = 2 × (rMZ – rDZ). Nowadays, however,
heritability is estimated via structural equation modeling based on systems of equations with
known components (e.g., trait variance and twin correlations) and unknown parameters
(e.g., heritability and shared environmental influences) which have to be estimated based on
the known information. This allows a more precise estimate of heritability.
For simplicity, we restrict the description of heritability estimation approaches to the
most commonly used design of twins reared together. Of course, there are many other
alternative designs and approaches that allow heritability estimates, such as adoption
studies, studies on twins reared apart, or extended twin family studies. In any case, the logic
of comparing the resemblance of family members with different degrees of genetic
relatedness remains the same. However, the precision of heritability estimations increases
with the number of known parameters (i.e. correlations within different family dyads), which
increases with the complexity of a genetically informative study design (Hahn et al., 2012;
Keller, Medland, & Duncan, 2010).
Genetic Contributions to Personality Trait Differences
In their meta-analysis of quantitative genetic studies (e.g., twin, adoption, and family
studies) on personality traits, Johnson and colleagues (2008) reported an average twin
correlation across the Big Five and related traits of r = .45 for MZ twins and r = .21 for DZ
twins (or other first-degree siblings) reared together, whereas the correlations between
other biological relatives, such as half siblings, or biologically unrelated family members,
such as adoptees or stepsiblings, were even smaller. The increase of the correlations for
different kinship dyads with an increase of the kinships’ degree of genetic relatedness
indicated a substantial heritability for personality traits. Heritability estimates ranged
8
between h² = .37 (for agreeableness) and h² = .48 (for extraversion) indicating some but little
variation in the level of heritability of the Big Five traits and related personality traits.
MZ twin correlations of personality traits are often more than twice as large as the
correlations between DZ twin siblings and between other first-degree relatives (e.g.,
between other siblings or between parent and child), indicating that not only additive but
also nonadditive genetic factors contribute to MZ twins’ trait similarity and thus to individual
differences in personality traits. Strong evidence for a significant contribution of nonadditive
genetic factors (in particular epistatic gene × gene interaction effects) came from a recent
meta-analysis on the heritability of personality traits (Vukasović & Bratko, 2015). The
authors compared broad-sense heritability estimates (additive + nonadditive genetic
components) for personality traits based on studies including MZ twins (h² = .47) and
narrow-sense heritability estimates (additive genetic component only) based on other
genetically informative designs, such as extended family and adoption studies (h² = .22). This
considerable difference suggests a significant contribution of nonadditive genetic factors to
individual differences in personality traits, because those influences can only act to increase
the similarity of twins and in particular (i.e., in case of epistatic gene × gene interaction
effects; see Figure 1) the similarity of genetically identical MZ twins. Thus, twin studies play a
very important role for the estimation of the total amount of genetic contributions to trait
differences (i.e., broad-sense heritability including additive and nonadditive genetic
influences).
A more fine-grained and more robust way to estimate the additive genetic
component in the presence of nonadditive genetic contributions to individual differences in
personality traits is the analysis of data from extended twin family designs (i.e., twins and
their parents, partners, and/or children) or from multiple dyads of within-generational (e.g.,
9
twins, siblings, half-siblings, and cousins) and between-generational relatives (e.g., parent-
child and grandparent-grandchild). Those studies yielded broad-sense heritability estimates
for personality traits that were quite comparable to the results derived from the classical
twin design (e.g., Hahn et al., 2012).
Most studies and the mentioned meta-analyses of the heritability of personality traits
did not correct for error of measurement or nonrandom biases in measures of personality
traits. If not taken into account, random error of measurement (ε) would be confounded
with estimates of nonshared environmental factors (ENS) and act to increase family
members’ dissimilarity in trait measures (see Figure 1). Thus, heritability estimates are often
attenuated to the degree to which personality traits are unreliably measured. After
correction for random error variance (h²* = h²/[1 – ε²]), the heritability estimates of the Big
Five personality traits are typically larger, ranging between 50% and 60% (Loehlin, McCrae,
Costa, & John, 1998; see also Kandler & Papendick, in press).
Although the correction for random error of measurement may provide more reliable
estimates of heritability, studies relying on one method of measurement (or rater
perspective) cannot rule out systematic nonrandom method (or rater) biases distorting
heritability estimates. Most behavioral genetic studies on personality traits have relied on
self-reports, despite the problem that self-ratings are often distorted due to self-raters’
response styles (e.g., lenient, median, severe, or extreme response tendencies) or
unwillingness to provide accurate trait scores (e.g., impression management due to social
desirability). Given that impression management or response tendencies are also heritable
to some degree, genetic variance in personality self-reports may not only reflect true genetic
differences in the personality traits actually measured but also unwanted genetic
components in systematic self-rater biases. In fact, twin studies using self-reports as well as
10
ratings from well-informed peers found evidence for both a genetic component in accurate
personality trait scores modelled as latent trait variables based on self-peer convergence
(i.e., correlation between self- and peer reports) and a genetic component in self-report
residuals (Kandler, Riemann, Spinath, & Angleitner, 2010). However, heritability estimates
for latent trait scores increased with the self-peer convergence which increased with the
information accuracy of peer assessments (i.e., the peer-raters’ familiarity with the assessed
twin) resulting in a decline of the genetic component in self-report residuals (Kandler,
2012a). In other words: heritability estimates for personality traits increased and genetic
variance in method-specific components decreased with the accuracy of measurement (see
Figure 2). This indicates that genetic variance in self-reports on personality traits primarily
reflect substance in terms of true genetically based personality differences rather than
artifact in terms of heritable rater biases.
[Figure 2 about here]
Larger average heritability estimates have been found for more accurate measures of
the Big Five personality traits and related facets, such as openness to actions (Kandler et al.,
2010), as well as for more specific nuances, such as openness to different foods (Mõttus et
al., 2016), but not for potential higher-order personality trait dimensions, such as plasticity
(Riemann & Kandler, 2010). The latter indicates that not each trait level within a broad
hierarchy of personality traits may reflect genetically anchored trait differences, but rather
differences due to proximate environmental or distant cultural variation (cf. Turkheimer,
Pettersen, & Horn, 2014).
In sum, previous genetically informative studies – in particular twin studies –
provided strong and robust support for a genetic basis of individual differences in
personality traits. Despite some variation in the size of heritability estimates across the Big
11
Five personality traits – extraversion often tended to show the largest heritability and
agreeableness the smallest (see Kandler & Bleidorn, 2015, for an overview) – these
differences are not statistically important (Vukasović & Bratko, 2015). The average broad-
sense heritability of personality traits amounts to h² = .50, whereby heritability estimates are
typically larger for more accurate measures of the Big Five personality traits and related
facets (Kandler et al., 2010; Kandler, 2012a). A heritability around 50%, however, does not
mean that 50% of a person’s personality is determined by his or her genetic makeup.
Heritability basically represents the genetic contribution to individual differences in a
specific trait within a specific group or population at a specific point in time. Heritability
estimates have few implications for particular individuals, causation, or developmental
processes. For a deeper understanding of the genetic basis of trait variation, we need to
apprehend how genetic factors can unfold their impact driving the development of
individual differences in personality traits.
Gene Expressions in Personality Traits
Genes do not directly influence personality traits. Genetic differences unfold via
individual differences in the protein synthesis, neuroanatomic structures, and neural as well
as hormonal systems. More than two third of human genes unfold their effects within the
brain and the peripheral nervous system – the neural substrate of human psyche and
individual consistencies in feeling, thinking, and acting. This internal pathway can account for
the specific associations between gene variants and personality traits. Whereas molecular
genetic studies have been successfully identified specific genes for various disorders whose
symptoms include personality change (e.g., Alzheimer disease), it has emerged as
exceedingly difficult to identify single gene variants responsible for individual differences in
Big Five traits or related personality traits. There is a large discrepancy between the hardly
12
replicable effects of specific gene variants, which could only account for a trivial proportion
of personality differences (< 1%; based on genome-wide association studies with large
sample sizes; de Moor et al., 2012), and the aforementioned substantial and robust
heritability estimates for personality traits. This discrepancy is known as the missing
heritability problem.
There are some possible explanations for the missing heritability in molecular genetic
studies. First, a small number of rare gene variants in a population may show large effects.
Those gene variants are difficult to detect within randomly drawn samples of unrelated
individuals but are completely shared by MZ twins and partly shared by other biological
relatives accounting for their similarity. So far, however, no evidence has been found for this
explanation. A second and more plausible explanation is that heritability of complex
personality traits is due to many genes of small effects. In line with this hypothesis, studies
have found evidence that the additive combination of multiple single nucleotide
polymorphisms (SNP; i.e., the smallest gene units that vary among humans) can account for
more than 5% of individual differences in complex personality traits, such as 18% for
extraversion, 12% for neuroticism, 11% for openness, 10% for conscientiousness, and 9% for
agreeableness (e.g., Lo et al., 2016). Still, those SNP-based heritability estimates are much
smaller than the estimates of substantial heritability derived from genetically informative
twin and family designs.
A third explanation for the missing heritability in molecular genetic studies may be
nonadditive genetic influences and in particular the already introduced phenomenon of
epistatic gene × gene interactions. Some genes can regulate the expression of other genes
and may, thus, increase or decrease the effects of these genes. As a consequence, two
carriers of the same gene variant may differ in the expression of that gene because of
13
differences in this gene’s regulation. Likewise, carriers of different gene variants may show
similar gene expression by virtue of regulations that promote similar outcomes. Those
epistatic gene × gene interactions are not considered in genome-wide association studies,
which primarily focus on the main-effect association (i.e., linear correlation) between gene
variants and traits. Heritability estimates of about 50% for personality traits are often the
result of twin studies. Twin studies have also consistently shown that MZ twins are more
than twice as similar on personality traits as DZ twins, indicating the presence of nonadditive
genetic influences. Particularly epistatic gene × gene interactions can only contribute to the
trait similarity within monozygotic twin pairs and the dissimilarity among other biological
relatives, because those effects are only shared by genetically identical MZ individuals. As
already mentioned, a recent meta-analysis on the heritability of personality traits (Vukasović
& Bratko, 2015) found that studies including MZ twins yielded higher heritability estimates
(on average h² = .47) compared to other quantitative genetic designs without MZ twins (on
average h² = .22). The difference (.47 – .22 = .25) informs about the potential contribution of
epistatic gene × gene interactions. Consequently, those effects could account for about half
of the genetic component and about a fourth of individual differences in personality traits.
Epistatic gene × gene interactions are thus one of the most promising explanations for the
missing heritability.
The impact of genes may not only depend on the presence of other regulating genes
but also on individual’s environmental influences. For example, the genetic influences on
individual differences in two personality traits related to extraversion and neuroticism have
been found to be lower for 17 years old adolescents who experienced lower levels of
parental regard (Krueger, South, Johnson, & Iacono, 2008). Conversely, genetic differences
may affect individual differences in the sensitivity to environmental stressors. For instance,
14
the effect of the experience of negative life events on the tendency to depression during the
absence of positive experiences has been found to be larger for people with a stronger
genetic diathesis primarily mediated by neuroticism (Kandler & Ostendorf, 2016). Moreover,
environmental factors can switch on and off the gene expression without altering DNA
sequence, referred to as epigenetic influences. Hence, genetic factors can interact with
environmental factors in many different ways. Those gene × environment interaction effects
further complicate the detection of main effects of specific gene variants on personality trait
differences and provide a further explanation for the missing heritability in molecular
genetic studies.
In sum, there are many promising explanations for the tremendous discrepancy
between the estimates of small heritability based on molecular genetic studies and the
substantial heritability estimates for personality traits based on twin studies. These
explanations concern the multifarious ways how genes can express within the individual’s
physiological internal milieu (many genes of small effects and gene × gene interaction) and
how they can interact with external environmental factors (gene × environment interaction
and epigenetic effects) that complicate the search for specific gene variants responsible for
individual differences in personality traits.
Individual environments can affect the expression of genes, but genetic differences
may also affect the individual’s environmental milieu. Differences in people’s genetic
makeup promote differing behavior, which may be more or less associated with individual
differences in personality traits and increases or decreases the probability of exposure to
certain environments, referred to as gene–environment correlation. Their individual genetic
makeup may drive people to actively select or evoke situations and environments that are
consistent with their personality traits. For example, genetic factors that act to make people
15
more agreeable may attract them to other agreeable people and evoke more positive social
reactions and support from others. Individual genetic differences may also drive people to
change or avoid situations and environments that are inconsistent with their personality
traits. For example, genetic influences that act to make people more conscientious may drive
them to modify or avoid chaotic situations or contexts. These examples illustrate that each
seemingly environmental influence on individual differences in personality traits may
actually reflect a genetic influence to some degree and that gene–environment correlation
can account for the intriguing finding that virtually everything is heritable – not only complex
personality traits, but also features of individuals’ environments.
Genetic influences on individual differences in experienced environments, such as life
events or family environments, may be partly mediated by personality traits, which may
affect the way people experience events and the probability of exposure to certain
environments by evoking social reactions, selecting and seeking out settings, changing and
creating situations. Some studies provided support for this hypothesis. For example, genetic
variance in personality traits accounted for about two third of the genetic influences on
individual differences in perceived parental support (Kandler, Riemann, & Kämpfe, 2009).
Gene–environment correlation can also act as propulsive mechanism of the development of
personality differences, in case the genetically driven individual experiences and exposure to
certain environments, in turn, reinforce or even change individual personality traits.
Genetic Contributions to Personality Trait Development
In their theory of genotype→environment effects, Scarr and McCartney (1983)
described gene–environment correlations as developmental mechanisms. Whereas genes
are the driving force behind the individual development, the individual environments and
related experiences shape the development. This concept considered the fact that some
16
experiences are not at random, but induced by the individual with a specific profile of (partly
heritable) preferences and traits. That is, the individual’s genetic makeup (i.e., the genotype)
unfolds its impact via behavioral tendencies (that are reflected by or associated with
personality traits to some degree) and affects, which and how environments are actually
experienced and what (direction of) influences these environments or experiences have on
the individual development. The environments in turn can elaborate, reinforce, and
sometimes change the traits that have initially induced them. Trait reinforcement or
stabilization is associated with positive gene–environment correlation, whereas negative
gene–environment correlation may accompany trait change. Whether positive or negative,
Scarr and McCartney further differentiated an active, an evocative, and a passive type of
gene–environment correlation (or genotype→environment effects).
Active gene–environment correlation represents the case in which people are
attracted to or avoid environments that match or do not match their (partly heritable)
personality traits. It also refers to situations in which people create or manipulate their
environments to increase the individual person-environment fit. Whether they are selected,
avoided, created or manipulated, environments in turn reinforce individual personality traits
and thus stabilize individual differences in personality traits. For example, as we know,
extraversion is partly heritable and extraverts are typically attracted to parties and larger
social networks that may reinforce their extraversion, whereas introverts are attracted to
smaller social networks and rather avoid parties that in turn stabilizes their introversion.
There are also cases of negative gene–environment correlation within this active type. For
example, people with a genetic diathesis to anxiety and depressive symptoms may seek
psychotherapeutic help, which in turn may not only have the potential to reduce their
symptoms but also the underlying high level of neuroticism.
17
Evocative gene–environment correlation reflects situations, in which the individuals
receive responses from the social environment that are evoked by their (partly heritable)
traits. On the one hand, genetically driven low agreeableness may lead to more experiences
of social conflicts, which may consequently reinforce low agreeableness. On the other hand,
low agreeableness is often expressed in asocial and criminal behavior. This increases the
probability of social exclusion and punishment, which in turn may have the potential to
reduce asocial and criminal behavior accompanying an increase in agreeableness. Similarly,
high openness to experiences may increase the probability to gain more and diverse
experiences. Positive experiences may act to increase openness, whereas the experience of
negative events may have the effect of lowering the individual’s level of openness.
Passive gene–environment correlation refers to situations, in which the biologically
related parents provide a rearing environment that is correlated both with their genetic
makeup and the one of their offspring. This passive type can only arise in biological nuclear
families, because solely biological parents are genetically related with their children
(adoptive parents would not) and provide family environments for their offspring. For
example, individual differences in openness are partly genetically influenced, and parents
who are more open to new and unusual experiences may provide a more democratic,
flexible, and liberal parenting style that may strengthen similar openness levels in their
offspring.
According to the theory of genotype→environment effects, the similarity between
individuals in the exposure to environmental influences and in the development of their
genetically influenced personality characteristics depends on their genetic relatedness. In
other words, individual trait differences due to gene–environment correlations reflect
genetic differences and would thus be confounded with heritability estimates, if not
18
explicitly modelled in quantitative genetic studies. Negative gene–environment correlation
would act to decrease the genetic component in individual differences, whereas positive
gene–environment correlation would act to increase the genetic contribution to individual
trait variation. Thus, the theory of genotype→environment effects provides an alternative
explanation for the genetic variance in personality traits beyond additive and nonadditive
genetic influences and for higher heritability estimates derived from twin studies compared
to other quantitative genetic studies, such as extended family and adoption studies.
Because of the confoundedness of the variance due to gene–environment correlation
with genetic variance one could argue that it makes no sense to disentangle genetic from
environmental variance in quantitative genetic designs. However, this confoundedness helps
to understand several empirical phenomena that would appear as implausible in the first
instance. First, the heritability of Big Five personality traits tends to increase during the first
two decades of life (Kandler, 2012b; Kandler & Papendick, in press; cf. Briley & Tucker-Drob,
2014). Kandler and Papendick have shown that after correction for error of measurement
the average heritability across all Big Five traits increased from 50% to 60% between birth
and age 20 (see Figure 3). Second, the stability of genetic influences on individual differences
in personality traits increases from childhood to young adulthood (Briley & Tucker-Drob,
2014; Kandler, 2012b; Kandler & Papendick, in press). Given an equal 4-year interval of time,
Kandler and Papendick reported an increase of the continuity of genetic factors from r < .80
after birth to a nearly perfect stability (r > .90) after age 30. Both findings are in line with the
theory of genotype→environment effects, because the general degree to which experience
is influenced by individual genotypes should increase with the individuals’ growing self-
determination from childhood to young adulthood. Thus, the genetic basis of individual
19
differences can prompt individual differences in the experience of environments that in turn
shape the development of individual differences in personality traits.
[Figure 3 about here]
Other empirical phenomena, such as the general decline of heritability estimates for
personality traits across the entire life span (see Figure 3), cannot be explained by the theory
of genotype→environment effects. Therefore, alternative explanations have to be taken into
account. The increase of heritability from childhood to adulthood and the decrease during
the adult years can also be explained by another already mentioned phenomenon of the
interplay between genetic and environmental factors, called gene × environment
interaction. As already mentioned, most findings on the heritability of personality traits (in
particular longitudinal trends) stem from studies of twins reared together. When twins grow
up in their shared familial environment, they may differently respond to the same
environment provided by their parents on the basis of their differing genotypes. Since MZ
twins are genetically identical, they respond in a similar fashion to the shared environment
and thus may become more similar than DZ twins in their personality traits. Consequently,
interactions between genetic factors and shared environmental factors would act like
genetic influences and would be confounded with heritability estimates, if not explicitly
modeled in twin studies. Thus, with increasing development during the first two decades of
life, when twin siblings share many environmental contexts (e.g., parental home) and age-
related experiences (e.g., school entrance), MZ twins may become more similar than DZ
twins as a function of their shared genetic makeup.
With the individual’s increasing autonomy during the coming of age and self-
responsibility after leaving parental home, twins experience more and more individually
unique environments (e.g., peers, work, or own family), whereas those shared with their
20
twin sibling diminish. Each sibling individually responds to his or her unique environments as
MZ and DZ twin siblings do. Consequently, interactions between genetic factors and
environmental influences not shared by twins act individually and would have unique impact
on the development. In other words, those nonshared gene × environment interactions
would have the effect of making twins less similar, regardless of their genetic relatedness,
and would be confounded with estimates of individual environmental effects not shared by
twins, if not explicitly modeled in twin studies. Thus, with ongoing development there is a
shift from the importance of shared to nonshared interactions between genetic and
environmental factors. And this shift can explain larger heritability estimates in childhood
and the decline of the genetic contribution to individual differences in personality traits
based on twin studies across the lifespan.
Conclusions and Future Directions
Quantitative genetic studies (in particular twin studies) have found that about 50% of
individual differences in personality traits are genetically influenced with less variation
across Big Five personality traits and hierarchical levels of trait dimensions. Heritability
estimates increase with increasing accuracy of personality trait measurement. However, the
substantial heritability of personality traits does not mean that 50% of individuals’
personality traits are caused by genes. Genes unfold their impact through many different
pathways – from molecular biological mechanisms inside the organic cell, via
psychophysiological and behavioral pathways to individual environments outside the
organism. The influences of single genes or gene variants on individual differences in
complex personality traits are difficult to detect, because many genes of small effects are
involved and they interact in many complex ways among each other (gene × gene
interaction) and with the environment (gene × environment interaction). Or as Turkheimer
21
and colleagues (2014) put it (p. 12.22): “Both genes and environments matter, but neither
genetic nor environmental effects can be broken down into discrete and specifiable
mechanisms at a lower level of analysis.”
Genetic factors can drive experiences and the development of traits through
genetically driven preferences, choices, and behavioral patterns across time. People select
and create their niches, they are attracted to or avoid environments, evoke social reactions,
and thus construct their own experiences. In this way, the genetic basis influences the
course of trait development. The genetic unfolding, however, depends on the access to or
the limitation of opportunities afforded by the environment, and people are differently
sensitive to same environments depending upon their partly heritable traits. Individual
differences in personality traits result from the product of both the individual genetic
makeup and experiences, which are individually filtered and constructed from the
opportunities provided by the environment. And those opportunities can vary across
individuals’ proximal environments (e.g., family circumstances) as well as distal
environments (e.g., cultural background). Until now, we know little about the role of cultural
variation for the interplay between genetic and environmental influences on personality
traits, because genetically informative (or environmentally sensitive) cross-national studies
are scarce (Dar-Nimrod & Heine, 2011). A promising approach for the etiological research on
personality differences and development would consider the complex interplays between
genes as well as between genetic factors and different micro-level and macro-level
environmental sources as propulsive mechanisms of the development of individual
differences in personality across the life span.
As we have shown, the analysis of heritability estimates and the stability as well as
change of genetic factors over the life course provide interesting insights into the role
22
genetic factors can play for personality trait development. But in no case genetic factors
should be considered without or beyond environmental circumstances and influences. In this
vein, genetically informative longitudinal studies suggest that genetic variance in personality
traits may partly mirror gene–environment interplays. The development of individual
differences in personality traits is an intriguingly complex and multilevel affair that we can
only understand if we aim to gain more insight into how genetic and environmental sources
work together and interact on many different levels in many different ways.
23
References
Briley, D. A., & Tucker-Drob, E. M. (2014). Genetic and environmental continuity in
personality development: A meta-analysis. Psychological Bulletin, 140, 1303–1331.
doi:10.1037/a0037091
Dar-Nimrod, I., & Heine, S. J. (2011). Some Thoughts on Essence Placeholders,
Interactionism, and Heritability: Reply to Haslam (2011) and Turkheimer (2011).
Psychological Bulletin, 137, 829-833.
De Moor, M. H. M., Costa, P. T., Terracciano, A., Krueger, R. F., de Geus, E. J. C., … Boomsma,
D. I. (2012). Meta-analysis of genome-wide association studies for personality.
Molecular Psychiatry, 17, 337–349. doi:10.1038/mp.2010.128
DeYoung, C. G. (2006). Higher-order factors of the Big Five in a multi-informant sample.
Journal of Personality and Social Psychology, 91, 1138–1151. doi:10.1037/0022-
3514.91.6.1138
Hahn, E., Spinath, F. M., Siedler, T., Wagner, G. G., Schupp, J., & Kandler, C. (2012). The
complexity of personality: Advantages of a genetically sensitive multi-group design.
Behavior Genetics, 42, 221–233. doi:10.1007/s10519-011-9493-y
John, O. P., Naumann, L. P., Soto, C. J. (2008). Paradigm shift to the integrative big five trait
taxonomy. In O. P. John, R. W. Robins, & L. A. Pervin (Eds.), Handbook of personality:
Theory and research (3rd ed., pp. 114–158). New York, NY: Guilford.
Johnson, A. M., Vernon, P. A., & Feiler, A. R. (2008). Behavioral genetic studies of personality:
An introduction and review of the results of 50+ years of research. In G. J. Boyle, G.
Matthews, & D. H. Saklofske (Eds.), The Sage handbook of personality theory and
assessment. Vol. 1: Personality theories and models (pp. 145–173). London, England:
Sage.
24
Kandler, C. (2012a). Knowing your personality is knowing its nature: The role of information
accuracy of peer assessments for heritability estimates of temperamental and
personality traits. Personality and Individual Differences, 53, 387–392.
doi:10.1016/j.paid.2012.01.004
Kandler, C. (2012b). Nature and nurture in personality development: The case of neuroticism
and extraversion. Current Directions in Psychological Science, 21, 290–296.
doi:10.1177/0963721412452557
Kandler, C., & Bleidorn, W. (2015). Personality differences and development: Genetic and
environmental contributions. In James D. Wright (Ed.), International encyclopedia of
the social & behavioral sciences (2nd ed., Vol. 17, pp. 884–890). Oxford: Elsevier.
Kandler, C., & Ostendorf, F. (2016). Additive and synergetic contributions of neuroticism and
life events to depression and anxiety in women. European Journal of Personality, 30,
390–405. doi:10.1002/per.2065
Kandler, C., & Papendick, M. (in press). Behavior genetics and personality development: A
methodological and meta-analytic review. In J. Specht (ed.), Personality Development
across the Lifespan. Elsevier.
Kandler, C., Riemann, R., & Kämpfe, N. (2009). Genetic and environmental mediation
between measures of personality and family environment in twins reared together.
Behavior Genetics, 39, 24–35. doi:10.1007/s10519-008-9238-8
Kandler, C., Riemann, R., Spinath, F. M., & Angleitner, A. (2010). Sources of variance in
personality facets: A multiple-rater twin study of self–peer, peer–peer, and self–self
(dis-) agreement. Journal of Personality, 78, 1565–1594. doi:10.1111/j.1467-
6494.2010.00661.x
25
Kandler, C., Riemann, R., Spinath, F. M., Bleidorn, W., Thiel, W., & Angleitner, A. (2013). The
Bielefeld Longitudinal Study of Adult Twins (BiLSAT). Twin Research and Human
Genetics, 16, 167–172. doi:10.1017/thg.2012.67
Kandler, C., Zimmermann, J., & McAdams, D. P. (2014). Core and surface characteristics for
the description and theory of personality differences and development. European
Journal of Personality, 28, 231–243. doi:10.1002/per.1952
Keller, M. C., Medland, S. E., & Duncan, L. E. (2010). Are extended twin family designs worth
the trouble? A comparison of the bias, precision, and accuracy of parameters
estimated in four twin family models. Behavior Genetics, 40, 377–393.
doi:10.1007/s10519-009-9320-x
Krueger, R. F., South, S., Johnson, W., & Iacono, W. (2008). The heritability of personality is
not always 50%: Gene–environment interactions and correlations between
personality and parenting. Journal of Personality, 76, 1485–1521. doi:10.1111/j.1467-
6494.2008.00529.x
Lo, M.-T., Hinds, D. A., Tung, J. Y., Franz, C., Fan, C.-C., Wang, Y., … Chen, C.-H. (2017).
Genome-wide analyses for personality traits identify six genomic loci and show
correlations with psychiatric disorders. Nature Genetics, 49, 152-156.
Loehlin, J. C., McCrae, R. R., Costa, P. T., & John, O. P. (1998). Heritabilities of common and
measure-specific components of the Big Five personality factors. Journal of Research
in Personality, 32, 431–453. doi:10.1006/jrpe.1998.2225
McAdams, D. P. (2015). The art and science of personality development. New York, NY:
Guilford.
McCrae, R. R. (2009). The physics and chemistry of personality. Theory & Psychology, 19,
670–687. doi:10.1177/0959354309341928
26
Mõttus, R., Kandler, C., Bleidorn, W., Riemann, R., & McCrae, R. R. (2016). Personality traits
below facets: The consensual validity, longitudinal stability, heritability, and utility of
personality nuances. Journal of Personality and Social Psychology.
doi:10.1037/pspp0000100
Polderman, T. J. C., Benyamin, B., de Leeuw, C. A., Sullivan, P. F., van Bochoven, A., …,
Posthuma, D. (2015). Meta-analysis of the heritability of human traits based on fifty
years of twin studies. Nature Genetics, 47, 702–709. doi:10.1038/ng.3285
Riemann, R., & Kandler, C. (2010). Construct validation using multitrait-multimethod-twin
data: The case of a general factor of personality. European Journal of Personality, 24,
258–277. doi:10.1002/per.760
Scarr, S., & McCartney, K. (1983). How people make their own environments: A theory of
genotype → environment effects. Child Development, 54, 424–435.
Specht, J., Bleidorn, W., Denissen, J., Hennecke, M., Huttman, R., Kandler, C., . . .
Zimmermann, J. (2014). What Drives Adult Personality Development? A comparison
of theoretical perspectives and empirical evidence. European Journal of Personality,
28, 216–230. doi:10.1002/per.1966
Stößel, K., Kämpfe, N., & Riemann, R. (2006). The Jena Twin Registry and the Jena Twin Study
of Social Attitudes (JeTSSA). Twin Research and Human Genetics, 9, 783–786.
doi:10.1375/183242706779462615
Turkheimer, E., Pettersen, E., & Horn, E. E. (2014). A phenotypic null hypothesis for the
genetics of personality. Annual Review of Psychology, 65, 12.1-12.26.
doi:10.1146/annurev-psych-113011-143752
Vukasović, T., & Bratko, D. (2015). Heritability of personality: A meta-analysis of behavior
genetic studies. Psychological Bulletin, 141, 769–785. doi:10.1037/bul0000017
27
Figure Caption Legend
Figure 1. Twin model for monozygotic (MZ) and dizygotic (DZ) twins: GA = additive
genetic factors; GD = nonadditive genetic factors due to genetic dominance effects within
gene loci (allelic gene interaction); GI = nonadditive genetic factors due to genetic
dominance effects between gene loci (epistatic gene × gene interactions); ES =
environmental factors shared by twins; ENS = environmental factors not shared by twins; ε =
error of measurement.
Figure 2. Heritability estimates for latent trait scores based on self-peer rater
convergence corrected for random error of measurement and rater-specific components.
Estimates are derived from three studies focusing on different levels of the trait hierarchy
(Kandler et al., 2010; Mõttus et al., 2016; Riemann & Kandler, 2010) based on the combined
sample of two multiple-rater twin studies (Kandler et al., 2013; Stößel, Kämpfe, & Riemann,
2006).
Figure 3. Trends of heritability estimates (the proportion of individual differences due
to genetic differences) for Big Five personality traits across the life span (by courtesy of
Kandler & Papendick, in press): Estimates are based on the results of the 30 genetically
informative longitudinal, cross-sequential, or age-cohort studies and weighted by sample
size (darker points carried more weight in the analysis).
28
Figure 1.
Figure 2.
29
Figure 3.
