I recently watched the Netflix documentary “Three Identical Strangers.” Without too much of a spoiler alert, triplets who were separated at birth come across each other by chance and discovered that the story of their separation was no accident. The film raises important issues about what makes us the people that we are and ethical issue regarding research studies and their impact on peoples lives. In training and lectures on personality disorder, child development and parenting, research regarding the nature of personality often causes the most debate. I have been meaning to write about personality for some time, this film has inspired me to share some intriguing and compelling research studies on this subject. It is a big area so I will do this over several parts. In Part I I will introduce some basic concepts and explore the research base regarding the genetic heritability / environmental influence on development. In Part II I will explore the specific research on personality traits itself. And then in Part III, I plan to explore ethics, challenges and implications of the research.
Interest in personality has a long history but the systematic study of it is relatively recent. As practicing psychiatrists in the early 1950s, Chess and Thomas (1984) were struck by the amount of blame mothers received for their children’s misbehaviour. Contrary to psychoanalytic theories which dominated psychology at the time, Chess and Thomas observed that even in infancy children exhibited what they referred to as a primary reaction pattern. This led them to instigate a pioneering study of 141 babies temperaments that is commonly known as the New York Longitudinal Study (click here).
This study started in 1956 and followed these children over several decades. It combined direct observation and detailed interviews with parents about their children. Chess & Thomas’s diligent profiling found nine key characteristics that were highly stable as early as the age of two or three months old. They also occurred in a wide diversity of populations, regardless of socio-economic status, culture, learning disability or premature birth. The 9 characteristics were:
level and extent of motor activity
the rhythmicity, or degree of regularity, of functions such as eating, elimination and the cycle of sleeping
acceptance or withdraw response to a new object or person
the adaptability of behaviour in light of environment change
the threshold, or sensitivity, to stimuli
the intensity, or energy level, of responses
the child's general mood or "disposition", whether cheerful or given to crying, pleasant or cranky, friendly or unfriendly
the degree of the child's distractibility
the span of the child's attention and his persistence in an activity
When Chess and Thomas analysed these characteristics, they found that some traits clustered together into three general types of temperament: easy, difficult and slow to warm up. The "easy children" (40 per cent of the sample) were characterised by positiveness in mood, regularity in bodily functions, a low or moderated intensity of reaction, adaptability and positive approach to, rather than withdrawal from, new situations. As infants, they quickly established regular sleeping and feeding schedules, were cheerful and adapted quickly to new routines, new food and new people. As they grew older they learned the rules of new games quickly, participated readily in new activities and adapted easily to school.
The "difficult children" (10 per cent of the sample) were characterised as those with irregular bodily functions, usual intensity in reactions, tendency to withdraw in the face of new stimuli. They were slow to adapt to changes in the environment. As infants they were often irregular in feeding and sleeping, were slower to accept new foods, took a longer time to adjust to new routines or activities and tended to cry a great deal. Their crying and their laughter were rated as being characteristically loud. Frustrations usually seemed to send them into violent tantrums. These children were, of course, something of a trial to their parents and required a high degree of consistency and tolerance in their upbringing.
The children classified as "slow to warm up" (15 per cent of the sample) typically had relatively low activity levels, tended to withdraw on their first exposure to new stimuli, were slower to adapt, were somewhat negative in mood and responded to situations with a low intensity of reaction. The remaining children (35 per cent) had such mixture patterns that they did not allow for classification into one of the three groups.
For Chess & Thomas, child nurturance was not about the issue of ‘bad’ personality. Rather they posited that there was a bad fit between the child’s temperament and family environment. The ‘goodness to fit’ model has been particular important in adoption services, where child temperament can be very different to adoptive parents (to read more click here).
Chess and Thomas’s work has been highly influential and has inspired multiple studies that continue to substantiate and modify their work. For example, a recent study found that behavioural inhibition in infants temperament predicted personality at 26 years old (click here). As adults, they were also more reserved, had fewer romantic relationships and experienced lower social functioning with friends and family.
Despite the impact of Chess and Thomas’s work and subsequent studies, they are rarely cited or recognised in the UK. Here social workers, family practitioners and mental health workers tend to favour what might be termed “radical environmental” approaches. This was once the domain of Behaviourists who believed that personality was learnt through positive and negative consequences of interacting with the envrionment. You can train a mouse to navigate a maze using food but no end of cheese will teach it to play the piano. Even Behaviourists began to recognise that there were biological constraints to what could be learned.
However, radical environmental models are still popular but in new forms. Nowadays, these models attribute all difficulties a person can experience to abusive parenting, trauma or social-economic pressures. What these models fail to recognise is that these events impact on “something.” They impact on an individual who must assimilate them into their emotional and psychological world. We are born with in-built systems to process information and expereinces and there is variation in the systems each person inherits. This is the basic starting blocks of personality. It is the consistent and stable patterns of how these systems express themselves that we might refer to as temperament. So, if babies demonstrate consistent patterns of behaviour from birth, where do these temperaments come from?
Genes & Environments
The answer is that babies are not blank slates but are born with underlying temperaments which are largely inherited from their parents, which become shaped and influenced by the environments in which they grow. The study of these developmental pathways is called Behavioural Genetics. Behavioural Genetic studies never focus on one specific source of personality development. Instead it is a field that is interested the contribution made by three broad domains in human development. These are:
Heritability is the proportion of variation in individual traits, skills, or physical attributes that can be attributed to genetic factors.
Shared Environmental factors includes those environmental influences that are shared by family members and act to enhance familial similarity. As such, these factors that make family members more alike. This includes home, parenting style and socio-economic status which is shared by individuals in the family.
Nonshared Environmental factors is a measure of environmental influences that are unique to each individual such as peer groups or one-off life events like accidents or illnesses. These environmental influences operate to make members of the same family different from one another. Though sample error, the uncertainties of the research study itself, is often included in this category.
Whilst the study of heritability of personality traits is often discounted as reductionist, this a primary misunderstanding. Behavioural Genetic studies account for all three sources of influence on the development of personality as standard. They have done the leg work that radical envrionmentalists have ignored. Conversely, disciplines that focus solely on environmental impacts (such as parenting or trauma on personality) rarely ever consider any other factors that might influence the outcome of their studies.
Behavioural Geneticists assess contributions from these three domains through examining the similarity and differences in people with similar or dissimilar genetic profiles. The principle area of research has been twin studies. This is because identical twins share the same genes. Non-identical twins only share 50 per cent of their genes. And then you have virtual twins, or adoptees, that are not biologically related to their parents or siblings but instead share common environments with genetically unrelated others. So, for genetic influence to prove true, you would expect the more closely genetically related twins to be more similar in their traits and behaviours in comparison to the other groups.
The Basic Assessment Stat
The principle statistical model that has been used in these comparison studies is Correlation Covariance (Don’t glaze over! I will explain simply but you can skip to the last line in bold). Correlation Covariance is a way of understanding how two sets of numbers relate to each other. For example, how is height and weight related to each other? If taller people were always heavier than shorter people, this would be a positive correlation of 1.00. If all taller people were always lighter than every shorter person, it would be a negative correlation of -1.00. If there was no relationships between height and weight it would be a correlation of 0.00. The actual Correlation Covariance of height and weight 0.6. This is to say taller people tend to be heavier than shorter people but there is a degree of variation. The correlation between height and weight co-varies. So Correlation Covariance is a simple way of expressing how two sets of numbers relate to each other. It is then possible describe the Correlation Covariance as a percentage of the population as well. Behavioural Geneticist use more sophisticated statistical tools but they still tend to report the Correlation Covariance. So, closer to 1.00 suggests very strong relationships between two things whilst 0.00 means there is no relationship. That’s all you need to know to understand the data reported in this blog.
So when we make comparison of identical twins, non-identical twins or adoptees, we can assess how well a given personality trait in one twin correlates with the other. These studies can be arranged in several different ways. For example, do identical twins separated at birth share similar levels of anxiety despite growing up different environments? How does this compare to non-identical twins in the same situation? High degrees of similarity would suggest a strong genetic influence. Conversely, do adopted virtual twins share similar traits such as extroversion? This would suggest the
influence shared environments on their development.
Thousands of these studies have now been conducted across hugely diverse populations and socio-economic groups with large samples. These studies have revealed a surprising level of consistency and replicability. The weight of findings of these studies have been summed up by Turkhiemer (2003) (click here) which he has described as the Three Laws of Behaviour Genetics:
First Law: All human behavioural traits are heritable.
Second Law: The effect of being raised in the same family is smaller than the effect of genes.
Third Law: A substantial portion of the variation in complex human behavioural traits is not accounted for by the effects of genes or families.
I will explore the research base underpinning each of these laws separately.
Traits are Heritable
Let’s look at the first Law, that all behavioural traits are heritable. Early studies demonstrated that identical twins are much more alike in temperament than non-identical twins and this has been supported by nearly every study since (click here). Typically the correlation in traits in identical twins is almost double that found in non-identical twins. Conversely virtual twins raised in the same household are no more alike in temperament that two strangers randomly picked out of a crowd (Nettle, 2014). This suggests a high rate of heritability of genes on personlity traits.
A recent large scale meta-analysis (a study of studies) of twin correlations (click here) was conducted that examined and reported variance for 17,804 traits from 2,748 publications including 14,558,903 partly dependent twin pairs. This study amalgamated virtually all published twin studies of complex traits like temperament into one huge study. It found that estimates of heritability cluster strongly across all traits with a reported heritability of 49%. For a majority (69%) of traits, the best explanation of this high covariance was simply that twins shared similar genes.
Even more recently, advances in genome analysis has reveal greater insight into the genetic underpinnings of a wide range of traits that are determined by multiple genes that all exert a low level of influence. This has given rise to much more sensitive assessment. This has found similar findings for cognitive abilities (Benyamin et al., 2014; Davies et al., 2015; St Pourcain et al., 2014), psychopathology (Davis et al., 2013; Gaugler et al., 2014; Klei et al., 2012;Lubke et al., 2012; Lubke et al., 2014;McGue et al., 2013; Ripke et al., 2013; Wray et al., 2014), personality (Rietveld, Cesarini, et al., 2013; Verweij et al., 2012; Vinkhuyzen et al., 2012), and substance use/drug dependence (Palmer et al., 2015; Vrieze, McGue, Miller, Hicks, & Iacono, 2013). This has also been validated in cross cultural studies (Rolland 2002; Katigbak et al 1996).
These research findings are consistent and replicable across wide ranges of traits, populations and research methods (for a readable review of 10 consistent findings click here). As Plomin (2018) has suggested, the challenge now would be to find a human trait that is not linked to a genetic underpinning to some degree. But these studies reveal something else that should not be forgotten. No Behavioural Genetic study has ever identified a trait that is 100 per cent genetic. Genes make a significant contribution to human temprement but are not the sole determinant. Environments still matter. We have already seen that Behavioural Genetic studies include the influence of Shared and Nonshared Environments in studies. So what does the environment contribute to human development?
Turkhiemer's (2003) Second Law is that studies consistently reveal that the contribution made by shared environments, family & socio-economic position, is surprisingly low. Bouchard (2004) reviewed studies to provide a comparison of heritable verses environmental influence on a wide number of traits that are described in Table 1 (click here for the full article ). This is not from a representative sample but a survey of studies. Remember, numbers close 1.00 show a very strong correlation whilst numbers near 0 show weak correlations.
Table 1. Estimates of Broad Heritability and Shared Environmental Inﬂuence and Indications of Nonadditive Genetic Effects and Sex Differences in Heritability for Representative Psychological Traits (Adapted from Bouchard, 2004)
We can see a number of key points in this table:
Personality traits score in a remarkably similar range regardless of the model utilised (Big 5 or Big 3 Models). The vast majority of studies consistently suggest that 40 per cent of personality traits are inherited.
Shared Environment (family & socioeconomic position) consistently show a very weak correlation across nearly all measures-except religiosity. This is an average score and the impact of shared environment on any given individual will vary in the sample. However, a general conclusion is that the infuence of shared environment is much weaker for the vast majority of people than is commonly assumed and often negligible in many traits.
Reviewing studies examining age ranges consistently reveal Shared Environments make a larger contribution in the early life course but these decrease dramatically as age increases. Studies consistently show that genetic influences strengthen over time. In early childhood parents have complete control of the child’s environment. But once children start schooling, develop their own peer groups, gain choice & express preferences, interact with a wider range of adults etc, this influence rapidly decays. Hence Nonshared Environmental influences also become stronger.
Shared Environments weak and non-enduring effect in general populations has been found in nearly every single population study. For example, Rhee, & Waldman’s (2002) meta-analysis of anti-social behaviour found the variance with shared environment was 0.16. A large-scale study explored the heritability of ADHD in 1,938 families with twins and siblings aged 4 to 12 years (Levy, et al 1997). It found very high genetic heritability covariance of 0.75 to 0.91 across familial relationships (twin, sibling, and twin-sibling) and across different definitions of ADHD. No evidence was found for shared environment contributing to the existence of the disorder. Likewise, conjoined twins who share the same environments (family and public) at all times are no more alike in personality than non-conjoined twins (Harris, 2006).
As Turkhiem (2003) summarised in response to the low influence of shared environments:
“It is no longer possible to interpret correlations among biologically related family members as prima facie evidence of sociocultural causal mechanisms. If the children of depressed mothers grow up to be depressed themselves, it does not necessarily demonstrate that being raised by a depressed mother is itself depressing. The children might have grown up equally depressed if they had been adopted and raised by different mothers, under the influence of their biological mother’s genes. For every behavior geneticist who continues to report moderate heritabilities as though they were news, there is an environmentalist who reports causally ambiguous correlations between genetically related parents and children.”
Genetic factors are rarely included in parenting models of development which means that they may dramatically over-report the influence of parenting. Furthermore, they also fail to recognise that the parenting style itself will also be determined by the parents own personality, which is partly genetically influenced. This might be considered the “nature of nurture” and studies are beginning to explore this area more deeply (click here for a detailed analysis).
Interestingly, the small correlations in shared environmental influence have been paralleled in primate studies. For example, studies have shown chimpanzee personality (whom we share 98.5 per cent of our DNA with) is composed of Dominance-related factors and five human-like factors. Genetic, Shared zoo, and Nonshared Environmental variance of these factors has also been estimated. Traits such as Dominance show significant heritability whilst the effects of “shared zoo” are almost negligible (Weis et al 2000).
Non-Shared Environments: The Biggest Puzzle?
In broad terms, genetic and shared environments have been mapped with high degrees of consistency. The mysterious element is the infleunce of non-shared environmental factors, which account for a substantial portion of the variation in complex human behavioural traits. This brings us to Turkhiem's (2003) Third Law- A substantial portion of the variation in complex human behavioural traits is not accounted for by the effects of genes or families. Genetic factors explain why siblings are similar, but they cannot account for everything. No trait is 100 per cent heritable. So, with Shared Environment factors showing weak correlations, the Nonshared Environment must be important in making us different.
Nonshared Environmental factors describe the effects of development that are unique to each and every person. These Nonshared Environments are not about the “events” that happen to us, but rather they measure the “effect” these events have on us. This includes our perception of these events rather than the innate magnitude of them. Therefore, any factor can be considered as a Nonshared Environmental effect, so long as it can be assessed separately for each child (child-specific).
As we grow up, we separate from our families so most of what happens to us is non-shared, identical twins included. Judith Rich Harris provides a compelling explanation of the drivers of non-shared environmental effects are largely determined by peer socialisation processes that simply over-power the best parenting intentions in the world as we leave home in degrees. To read interesting summarys of her two books click here and here.
Divorce, school, peer group, heartbreak, physical illness, accidents, luck and misfortune occur in all our lives and have their own personal sting. This has given rise to the “gloomy prospect” that the sheer randomness of Nonshared Environments may be impossible untangle. Furthermore, how might we understand the relationship between genes and these environments? Do these environmental effects shape our temperaments or is it simply that our genetic temperaments lead us into particular environments? As gene expression changes with age, do we have to factor ‘timing’ into the chicken and egg equations as well?
Some success has been found in the identification of different experiences of parenting amongst siblings within families. As I have suggested before, parenting is often determined by where the resources of a parent meet the temperament of the child. As children differ in families, the parental response cannot be the same for every child. Plomin et al (2003) study on adjustment, personality and cognition has found non-shared environment effects within families. There were correlations of 0.01 for family constellation, 0.02 for differential parental behaviour, 0.02 for differential sibling interaction and 0.05 for differential peer or teacher interaction. These correlations are low but can add up quickly -and they may also be bidirectional.
Burt et al (2005) (click here) studied the origins and causes of childhood externalizing symptoms (poor impulse control) and parent–child conflict from the ongoing Minnesota Twin Family Study. A total of 1,506 same-sex twins were assessed at ages 11 and 14 and their parents too. The results revealed that conflict and externalizing behviour in young people was mostly genetic. However, there was also a relationship between conflict and the externalised disorders. The results suggested that parent–child conflict partially results from parental responses to their child's heritable externalizing behaviour, which in turn amplified the young person’s symptoms further. These results suggest a “downward spiral” of interplay between parent–child conflict. Poor impulse control impacted on the parent, who responded with harsher parenting. This in turn acerbated poor impulse control. Here we see a dominant genetic effect being amplified by a much lower but significant environmental effect. And it brings us closer to understanding how genetic and environment interact.
In the first part in of this series I hoped to introduce you to the essential principles and concepts of Behavioural Genetics. And, by default, challenge some assumptions about the nature of this research for those new to the topic. Firstly, it is evident that infants are born with temperaments. However, the study of their development is not restricted to examining genetics but all sources of influence including shared and non-shared environments. Secondly, the research findings in Behavioural Genetics have proved highly consistent and replicable in ways that is simply not observed in any other field of human study. Whilst these can be summarised by Turkhiemer’s Three Laws, this summation should not betray the complexity and nuance that resides within them or that emerging research will not challenge them. In summary, we bequeath our children much in terms of their genetics and their socio-economic standing. But we must also recognise that they are also highly sensitive to the world that they will create around themselves. Perhaps that is the real take home message for parents. In the next instalment, I will focus on the main personality traits themselves, what they are and what shapes their expression in normative and personality disordered individuals
Benyamin, B., Pourcain, B., Davis, O. et al. (2014) Childhood intelligence is heritable, highly polygenic and associated with FNBP1L. Mol Psychiatry 19, pp 253–258 .
Burt, A.S. et al (2005) How are parent–child conflict and childhood externalizing symptoms related over time? Results from a genetically informative cross-lagged study Volume 17, Issue 1, pp. 145-165
Davies G, Armstrong N, Bis JC, Bressler J, Chouraki V, Giddaluru S, Deary IJ. (2015) Genetic contributions to variation in general cognitive function: A meta-analysis of genome-wide association studies in the CHARGE consortium (N=53 949) Molecular Psychiatry.20:pp 183–192.
Davis LK, et al (2013) Partitioning the heritability of Tourette syndrome and obsessive compulsive disorder reveals differences in genetic architecture. PLoS Genetics.;9:e1003864.
Levy, F. et al. (1997) Attention-Deficit Hyperactivity Disorder: A Category or a Continuum? Genetic Analysis of a Large-Scale Twin Study. Journal of the American Academy of Child & Adolescent Psychiatry, Volume 36, Issue 6, pp 737-744
Gaugler T, et al (2014). Most genetic risk for autism resides with common variation. Nature Genetics, 46: pp 881–885.
Harris, J. R. (2006) No Two Alike: Human Nature and Human Individuality. W.W. Norton.
Katigbak, M. S. et al (1996). Cross-cultural generalizability of personality dimensions: Relating indigenous and imported dimensions in two cultures. Journal of Personality and Social Psychology, 70(1), pp 99–114.
Klei L, et al (2012) Common genetic variants, acting additively, are a major source of risk for autism. Molecular Autism.; 3:9.
Little, B.R. (2014) Me, Myself, and Us: The Science of Personality and the Art of Well-Being. PublicAffairs
Lubke, G.H. et al (2012) Estimating the genetic variance of major depressive disorder due to all single nucleotide polymorphisms. Biological Psychiatry. 72:pp 707–709.
Lubke, G.H. et al (2014) Genome-wide analyses of borderline personality features. Molecular Psychiatry.;19:923–929. doi: 10.1038/mp.2013.109.
McGue M, et al (1993) Personality stability and change in early adulthood: A behavioral genetic analysis. Developmental Psychology. 29:pp 96–109.
Palmer R.H.C. et al (2015) Examining the role of common genetic variants on alcohol, tobacco, cannabis and illicit drug dependence: Genetics of vulnerability to drug dependence. Addiction. 110:pp 530–537.
Plomin, R, et al (2003) Why are children in the same family so different? Nonshared environment a decade later, Can J Psychiat, 2001, vol. 46: pp. 225-33.
Plomin, R. (2018) Blueprint: How DNA Makes Us Who We Are. MIT Press
Rietveld CA, et al (2013) Molecular genetics and subjective well-being. Proceedings of the National Academy of Sciences (USA), 110:pp 9692–9697.
Ripke S, et al (2013) Genome-wide association analysis identifies 13 new risk loci for schizophrenia. Nature Genetics.45:pp 1150–1159.
Rhee, S. H., & Waldman, I. D. (2002). Genetic and environmental influences on antisocial behavior: A meta-analysis of twin and adoption studies. Psychological Bulletin, 128(3), 490–529
Rolland, J.P. (2002)The Five-Factor Model of Personality Across Cultures pp 7-28 | The Cross-Cultural Generalizability of the Five-Factor Model of Personality.
St Pourcain B, et al (2014) Common variation near ROBO2 is associated with expressive vocabulary in infancy. Nature Communications. 5:pp 4831.
Verweij K.J.H. et al (2012) Maintenance of genetic variation in human personality: Testing evolutionary models by estimating heritability due to common causal variants and investigating the effect of distant inbreeding. Evolution, 66:pp 3238–3251.
Vinkhuyzen A.A.E. et al (2012) Common SNPs explain some of the variation in the personality dimensions of neuroticism and extraversion. Translational Psychiatry. 2(4): e102
Vrieze S.I. et al (2013) Three mutually informative ways to understand the genetic relationships among behavioral disinhibition, alcohol use, drug use, nicotine use/dependence, and their co-occurrence: Twin biometry, GCTA, and genome-wide scoring. Behavior Genetics. 43:pp 97–107.
Weiss, A. et al (2000) The Heritability of Personality Factors in Chimpanzees (Pan troglodytes) Behavior Genetics. 30, pp 213–221.
Wray N.R. et al (2014) Research review: Polygenic methods and their application to psychiatric traits. Journal of Child Psychology and Psychiatry. 55:pp 1068–1087.