A genetic study of nearly 80,000 young children found that activity, shyness, sociability and emotionality share some genetic roots with ADHD, autism and personality traits measured much later in life. The overlaps cannot predict an individual child’s future, but reveal biological continuity across development.
At any toddler gathering, the differences are hard to miss. One child races across the room and rarely pauses. Another stays close to a parent’s leg. A third child seeks out other children immediately while another watches from a distance.
For parents, this can feel deeply personal. Is my child just lively? Too shy? More emotional than others? Am I doing something wrong?
A new study published in Nature Human Behaviour suggests that this may be the wrong way to frame the question.
Angelica Ronald, Professor of Psychology and Genetics at the University of Surrey in the United Kingdom, and colleagues analysed genetic and behavioural data from almost 80,000 infants and toddlers. They identified 10 regions of DNA associated with activity, shyness, sociability and emotionality.
The broader genetic patterns linked to these traits also overlapped, in separate datasets, with attention-deficit/hyperactivity disorder (ADHD), autism and adult neuroticism – the tendency to experience negative emotions more readily.
“We are not saying that a shy toddler cannot become an extroverted adult. The genetic overlap is only partial, and children’s environments and experiences also shape how they develop. What is interesting is that some of the same genetic variation can still be seen across such different stages of life,” explains Angelica Ronald.
Why genetic research has largely overlooked toddler temperament
In developmental psychology, early differences in activity, emotionality, shyness and sociability are called temperament. They can influence how children explore, learn and interact with others – and the responses they receive from parents, childcare workers and other children. Temperament is therefore relevant far beyond the home, including in childcare, education, clinical practice and public health.
Twin studies have shown that early behaviour is partly heritable, but they cannot identify the specific DNA differences involved.
“Twin studies tell us that genes matter, but they do not discover which genes or how they work. To answer that, you need a gene discovery study,” says Angelica Ronald.
Genome-wide association studies have identified genetic variants linked to many common traits and diseases. But most large biobanks contain adults rather than infants and toddlers, making early-childhood genetics much harder to study.
“There are very large biobanks for older people, but there is not the same advocacy for infants and toddlers. They cannot speak up for themselves. And there are not large biobanks for this age group, so people like me cannot just access one dataset and do the study easily,” says Ronald.
This leaves a gap during a formative period when children are learning to move, express emotions and form relationships. The new study asks whether common genetic variants – small DNA differences that are widespread in the population – help to explain ordinary differences in early behaviour.
“This would complement existing research that focuses on extreme, severe early-onset conditions,” says Angelica Ronald.
The study therefore adds genetics to a picture of early development in which both rare conditions and children’s surroundings are already known to matter.
Finding tiny genetic effects required nearly 80,000 toddlers
Ronald and colleagues conducted a genome-wide association study. Instead of starting with a suspected gene, they compared common DNA variants across the children to determine whether any were consistently linked to temperament.
“Each genetic variant is likely to make only a tiny difference to complex behaviour. This means that you need very large samples but also good measures of behaviour. We searched internationally for cohorts that had measured traits differing widely between children, could be assessed consistently and were already known to be partly heritable,” explains Angelica Ronald.
Parents or caregivers had rated almost 80,000 infants and toddlers during their second and third years of life on emotionality, activity, shyness and sociability. Such reports capture behaviour across far more situations than researchers can observe directly but can also be influenced by expectations and memory.
“Parents see their child across a range of situations – at three in the morning, dropping them off at daycare, with siblings, morning, noon and night. There can be biases, but there is always bias and error in whatever data you have,” notes Angelica Ronald.
How researchers separated stable temperament from measurement noise
Direct observation, video recordings or activity sensors could have provided more objective measurements but were not feasible at this scale. Because the cohorts had used different questionnaires, the researchers asked another group of families to complete them and tested how closely the measures agreed.
“Sometimes cohorts have not used exactly the same measure. Then you have to ask whether you leave them out or check whether the measure is about the same,” explains Angelica Ronald.
When possible, the researchers combined ratings from more than one age. A child rated as highly active at both age two and age three years, for example, was more likely to reflect a stable difference than a temporary phase.
“We wanted to know whether the associations might inadvertently be picking up socioeconomic status or other differences between families. In the twin cohorts, we therefore compared siblings within the same family as well as children from different families. The signal did not change substantially, which suggests that our results did include direct genetic influences and were not being driven by socioeconomic factors,” says Angelica Ronald.
The researchers combined the small effects of many variants into genetic scores. Scores built from one set of cohorts were linked to the same traits in cohorts left out of the calculation, although they explained only a very small share of the differences between children. This provided another check that the associations were not confined to the data used to identify them.
The team then compared the toddler patterns with diagnoses and personality traits measured in independent studies.
“You would not find genetic overlap if the signal in the toddler study was just noise. For example, ADHD diagnosis later in life and parent-rated toddler activity have very different sources of error, so overlap between them is another good sign,” adds Angelica Ronald.
Shyness showed the strongest genetic signal
In the analyses combining ratings from both ages, the common variants captured by the study accounted for about 15% of the variation in shyness, 10% in activity, 7% in emotionality and 3% in sociability.
These figures do not represent the traits’ total genetic contribution: they include only the common variants measured by this method. The same ranking appeared in the largest individual cohort, and the researchers identified 10 regions of the genome associated with at least one trait.
“The traits were connected, but not all in the same way. Shyness and sociability tended to point in opposite directions, while shyness and emotionality shared some genetic variation. Early temperament therefore appears to involve patterns of activity, emotion and social behaviour that are partly genetically related,” says Angelica Ronald.
Comparisons with later-life datasets showed genetic overlap between higher toddler activity and ADHD, between higher emotionality and adult neuroticism and between greater shyness and lower adult extraversion. This points to shared genetic patterns; it does not mean that the early traits and later outcomes are the same.
“Earlier studies have linked activity among young children with later ADHD and toddler temperament with later personality. What is different here is that we see the connection through shared genetic patterns, which can teach us something about the underlying biology,” notes Angelica Ronald.
These are similarities across large groups and not forecasts for individual children. An active toddler does not necessarily have or later develop ADHD, and a shy toddler can become a socially confident adult.
“You are asking a lot of the data. Many of the toddler data come from when children are about 18 months old, and the adult personality data are from much later in life. There are many years in between. So to still see some genetic overlap was really fascinating,” adds Angelica Ronald.
The link with autism appeared in two analyses. Variants associated with lower sociability and higher emotionality also tended to be associated with autism in an independent dataset. A separate analysis found that genes highlighted by the temperament results included more autism-linked genes than expected. Because this second analysis involved only a few genes, however, it should be treated as another clue rather than independent confirmation.
The genetic signals are starting points – not explanations
Two signals may help researchers investigate the underlying biology. The activity signal coincided with another genetic signal linked to how strongly RHEBL1 is expressed in the adult cortex; the same region has previously been associated with educational attainment and cognitive performance. The emotionality signal showed a similar pattern involving MR1, a gene best known for its role in the immune system.
The researchers did not find single genes “for” shyness, activity or sociability. Temperament is highly polygenic: many variants each nudge behaviour by a tiny amount. Combining these effects into a genetic score can help researchers to study development, but the scores still explain far too little to predict an individual child’s outcomes.
Adding cohorts from Japan and South Africa produced broadly similar patterns, but only one of the identified regions still met the study’s strict statistical threshold. These cohorts were much smaller, and most participants still had European ancestry. The findings therefore cannot be assumed to apply equally across populations. The authors emphasise that larger and more diverse samples are needed.
“This study gives other researchers a starting point. We have made the genetic results publicly available, so other researchers can now ask new questions about this early stage of life that the genetics field has previously overlooked,” adds Angelica Ronald.
Genetics may help reveal which environments matter
For now, the findings can support research but cannot be used to assess individual children. One unexpected use is to study the environment more rigorously: because children differ genetically from the start, part of an apparent link between parenting, childcare, stress and later outcomes may reflect inherited predispositions.
“It sounds counterintuitive, but genetics may help us to understand the early environment more accurately. Once you account for genetic predisposition, you can ask whether parenting, childcare or other experiences are still linked to young children’s behaviour. That gives you a cleaner test of the environment,” says Angelica Ronald.
The findings may also help to explain why children carrying the same rare deletion or mutation can have very different symptoms and support needs. One possibility is that the rare change acts against a different background of thousands of common genetic variants.
“Children with a rare early-onset condition can have the same genetic deletion or change but show the condition very differently. Some may be severely affected, some mildly affected and some almost unaffected. The rest of their chromosomes – the common genetic background – is often ignored,” notes Angelica Ronald.
The genetic maps may eventually show whether predispositions towards activity or emotionality modify how a rare condition appears. This remains a research possibility and not a tool for prognosis or treatment.
“The point is not to tell parents what their child will become. It is to understand why children differ from the earliest years and, one day, to give families better information about the support a child may need,” concludes Angelica Ronald.
