Temperament emerges early in life. Some infants cry more readily than others, some are more energetic and some take more time to warm up to strangers. Now a new SFARI-funded genome-wide association (GWAS) study finds that genetic variants shape these early differences, some of which also map onto autism, attention-deficit/hyperactivity disorder (ADHD) and adult personality traits¹.

The effort was jointly led by Stephan Sanders of the University of Oxford, Alexandra Havdahl of the Norwegian Institute of Public Health and Angelica Ronald of the University of Surrey. Nearly 79,000 children took part, drawn from five European groups and two smaller ones in Japan and South Africa. Parents completed assessments when their children were between one and three and a half years old, with the exact timing varying somewhat by cohort.

Using genome-wide complex trait analysis, the team estimated the heritability of emotionality, activity, shyness and sociability, meaning how much of the differences between children in each trait can be explained by genetics. Shyness showed the highest heritability, at about 15 percent, followed by activity at roughly 10 percent, emotionality at 7 percent and sociability at just over 3 percent. These figures are lower than some earlier heritability estimates from twin studies, which is typical for this kind of genetic analysis².

The GWAS identified ten genetic loci with genome-wide significance. Two stood out for showing strong colocalization with gene expression in the adult human cortex, based on data from an existing brain tissue study. The first locus, near a gene called RHEBL1 (Ras Homolog Enriched in Brain-Like 1), showed an inverse relationship: Higher RHEBL1 expression in the brain was linked to lower activity levels in the third year. The RHEBL1 protein contributes to mTOR signaling, a cellular pathway involved in brain development and synaptic plasticity that was already implicated in autism, schizophrenia and depression³. The allele linked to lower activity has also been tied, in separate studies, to more years of education and stronger cognitive performance, suggesting a genetic association between early activity level and later cognitive outcomes.

The second locus, near a gene called MR1 (Major Histocompatibility Complex Class I-Related protein 1), also showed an inverse relationship, with higher MR1 expression in the brain being linked to lower emotionality. The MR1 protein plays a role in immune signaling, presenting microbial metabolites to a specialized set of T cells, making this result part of a growing body of evidence connecting immune function to early brain and behavioral development⁴. Unlike RHEBL1, MR1 has not previously been linked to any behavioral trait.

The researchers also asked whether early childhood temperament shares genetic roots with autism and ADHD. Genetic variants connected to higher sociability tended to correlate with lower likelihood of autism, whereas variants associated with higher emotionality tended to correlate with higher autism likelihood. Activity level was strongly linked to ADHD, consistent with prior observational research linking early motor activity to later ADHD diagnoses. Shyness, meanwhile, showed a lower genetic likelihood of co-occurring with ADHD.

Emotionality and shyness also showed genetic ties to adult personality. Emotionality correlated with adult neuroticism, and shyness correlated negatively with adult extraversion, consistent with how psychologists already define these adult personality traits.

Because non-European cohorts with relevant early childhood data are scarce, few loci could be tested for replication at all, and of those, the authors found that just one, the RHEBL1 locus linked to third-year activity, held up when they expanded their work to include cohorts from Japan and South Africa.

Children in the same family share a household but not identical genes: siblings inherit only some of the same DNA. That distinction lets researchers separate genetics from environment. A sibling with more of the genetic variants linked to a trait, such as shyness, tended to score higher on it than their brother or sister—evidence that genetics, not environment, explained the pattern. Still, parent-reported data carry some bias, and sociability’s lower-than-expected heritability may partly reflect how big a child’s built-in social circle is, because a child with more siblings has more people to interact with regardless of their own personality.

Together, the findings offer some of the first molecular genetic evidence connecting infant and toddler temperament to the biology of later neurodevelopmental and psychiatric conditions. The RHEBL1 and MR1 genes stand out as leading candidates for further investigation. The authors argue that this kind of foundational work matters because support and intervention tend to be most effective when started early. A clearer understanding of what drives early temperament could eventually help shape how those support systems are designed.

1. Hollowell A. et al. Nat. Hum. Behav. Epub ahead of print (2026) PubMed
2. Vollrath M.E. et al. J. Pers. Soc. Psychol. 126, 660-675 (2024) PubMed
3. Costa-Mattioli M. and Monteggia L.M. Nat. Neurosci. 16, 1537-1543 (2013) PubMed
4. McWilliam H.E.G. and Villadangos J.A. Nat. Rev. Immunol. 24, 178-192 (2024) PubMed