A major international study has concluded that brain scans cannot reliably predict whether a child will develop social anxiety later in life, despite decades of research linking inhibited childhood behaviour to the disorder. The paper, led by Leiden University neuroscientist Janna Marie Bas-Hoogendam and American child and adolescent psychiatrist Daniel Pine, was named Paper of the Year by the American Academy of Child and Adolescent Psychiatry (AACAP) at its annual meeting in Atlanta on 30 October, where it also received the Norbert and Charlotte Rieger Award for Scientific Achievement.
Inhibited behaviour in childhood — such as freezing when a stranger enters the room, retreating to a parent's lap, or reluctance to approach other children — is strongly associated with social anxiety disorder later in life, according to the researchers. The question was whether this inhibition is also reflected in the structure of the brain, and whether brain scans could therefore help identify early signs of the disorder.
Bas-Hoogendam and Pine set out to answer this question with a Rubicon grant from the Dutch Research Council (NWO), awarded in 2020. The grant is officially intended to support researchers in setting up a research project abroad, but because of the Covid-19 pandemic, Bas-Hoogendam worked from home with Pine and his team at the National Institute of Mental Health (NIMH) in Bethesda, USA, Leiden University reported.
The paper, published in 2025 in the Journal of the American Academy of Child and Adolescent Psychiatry (JAACAP), is titled "Structural Brain Correlates of Childhood Inhibited Temperament: An ENIGMA-Anxiety Mega-Analysis." It involved hundreds of brain scans from research centres around the world, collected through the international ENIGMA-Anxiety working group. The study has dozens of authors, reflecting the many researchers who contributed data.
The researchers used a registered report format: they submitted their introduction and methodology for peer review before conducting any analyses. Only after the methodology was approved and published in the journal did they carry out the analyses, following the preregistered plan precisely. This allowed them to publish the full article regardless of the results.
"We looked for associations between the degree of behavioural inhibition in children and their brain structure," Bas-Hoogendam explained, according to Leiden University. "We wanted to know whether early signs of anxiety that children might develop later in life could already be detected in the structure of their brains."
The analyses suggested this was not the case. "Most of our findings were null results, which are normally quite difficult to publish," Bas-Hoogendam said. "But in this case, we could publish them because the methodology had been approved in advance."
The studies included in the paper used different methods to measure behavioural inhibition. In some, researchers placed young children in unfamiliar lab settings — for instance, having a stranger come into the room or introducing a moving robot — and documented their responses. In other cases, caregivers and educators provided accounts of how they viewed the child's actions, and in additional research, adolescents and young adults recalled their own childhood conduct.
Across all these different methods, the researchers found no significant association between brain structure and inhibited temperament. They also conducted separate analyses for each of the three subgroups, but again found no consistent, statistically significant associations.
Bas-Hoogendam emphasised the importance of publishing null findings. "But publishing these kinds of null findings is extremely important to ensure that the scientific literature accurately reflects the research that has been conducted," she said, according to Leiden University. "This helps prevent certain associations and effects from appearing stronger than they actually are, while results that fail to confirm those associations remain hidden away in a drawer. This is known as the 'file-drawer effect'."
One of the features that makes this research unique, the researchers noted, is the ability to combine data from so many different studies. Previous studies were often small-scale, involving relatively small groups of children and adolescents. That is understandable, Bas-Hoogendam explained, as this combined examination of behavior and neurobiology is costly and demands extensive labor. Yet a drawback is that those prior investigations, each employing distinct emphases and methods, frequently yielded conflicting results.
By conducting a mega-analysis within the ENIGMA consortium, researchers can assess whether consistent and reproducible patterns can be identified. "This type of analysis produces more reliable results, which is extremely important for developing effective interventions, for example. You need a solid starting point for that, and large datasets are essential for establishing such a foundation," Bas-Hoogendam said.
For university students, particularly those studying psychology, neuroscience, or medicine, the study serves as a reminder that not every promising avenue of research yields the hoped-for results — and that null findings are themselves valuable contributions to science. The paper's recognition by the AACAP underscores the importance of rigorous, pre-registered research that can withstand the test of replication, even when the answer is "no."