When adolescence arrives, many young people feel as though every slip is under scrutiny. For some, that heightened self‑awareness diminishes with age; for others, it evolves into marked anxiety.
A new investigation conducted by scholars at the USC Dornsife College of Letters, Arts and Sciences offers clues about why this happens and could eventually help identify at‑risk teens before anxiety fully sets in.
The study, appearing in JAMA Network Open, found that merging two distinct brain imaging modalities produces a clearer forecast of which adolescents are likely to experience increased anxiety later on. It deepens our understanding of how the teenage brain reacts to errors and why these reactions differ across individuals.
The participants were part of a rare, long‑term project financed by the National Institute of Mental Health, which has tracked children from birth through adolescence to examine how early temperament shapes brain development and mental health outcomes. Many were tagged early in life as having a "fearful temperament"—a tendency toward shyness or caution when meeting new people or facing novel situations. Although this trait is linked to greater anxiety risk, not all children with it develop anxiety.
"We wanted to understand why some children go on to develop anxiety while others do not, even when they share similar early traits," said Emilio Valadez, assistant professor of psychology at USC Dornsife and lead author of the paper. "By exploring how the brain handles mistakes, we hoped to spot indicators of later anxiety risk."
To uncover these indicators, the researchers asked teens to perform a simple computer task that was designed to elicit minor mistakes—pressing an incorrect button in response to a rapid series of arrows. While the task was underway, the participants' brain activity was recorded using two technologies: electroencephalography (EEG), which captures electrical activity over time, and functional magnetic resonance imaging (fMRI), which shows the spatial origin of that activity.
Each method supplies a different piece of information. EEG is incredibly fast but gives only a rough view of location; fMRI is slower but offers high spatial precision.
The team scanned subjects twice—in 2017 at age 13 and again in 2019 at age 15—and created a novel data‑fusion strategy that combined the EEG and fMRI outputs. This "EEG‑fMRI fusion" produced a far more detailed picture of how the brain responds to errors and how those patterns evolve during adolescence.
"It’s like using both eyes together to gain depth perception," Valadez described. "The brain’s activity suddenly became much clearer."
Analyses revealed that EEG or fMRI alone provided limited predictive power regarding changes in anxiety. However, when the two were integrated, the results were striking.
The fused data accounted for roughly 25 % of the variation in anxiety change between ages 13 and 15, surpassing predictions based solely on early temperament, age, sex and baseline anxiety.
"This was a welcome surprise," Valadez said. "We anticipated perhaps five or ten percent, but twenty‑five percent is a significant leap. I was initially concerned I had miscalculated."
The study also showed that early temperament influenced how specific brain regions related to later anxiety. For teens who were markedly shy or cautious as infants, heightened activity in the dorsal anterior cingulate cortex—an area that signals errors and potential threats—was associated with greater future anxiety. Conversely, the posterior cingulate cortex appeared protective: Among children with fearful temperaments, increased growth in activity of this region over time correlated with lower anxiety risk.
In lay terms, the brain’s error‑processing patterns and their evolution during a critical developmental period help determine whether anxiety will ameliorate or intensify.
These findings hint that, in the future, brain imaging could assist in pinpointing which teens are most vulnerable to anxiety disorders before full symptom onset, potentially enabling earlier, customized interventions.
"This study doesn‘t immediately change daily life for families," Valadez noted. "But it moves us toward earlier anxiety risk prediction, which could help stop mental health problems before they commence."
The research also emphasizes the value of combining distinct imaging techniques.
"No single tool offers the full picture," Valadez added. "Data integration across methods yields a richer understanding of how the brain supports mental health."
Looking ahead, the team intends to investigate whether similar predictive power can be achieved at even earlier ages—perhaps when children are 8 or 9—and to explore other brain functions, such as memory or attention, that might illuminate anxiety risk.
"Our ultimate objective," Valadez said, "is to read the brain’s narrative early enough to identify children who may need extra support and provide that support before anxiety becomes a lifelong challenge."