Predicting Adolescent Mental Health Risks with AI

Senior Researcher Predicting Mental Health Issues Using Artificial Intelligence For decades, Professor Terry Moffitt, a senior clinical psychologist and world-renowned mental health researcher, has maintained deep international connections.

Recently, she was to receive an honorary doctorate from a major university, but conflict disrupted the plans, forcing her to move between shelters.
Now, Professor Moffitt discusses the new system she developed with her partner: an algorithm that identifies mental health risks among adolescents even before symptoms appear.
Where does it all start? In sleep: “Sleep proved to be a stronger predictor than trauma history.” Her connection to research did not begin in academia but through her personal life.

She married a professor from a scholarly family with a rich intellectual and academic background. Since then, her life intertwined with international research communities, and the university became a second home.
Observing Mental Health Problems in Adolescence Throughout her scientific career, Professor Moffitt has been recognized as a leading researcher in developmental psychology and mental health. “
My research follows large groups of people born in the same year from birth to death,” she explains. “One study follows 1,000 people born in 1972.

In youth, I studied the links between mental and physical health. Today, in their 50s, the research is ideal for examining the impact of mental health on aging.” She highlights two significant scientific achievements.
The first is a theory identifying two types of youth involved in legal conflicts.
“One type is ‘life-course persistent’ young people with antisocial behavior starting in childhood and continuing through life, repeatedly harming others. Fortunately, they are rare,” she describes.

“The second, more common type, includes youth whose legal issues are temporary they grow out of them and become normative citizens.”
This theory and its subsequent research influenced youth justice systems in many countries.
The second recent achievement comes from aging and biotechnology.
“My team developed a blood test measuring the pace of aging called DunedinPACE.
It is used in biotechnology for assessing interventions aimed at slowing aging and preventing diseases in clinical trials.”

Recently, Professor Moffitt led a breakthrough project using artificial intelligence to predict mental health risks in adolescents before any symptoms appear.
The goal of the system is early intervention to prevent deterioration to severe clinical conditions. Adolescent mental health risks are sharply rising, though causes remain debated.
“There are many hypotheses,” she explains.
“Does social media harm? Findings are inconclusive.
Perhaps the fact that most families have only one or two children makes parents more anxious?

There are also more psychiatric medications available, so people who previously would not see a doctor now request prescriptions for ADHD, depression, or anxiety.
Expectations for adolescents are higher today.”
She emphasizes: “The clearest fact is that most adults with mental health issues experienced their onset in adolescence.
That is why attention to adolescent mental health is so important.” Sleep as a Predictor From this understanding came the ambitious project.
“This is a classic modern collaborative science project,” she says.

“My partner and I conducted initial research showing that a single index, composed of all mental health symptoms, is the best way to predict who will need treatment in the future.
We collected data on the mental health of 10,000 American adolescents.
These data were used by the AI team to develop the model.
I value this project because collaborative science achieves much more, much faster, and the result will help many adolescents.”
Sleep turned out to be the most accurate predictor of mental health risk.

“Sleep, sleep, and again sleep,” emphasizes Professor Moffitt.
Of all the health and mental metrics we examined, sleep problems were the best indicator of future mental illness.
The good news is that effective treatment for sleep issues already exists cognitive-behavioral therapy for insomnia (CBT).
It improves sleep quality, and we are now testing if this treatment in adolescents also reduces future mental health risk.”
The model is highly accurate among the large sample of American children it was developed on.