Remember the butterflies in your stomach and the tingling sensation that gives you shivers when you are just about to pitch an exciting project idea to your professor? Or when you finally deliver your handmade gift that you spent countless hours perfecting? Last summer, Marco Leyton, a professor in the Department of Psychiatry at McGill, experienced a similar feeling. He and his research team spent months trying to find errors in their three-factor model that can predict a lifetime history of multiple mental illnesses by tapping into just three factors: Biology, behaviour, and childhood trauma.
“Not only did the three factors predict who had a psychiatric problem, but the strength of the effect was extraordinary,” Leyton wrote in an email to The McGill Tribune. “We then spent the next few months searching for an error but couldn’t find one. It was an exciting summer.”
Fortunately, the model was accurate and could predict the participants’ lifetime history of psychiatric illnesses with 90 per cent accuracy based on incidences of childhood trauma, temperamental traits, and midbrain dopamine regulation.
The 52 participants, who were followed since birth, showed various psychiatric illnesses such as attention deficit hyperactivity disorder, mood and anxiety disorders, substance use disorders, eating disorders, and more. The fact that the three-factor model could predict a wide variety of psychiatric illnesses, Leyton argued, bolsters the notion that they may have common origins.
“Comorbidity is a norm: People who meet criteria for one disorder are also likely to meet criteria for other disorders either at the same time or in succession,” said Leyton, whose research focusses on finding causes of addiction-related psychiatric illnesses.
The striking strength of this model comes from the team’s ability to assess all three factors together for the first time.
“Childhood trauma is the most quintessential, unanimously known risk factor for every psychiatric disorder, unfortunately,” said Maisha Iqbal, the first author of the paper and a neuroscience master’s student at McGill.”
The team assessed childhood trauma from a self-report questionnaire that included questions about emotional and physical neglect and abuse. However, due to many contributing factors, including genetic predisposition, family history, resilient brains, and flexible coping skills, some people were able to live their adult lives relatively unscathed. Thus, individual factors alone cannot accurately predict the onset of psychiatric illnesses.
Researchers combined this with scores obtained from another questionnaire assessing participants’ externalizing traits between the ages of 11 to 16. These included their temperamental traits, aggression, and impulsivity.
Adding positron emission tomography scan data to the model revealed that poorly regulated dopamine increased the prediction accuracy of the model even further. Dopamine is a chemical produced in the brain that influences mood, and triggers feelings of reward, pleasure and motivation. It is the same chemical that makes people feel rewarded when someone likes their post on Instagram, or makes them feel punished when they get scolded. It is also involved in pathways regulating drug addiction and movement disorders. When misregulated, it affects one’s motivation, attention, emotional and behavioral responses to situations, posing a biological risk for various attention and mood disorders.
Administering early diagnosis and intervention protocols for psychiatric illnesses has the potential to greatly improve patient well-being at various levels. Studies like this one may help convince policymakers to encourage the use of predictive algorithms, like the three-factor model, in clinics—which are often discredited due to the inaccuracy of the ones currently available.
Further complications arise if models include neuroimaging or biological analyses, because of their logistical limitations as well as a need for higher levels of expertise to run them. But with the emergence of new technologies aiming to make portable and cost-effective neuroimaging devices, this research still holds promise.
Note that this study establishes only a predictive, rather than a causal link between these factors and psychiatric illnesses. The team’s next steps are to replicate this effect in a larger and more diverse cohort of participants.