AI Technologies May Help to Improve Early Identification of Youth with ADHD At Risk of Substance Use

Substance use disorder (SUD) – also known as drug addiction – is one of the most problematic co-occurring conditions with ADHD. Children with ADHD are twice as likely to develop SUD later in life, compared to those without ADHD [1, 2].  Identifying children and adolescents who are at risk, even before they ever use substances, can be an important step in preventing SUD. But how do you know who is at risk? Artificial Intelligence (AI) technologies are promising tools to identify risk factors and predict the likelihood of an individual developing SUD later in life.

Image by rebcenter-moscow

SUD is a mental illness that involves impulsive use of substances such as alcohol, marijuana, nicotine and opioids. Substance use often starts with experimentation in adolescence, such as cigarette smoking and drinking. Studies have found that adolescents with ADHD often start these experiments at younger ages, more quickly become heavier users and develop more severe functional impairments [3].  Indeed, younger ages of first use of a substance is associated with a higher risk of abusing the substance later in life. 

While SUD causes significant harm, and personal and societal burdens, it is actually a preventable and treatable condition. Early identification of at-risk youth is especially important because it would allow for more targeted early interventions.  Studies have shown that school- and community-based prevention programs can be highly effective at reducing substance use. Treating and managing ADHD, both with medication and behavioral therapy, has also been found beneficial in terms of reducing substance use [4, 5].  Being able to screen and identify the high-risk adolescents before their first attempt and being able to further “discourage” and “prevent” such onsets of substance use would probably be the most cost-effective prevention program for SUD.

But how do you know who is high-risk and who is not, even before they start using drugs? One method is to use Artificial Intelligence (AI) technologies in combination with very large databases.

Machine Learning Prediction Models for SUD

Machine learning is a special subset of AI technologies, that can learn hidden patterns and attributes in a completely data-driven fashion. Machine learning algorithms learn from the iterative exposures to a large amount of data, and subsequently make predictions based on what the model has learned from these training examples.

In this recently published EU-funded study in the Journal of Child Psychiatry and Psychology [6],  my colleagues and I applied machine learning models to data collected from the Swedish national registers. These registries contain family and health data of millions of people, including information on clinical diagnoses such as ADHD and SUD. For our research, we focused on children with ADHD and trained various models to predict those who would eventually have a diagnosis of SUD, and those who would not. More than 19,000 children with ADHD were used in the study. The collected information that we used to train the models included psychiatric and somatic diagnoses, family history of these disorders, socioeconomic status, and birth complications.

The machine learning algorithm produced two useful models to predict SUD in children or adolescents with ADHD. The first one makes a prediction at age 17 for future SUD diagnosis between age 18-20. This is an important period when young adults, often leaving home for the first time, are more subjective to peer influence and start their first use of substance. The second model, a longitudinal model, makes a yearly prediction at every age from age 2 to 17 for SUD diagnosis in up to 10 years in the future.  We found that both models were able to make significant predictions. This means that when we tested the models on part of the dataset (that was not used to train the model), the model was more often correct to predict SUD outcome than could be expected from chance. You can interpret this as that the model had learned certain parameters in the data that predict whether a person with ADHD will develop SUD or not.

Early Detection of SUD Risk

One important discovery was that the longitudinal model was able to make significant prediction at as early as age 2 for up to 10 years into the future. The earliest age of a valid SUD diagnosis in our dataset was at age 12. Using a method called “supervised learning”, we “taught” the model to identify those children at age 2 who would be diagnosed with SUD at age 12.  Such supervised training was carried out for each age for any given outlook of 1, 2, 5 or 10 years.  Being able successfully predict at-risk children years before their first attempt of substance use is a really promising result.  Furthermore, the yearly prediction at every age provides a monitoring system tracking the risk for SUD, either increasing or decreasing risk over years.

What have we learned from our prediction models?

It is important to stress that the predictions of these models were not perfect. The algorithms are not always correct at identifying who will develop SUD and who does not. However, they do give insight into some important predictors that are more informative than others. For example, the early detection and diagnosis of ADHD and socioeconomic status.  At the population level, the early diagnosis of ADHD was associated with a lower risk of developing SUD later in life, while adverse socioeconomic status was associated with a higher risk.  However, we could not identify any predictors at the individual level that contributed their increased or decreased risks.  The ability to identify such specific risk for each individual would be extremely useful when targeted interventions can be most effectively applied.

Despite that the prediction accuracies were modest and not ready for real-life deployment yet, these findings clearly have many, broad implications for policy-makers, parents, teachers and clinicians. Given that more and more data will become available to train these AI models, more accurate and generalizable predictions can be made at early ages regarding the disease risk, it is imperative to develop more effective and personalized measures for prevention and risk-reduction. We will therefore continue our work to expand the prediction accuracy of our models.

Overall, we provided a proof-of-concept that machine learning AI technologies – by leveraging the large volume of data, such as those of national registers and other electronic health records – can be used to predict disease risk, such as SUD, long before the disease onset.

Dr. Yanli Zhang-James is an associate professor at SUNY Upstate Medical University Department of Psychiatry and Behavioral Sciences. She is involved in the CoCA Project.

1.         Ercan, E.S., et al., Childhood attention deficit/hyperactivity disorder and alcohol dependence: a 1-year follow-up. Alcohol Alcohol, 2003. 38(4): p. 352-6.

2.         Wilens, T.E., et al., Does ADHD predict substance-use disorders? A 10-year follow-up study of young adults with ADHD. J Am Acad Child Adolesc Psychiatry, 2011. 50(6): p. 543-53.

3.         Kousha, M., Z. Shahrivar, and J. Alaghband-Rad, Substance use disorder and ADHD: is ADHD a particularly “specific” risk factor? J Atten Disord, 2012. 16(4): p. 325-32.

4.         Molina, B.S., et al., Delinquent behavior and emerging substance use in the MTA at 36 months: prevalence, course, and treatment effects. J Am Acad Child Adolesc Psychiatry, 2007. 46(8): p. 1028-40.

5.         Schoenfelder, E.N., S.V. Faraone, and S.H. Kollins, Stimulant treatment of ADHD and cigarette smoking: a meta-analysis. Pediatrics, 2014. 133(6): p. 1070-80.

6.         Zhang-James, Y., et al., Machine-Learning prediction of comorbid substance use disorders in ADHD youth using Swedish registry data. J Child Psychol Psychiatry, 2020.

What do rewards have to do with mental health problems?

Photo by Jacqueline Munguía

What do you think of when I say “rewards”? Perhaps you think of the points you collect every time you shop or the badges you get when playing a videogame. Well, then you’re right!  A reward can be anything. A good grade, going on a trip with friends, a smile, and even that dessert you crave in the middle of the night. Rewards are any stimuli with the potential to make us seek and consume them, and if we like, we will probably want to get them again [1].

Actually, you crave that dessert because you ate it once, and you liked it so much that your brain learned that eating that dessert again will make you feel good. This happens because of a neurotransmitter called “dopamine” released when you eat the dessert, giving you that little rush of pleasure. Now your brain knows what you like and will want more of that.

By now, you probably have realized that rewards are present in virtually everything we do in our daily lives. That is why seeking and consuming rewards are considered to be a fundamental characteristic of human behavior. These rewards that we keep consuming guarantee that we stay alive by eating and drinking water, for example. Rewards also have a huge influence on how we experience positive emotions, motivate ourselves to perform tasks, and learn new things [2].

What about the relationship between rewards and mental health problems?

Although rewards are natural stimuli that make us keep doing healthy and nurturing things, it can also become a problem. Rewards are not the problem itself, but some people can have an unhealthy behavior towards rewards. That’s where mental health problems come in. Did you know that most mental health conditions have alterations in how rewards are processed in the brain? It’s so common that these so-called reward processing alterations are now considered a “transdiagnostic feature,” meaning we can find them across different mental health conditions [3].

Reward processing is a term to refer to all aspects related to how we approach and consume rewards. For instance, how you respond after getting a reward (responsiveness), how motivated you are to go after a reward (drive/motivation), how impulsive you are when trying to get new and intense rewards (fun-seeking/impulsivity). So, as you can see, it’s not only about getting the rewards, but many different things play a role in a simple action we do.

Let’s think of an example: You are going to a party with your best friends. You are motivated to go out with your friends because you’re always happy when you are around them [this is the drive/motivation]. Once you are at the party, you meet your friends, talk, laugh and are happy you decided to join because you’re feeling that rush of pleasure [this is the responsiveness aspect]. At some parties, things can get a bit out of control, and some people might do risky things on the spur of the moment, like binge drinking. You refuse to binge drink because you thought of the risks, and you don’t want to be in trouble later [that’s the third aspect, the fun-seeking/impulsivity].

Now, let’s think of how that party would be for people with reward processing alterations. In the case of a very high drive, they would be super motivated to hang out with friends. On the other hand, if they have low responsiveness, they wouldn’t be able to have fun at the party even though all of their friends are there and the party is super fun. Lastly, in the case of high fun-seeking/impulsivity, they wouldn’t think of the risks and consequences and engage in binge drinking anyways.

As I mentioned before, these alterations play a role in different mental health conditions. They can affect one or more aspects of reward processing, and they can be either lower or higher than average. For example, people with ADHD can show higher risk-taking, meaning that they are more susceptible to take big risks without thinking about the consequences [4]. This impulsive behavior might be a reflection of the altered fun-seeking aspect of reward processing. Another example is the lack of interest in social interactions in people with autism spectrum disorders [5]. This lack of interest might reflect a reduced drive/motivation to go after social rewards.

These are just some examples of what reward processing alterations might look like in the context of mental health problems. There are still a lot of open questions. As part of my PhD research, I am trying to answer some of them. For example, which came first? Are reward processing alterations causing mental health problems, or are they just mere symptoms of these conditions? If you want to learn more about this topic, stay tuned as more blog posts will come!

Dener Cardoso Melo is a PhD candidate at the University Medical Center Groningen (UMCG). He is using data from the CoCA project together with other datasets to investigate the potential causal role of reward processing alterations in different mental health conditions.


  1. Schultz, W. (2015). Neuronal reward and decision signals: From theories to data. Physiological Reviews, 95(3), 853-951. doi:10.1152/physrev.00023.2014
  2. Wise, R. A. (2002). Brain reward circuitry: Insights from unsensed incentives. United States: Elsevier Inc. doi:10.1016/S0896-6273(02)00965-0
  3. Zald, D. H., & Treadway, M. T. (2017). Reward processing, neuroeconomics, and psychopathology. Annual Review of Clinical Psychology, 13(1), 471-495. doi:10.1146/annurev-clinpsy-032816-044957
  4. Luman, M., Tripp, G., & Scheres, A. (2010). Identifying the neurobiology of altered reinforcement sensitivity in ADHD: A review and research agenda. Neuroscience and Biobehavioral Reviews, 34(5), 744-754. doi:10.1016/j.neubiorev.2009.11.021
  5. Stavropoulos, K. K., & Carver, L. J. (2018). Oscillatory rhythm of reward: Anticipation and processing of rewards in children with and without autism. Molecular Autism, 9(1), 4. doi:10.1186/s13229-018-0189-5

Prevalence and cost of ADHD comorbidity

Do individuals with ADHD more often suffer from depression, anxiety, substance abuse or severe obesity, than individuals without ADHD? Are there differences between men and women in how often this is the case? Does having ADHD in addition to one of these conditions result in higher health care costs?

The short answers to these questions, are yes, yes and yes. In the CoCA-project, researchers have investigated these questions using very large datasets including Scandinavian birth registries that contain information of millions of people. This allows us to get a better understanding of how often conditions occor, how often they occur together, and how often they occur in men vs women. Furthermore, we have investigated health insurance data from Germany to study patterns of health care costs associated with ADHD and its comorbid conditions.

The interpretation of these data is however not simple. That is why we have recorded a webinar with dr. Catharina Hartman from Groningen, The Netherlands. She is the leader of these studies and can explain what these findings can and cannot tell us. The webinar ends with implications for policy makers and health care professionals, based on these findings.

ADHD and cannabis use

It is not uncommon for individuals to suffer from two or more psychiatric disorders at the same time. The appearance of these disorders frequently follows a specific order, and one disorder may predispose to others, all of which in combination contribute to the worsening of the quality of life of the individuals who suffer them. This is usually associated with more severe symptoms and worse prognosis. In addition, making a diagnosis and applying personalized treatments becomes more challenging in this context. By investigating the genetic overlap between disorders, we gain better understanding of why the disorders frequently co-occur.

In mental health, substance use disorders often appear when there is another mental condition. This is the case for attention-deficit/hyperactivity disorder (ADHD) and substance use disorder, where individuals with ADHD are more likely to use drugs during their lifetime than individuals who do not have ADHD. In particular, cannabis is the most commonly used substance among individuals with ADHD, which can also lead to the use of other drugs and to the worsening of their symptoms. ADHD is one of the most common neurodevelopmental disorders, affecting around 5% of children and 2.5% of adults, and is characterized by attention deficit, hyperactivity and impulsivity. Both ADHD and cannabis use are conditions determined partly by environmental factors but where genetic factors also play an important role.

We recently investigated the genetic overlap between ADHD and cannabis use, and found that the increased probability of using cannabis in individuals with ADHD, can be, in part, due to a common genetic background between the two conditions. We identified four genetic regions involved in increasing the risk of both ADHD and cannabis use, which could point to potential druggable targets and help to develop new treatments. In addition, we confirmed a causal link between ADHD and cannabis use, and estimated that individuals with ADHD are almost 8 times more likely to consume cannabis than those who do not have ADHD. This evidence goes in line with a temporal relationship, where the ADHD appears in childhood and the use of cannabis during adolescent or adulthood. This suggests that having ADHD increases the risk of using cannabis, and not vice versa.

This research has only been possible thanks to large international collaborations by the Psychiatric Genomics Consortium (PGC), iPSYCH, and the International Cannabis Consortium (ICC), where the genomes of around 85 000 individuals were analysed.

Overall, these results support the idea that psychiatric disorders are not independent, but have a common genetic background, and share biological pathways, which put some individuals at higher risk than others. This will help to overcome the stigma of addiction and mental disorders. In addition, the potential of using genetic information to identify individuals at higher risk will have a strong impact on prevention, early detection and treatment.

Further reading

María Soler Artigas et al. Attention-deficit/hyperactivity disorder and lifetime cannabis use: genetic overlap and causality, Molecular Psychiatry (2019) –

About the author

María Soler Artigas is postdoctoral researcher at the Psychiatry, Mental Health and Addictions group at Vall d’Hebron Institut de Recerca (VHIR), also part of the Biomedical Research Networking Center in Mental Health (CIBERSAM). Her research is part of the CoCA consortium that investigates comorbid conditions of ADHD.