Persistence and emergence of ADHD in young adulthood: Unravelling aetiology using genetics in a population-based longitudinal cohort

While ADHD was historically thought to only affect children, it is now recognized that it may continue into adulthood. Studies identify about 2.5-5% of adults as having ADHD, which is associated with poor outcomes including unemployment, substance abuse, and even higher mortality. I recently conducted a study that found not only may ADHD continue into adulthood, but the disorder may also newly begin in young adulthood. This is a novel finding, as ADHD is currently understood to begin in childhood. My findings are supported by results from two other recent studies in Brazil and New Zealand that also identified late-onset ADHD groups. Currently, little is known about those adults who did not have ADHD in childhood but developed it in adulthood. Who are individuals with late-onset ADHD? Did they have a disposition to develop ADHD as children, but grew up in very supportive households that masked their disorder until later life? Or do these people not have ADHD in adulthood, but rather another disorder with symptoms that could look like ADHD?

Genetic studies can provide crucial insights in answering such questions about late-onset ADHD, and indeed broader questions about ADHD in young adulthood. One way to understand how late-onset ADHD might be similar or different to childhood ADHD is to investigate whether people with late-onset ADHD have higher levels of risk genes for childhood ADHD. If so, this would support the idea that people with late-onset ADHD would have shown ADHD in childhood, but the disorder was perhaps masked by a supportive family environment. In contrast, if people with late-onset ADHD do not have more risk genes for ADHD, but rather for other mental health disorders, this would suggest that late-onset ADHD may be explained by other mental health problems with symptoms similar to ADHD.

The age at which ADHD symptoms may appear varies broadly from person to person, with some individual's symptoms beginning in infancy, while for others symptoms may not emerge until adolescence or later. I will examine how genes affect whether an individual develops either childhood or adult-onset ADHD, as well as how genes affect when ADHD symptoms first begin across a range of ages over development. It is not known whether some genes may increase risk for a very early onset of ADHD (e.g. infancy) whereas others increase risk for later onset (i.e. late childhood), and whether genes that affect when ADHD begins are different from those that affect whether an individual develops ADHD in the first place.

Genes do not act alone to increase risk for ADHD. Rather, people's genes interact with their environment to influence the development of ADHD. For example, even people with many ADHD risk genes may not develop the disorder if they grow up in a protective environment; likewise, people with few ADHD risk genes might develop the disorder if they grow up in adverse environments. I will explore aspects of the social and family environment that could protect against or exacerbate genetic risk for ADHD. One pathway through which the environment can influence risk of ADHD is through epigenetics changes, which can turn genes 'on and off'. In this way, two identical twins who share 100% of their DNA may express different genes depending on epigenetic variations. I will compare epigenetic differences among twin pairs in which one twin has late-onset ADHD and the other does not, to identify biological changes that may be markers of young adult ADHD.

To-date, the majority of ADHD research has assumed that all cases of adult ADHD began in childhood. However, I and other researchers have recently found this may often not be the case. Understanding how, why and when adult ADHD may develop is a crucial question and has broad implications for the diagnosis and treatment of adult ADHD.

ADHD is understood as a childhood-onset neurodevelopmental disorder. However, in my recent longitudinal study, less than one-third of individuals with young adult ADHD had ADHD persistent from childhood, while over two-thirds did not have the disorder in childhood. Many questions remain as to the nature of persistent and 'late-onset' ADHD. The aim of my proposed project is to combine data from large-scale genetic studies with a population-based cohort study to investigate the persistence and emergence of ADHD in young adulthood.

More specifically, I aim:
(1) to assess whether late-onset ADHD is associated with: (a) childhood ADHD polygenic risk score, suggesting childhood and late-onset ADHD share genetic aetiology, or (b) polygenic risk scores for other mental health disorders, such as alcohol dependence and depression, which would suggest late-onset ADHD may be accounted for by other psychopathology that mimics ADHD in young adulthood;
(2) to utilize genetic data available in the Psychiatric Genomics Consortium to perform GWAS analyses examining ADHD age of onset as a quantitative trait, as well as to examine whether ADHD polygenic risk score is associated with age of onset among clinical cases of ADHD;
(3) to determine whether risk and protective factors in the social environment moderate genetic risk, as assessed by ADHD polygenic risk score, for the emergence and persistence of ADHD in young adulthood;
(4) to investigate whether DNA methylomic variation: (a) is associated with polygenic risk score for ADHD in a population-based cohort study, and (b) differs between monozygotic twins discordant for late-onset ADHD (n=49 pairs).

This project has important implications and opportunities for researchers and clinicians in understanding young adult ADHD. Investigating the genetic and epigenetic underpinnings of young adult ADHD will provide invaluable insights that will affect the direction of future research and could influence clinical decision-making.

ADHD is one of the most prevalent behavioural disorders in childhood and remains relatively common in adulthood, affecting approximately 2.5-5% of adults. Adult ADHD is also associated with several negative and costly outcomes including unemployment, substance abuse and even higher mortality. Untreated ADHD is associated with lower productivity at work, as well as more volatile behaviour; one study found that individuals with ADHD are more likely to commit crimes in periods when they are untreated compared to times when their ADHD is treated. Therefore it is of the utmost importance to recognize ADHD in adulthood and offer individuals with ADHD appropriate and effective treatment.

Recent research by myself and others has found that a large proportion of the adult ADHD population did not have the disorder in childhood, and instead had "late-onset" adult ADHD. A better understanding of late-onset ADHD has important impact in several areas. Most immediately for clinicians and nosologists, whether to require childhood-onset of ADHD is a topic of debate. The genetic and gene-environment analyses in the current proposal will shed light on the nature of late-onset ADHD and its relationship to childhood-onset ADHD, which will inform the debate around whether childhood onset should be required to receive a diagnosis of ADHD in adulthood. This has direct impact on individuals with ADHD, who may be turned away from treatment or not reimbursed by insurance companies if they do not meet this age of onset criterion. In the longer term, a better understanding of the aetiology of late-onset ADHD could inform treatment for adult ADHD. For example, whether childhood-onset and late-onset ADHD respond similarly to medication or behavioural treatment is an open question. Disentangling the genetic make-up of young adult ADHD will inform clinical treatment decisions and, in the future, potentially point to new areas of drug development.

This research could lead to advances in personalised medicine. For example, ADHD polygenic risk scores, in combination with assessment of the family environment, could someday be used to identify children with ADHD who are at high risk for a particularly persistent course of the disorder and who could especially benefit from increased intervention and treatment. These methods may also identify children for whom elevated ADHD symptoms in childhood represent a milder, transient developmental phase and who are likely to outgrow their ADHD without intensive intervention. The findings from the current proposal will make a significant contribution to understanding how genetics affect the course of ADHD across development and move towards the goal of using this information in the service of personalised medicine. This could benefit individuals living with ADHD, as well as their friends and families, by helping to tailor treatments to be most effective for the individual.

Additionally, genomic and epigenetic research may help to identify important new biological pathways and biomarkers associated with adult ADHD. GWAS analyses examining genetic variants associated with ADHD age of onset could identify novel neurobiological pathways associated with the timing of when ADHD symptoms appear. Epigenetic findings could point to novel biomarkers of disease risk and potential targets for further investigation in neurobiological systems.

Finally, an important aspect of this proposal is international, cross-cohort collaboration. In the current proposal, I plan to continue and extend my collaboration with investigators at the Pelotas Study in Brazil and the Dunedin Study in New Zealand. My longer-term goal is to expand this collaboration of ADHD research beyond these two cohorts to include a broader consortium of investigators studying ADHD over the life course. It is my hope that increased communication among researchers in this area can lead to increased replication, sharing of results and novel discoveries.