Does late‐onset attention‐deficit/hyperactivity disorder exist? JCPP Annual Research Review 2019

journal covers-JCPP
Bringing together empirical research, clinical studies and reviews in order to advance how we understand and approach child and adolescent mental health. Twitter @TheJCPP

Posted on

JCPP Annual Research Review: Does late‐onset attention‐deficit/hyperactivity disorder exist?

Professor Philip Asherson and Dr Jessica Agnew‐Blais discuss their paper.

You can listen to this as podcast on SoundCloud or iTunes.


Professor Philip Asherson

Philip Asherson, MB,BS, MRCPsych, PhD is Professor of Molecular Psychiatry at the MRC Social, Genetic and Developmental Psychiatry centre at the Institute of Psychiatry, King’s College London, and consultant psychiatrist at the Maudsley Hospital.

Philip’s research interests include genetics of ADHD and related neurodevelopmental disorders; clinical and genetic studies of ADHD in adults; mapping genes in common complex neuropsychiatric disorders; functional studies aimed at delinating the brain processes that mediate genetic risk on ADHD.

Dr Jessica Agnew-Blais

is a Research Fellow at the MRC Social, Genetic & Developmental Psychiatry Centre, King’s College London. Her research takes a life course approach to psychiatric epidemiology and focuses on understanding risk and protective factors in the progression of attention deficit hyperactivity disorder (ADHD) across development. Dr. Agnew-Blais received her doctorate in Epidemiology from the Harvard School of Public Health and her BA from Stanford University in the Department of Human Biology.


[00:00:36.345] Professor Philip Asherson: Hello, my name is Philip Asherson.  I’m a Professor of Psychiatry at King’s College London.  I work in the Social, Genetic and Developmental Psychiatry Department.

[00:00:45.510] Dr. Jessica Agnew-Blais: And I’m Jessica Agnew-Blais.  I’m an MRC Postdoctoral Research Fellow, also at the SGDP Centre at the Institute of Psychiatry, Psychology and Neuroscience at King’s College London.  Philip, can you give a bit of background on the age of onset of symptoms of ADHD?

[00:01:03.763] Professor Philip Asherson: Well, I just want to go back to the way we first thought about ADHD, and it seems that most of the early descriptions of ADHD were of young children, who were overactive, restless, inattentive.  And so, really, we’ve always thought, or for a long time, we thought of ADHD as mainly a childhood disorder and in fact, people thought it was only a childhood disorder and people would, kind of, grow out of it by the time were – they were adults.  And then, I – people began to realise that if you followed up children with ADHD as they grew older, there were cases of ADHD that persisted through into adulthood.  And so, people began to realise and understand that actually, ADHD can, and often does, persist through into adulthood.

Because of this idea of it being an early-onset neurodevelopmental disorder, people started to think about how do we define this?  And in the DSM-III, that was the first time a specific age was given, and they decided that ‘age seven’ was the criteria.  So, you would have symptoms and impairments of ADHD prior to the age of seven.  And this criteria was carried on through into the DSM-IV.  So, for a very long time, the minimal criteria was “Some symptoms with impairments prior to seven.”  So, it was interesting that more recently, in DSM-5, that criteria has changed, it’s become older, and “Several symptoms, with or without impairment.”  So, very different, quite a big change from what went before.

So, the reasons for this were largely coming from people, sort of, looking at the validity of the age seven criteria, and what they found was that the age of onset of age seven criteria had very little prediction.  It didn’t seem to predict the severity of the disorder or the treatment response, for example, or even some of the neurocognitive deficits.  And so, it didn’t feel right that age of seven would be included in the criteria ‘cause it had no predictive value.  And the next step was really to home in on this age, and it was decided that 12 was perhaps the most appropriate age.

So, more recently, Temi Moffitt published this paper in the American Journal of Psychiatry and that got a lot of attention, and this was a population study, it was a prospective study, from childhood, right through to adulthood.  In terms of ADHD, the data in childhood went up to the age of 15 and then, in adulthood, they had more data at the age of 38.  But you’ll notice they had no data in-between 15 and 38.  And the key finding was, first of all, that a lot of people with ADHD seemed to be growing out of the syndrome as time – as – by the age of 38.  But perhaps the more interesting and novel finding was that there was a lot of people who had ADHD symptoms as adults, symptoms and impairments, and they were just as impaired as the childhood onset group.  When you looked back into their childhood, they know – they didn’t have significant symptoms of ADHD as children.

So, subsequent studies also used population designs.  The Brazilian group looked at using their Pelotas Study and then, the E-Risk Study in the UK.  But those studies went up to about the age of 18 and perhaps surprisingly, they had very similar findings.  They found there was a group of young adol – young adults, or older adolescents, who met ADHD criteria currently, but when you looked back, they wouldn’t have met ADHD criteria before the age of 12.  So, I was wondering, Jessica, you’ve looked into this in a bit more detail.

[00:04:43.586] Dr. Jessica Agnew-Blais: Hmmm hmm.

[00:04:43.586] Professor Philip Asherson: What do we know about, you know, when these symptoms arose and did they all have symptoms, sub-threshold symptoms, or were there people who had none?

[00:04:50.400] Dr. Jessica Agnew-Blais: Sure.  That’s a really good question.  One of the studies that’s looked, in the most detailed way, at this idea of sub-threshold symptoms in these late-onset cases is by Cooper et al and the ALSPAC cohort, which is another longitudinal cohort in the UK.  And they took an approach where they defined a potential late-onset group using their scores on the Strengths and Difficulties Questionnaire.  So, those who fell below the cutoff for ADHD, at age 12, were – and above the cutoff, at age 17, were classified as potential late-onset cases.

And then, among this group, they further subdivided into a group that were misclassified as late-onset.  So, these are individuals who, when you looked at their SDQ scores, the age of seven, eight, nine and 12, had slightly raised, high or very high scores in childhood.  And they also had a group who were genuine late-onset.  So, these individuals had only low to average scores on the SDQ across the ages of seven, eight, nine and 12.  So, they found that this misclassified group, who had some elevated ADHD symptoms in childhood, accounted for about 75% of this potential late-onset group.  So, it appears that a large proportion of those with late-onset ADHD may show some evidence of the disorder in childhood.

Another study to investigate this question is the MTA Study, which is a large trial of ADHD treatment that also included a control group without childhood ADHD.  So, this is the local normative comparison group.  This group was followed many times after baseline, with quite comprehensive assessments of ADHD and other mental health problems, allowing for a detailed account of whether ADHD could emerge after childhood.  Using rather stringent criteria to define late-onset ADHD, this study found that overall, about 3% of this control population developed the full syndrome after age 12.

[00:06:43.680] Professor Philip Asherson: So, yeah, just a brief comment.  I found the MTA Study particularly interesting, ‘cause I mean, they had really high quality measures, that made it a really nice study and similar to clinical diagnostic procedures.  And the whole way they went about their study, they tried really hard to remove people from the dataset who could possibly have had ADHD in childhood, or if there was another explanation for later-onset ADHD.  So, for example, I know there was a clinical process in which many people with comorbidities, they made the judgment that the comorbidity might be causing their ADHD and therefore, they didn’t have ADHD.  But I have to say, I think that is a hard call, and we were going to think a little bit about, you know, why the data might be – you know, could there be explanations that would give rise to an apparent late-onset ADHD when, perhaps, it doesn’t exist?  There may be various biases or reasons that we had to take account of.

I mean, one thing – I mean, a very common explanation for late-onset of ADHD is these are perhaps children with ADHD who are very well compensated, they’re very well supported, and this idea that people with very high IQ are much less – are perhaps less likely to display ADHD as children, but it might all come out later, you know, when they go to university or have to do more on their own.  So, I found it quite interesting that the data doesn’t really seem to strongly support that view.  There are some other interesting explanations.  I was thinking about rater effects.

[00:08:20.700] Dr. Jessica Agnew-Blais: Hmmm hmm.

[00:08:22.958] Professor Philip Asherson: What do we know about that?

[00:08:23.015] Dr. Jessica Agnew-Blais: That’s a good question.  So, some people have hypothesised that this influx of individuals into the ADHD population in young adulthood could be due to a change in reporter of ADHD symptoms that we have in some of these studies.  So, for example, in the E-Risk and Pelotas studies, childhood ADHD symptoms are reported by parents and Teachers, whereas young adult ADHD is based on self-report.  So, it could be when you change who is reporting on ADHD symptoms, you have different people enter the ADHD population.  However, we concluded that this couldn’t explain all of late-onset ADHD, as several studies, such as the ALSPAC cohort and the Brazilian High-Risk Study, used parental report both in childhood and in young adulthood, and they still identified large proportions of late-onset ADHD.

[00:09:15.783] Professor Philip Asherson: So, it seems like there are these cases who had very few symptoms of ADHD as children, even though they’re presenting with both symptoms and impairments as young adults.  But is it true that they were entirely free of problems, or did they have other potential precursors, perhaps, or other syndromes?

[00:09:36.120] Dr. Jessica Agnew-Blais: Hmmm.  I think we’re beginning to see with some of the more recent studies that while these late-onset ADHD individuals didn’t meet full criteria in childhood, or maybe didn’t even have elevated ADHD symptoms, they often show signs of other externalising problems, like conduct disorder or oppositional defiant disorder.  So, it seems that while they may be free of the ADHD syndrome per se, they do have some other precursors, potentially, in childhood, prior to meeting ADHD criteria in young adulthood.

[00:10:08.490] Professor Philip Asherson: So, a question has arisen of whether, sort of, early-onset and late-onset forms of ADHD could be distinct disorders.  I think we need to do more genetic research to see what is the overlap between early-onset and later-onset forms of ADHD?  But what I have learnt from twin studies is it does look as if new genetic effects come online at different developmental ages.  And so, to me, it wouldn’t be at all surprising if different maturational processes, you know, that are starting, you know, both in early childhood, but also through adolescence.  You know, it could be that the lags or the delays or the deficits are arising during adolescence, not only during early childhood.

The other idea is that later-onset forms of ADHD may just be a milder form.  You know, perhaps there’s less genetic loading, and then, also, it might be that later-onset forms, there’s more environmental risks.

[00:11:04.186] Dr. Jessica Agnew-Blais: Hmmm hmm.

[00:11:04.186] Professor Philip Asherson: But again, you know, like, maybe stress brings it out, as you’re getting older.

[00:11:07.790] Dr. Jessica Agnew-Blais: Hmmm.

[00:11:07.790] Professor Philip Asherson: But again, we really lack any data to…

[00:11:08.840] Dr. Jessica Agnew-Blais: Hmmm hmm.

[00:11:10.493] Professor Philip Asherson: …make any clear statements on this.

[00:11:12.937] Dr. Jessica Agnew-Blais: So, to summarise, we found that about one to 2% of those without childhood ADHD would have onset of the full syndrome after age 12.  However, many late-onset ADHD cases may show some ADHD symptoms in childhood, or evidence of other disorders, such as conduct disorder, or other externalising problems.  However, more research is needed, especially more longitudinal research, following these individuals with late-onset ADHD forward, because what we really don’t know is whether ADHD may onset after childhood and then become more of a trait-like syndrome that continues for the rest of these individuals’ lives.  Whether there might be some, sort of, adolescent limited ADHD that has a peak in late adolescence and young adulthood and then, wanes later in life.  Or if these individuals may have some, sort of, more fluctuating course of ADHD, where at different times in their lives, experiencing different stressors, they may fall above threshold and other times, fall below, and wax and wane over their life course.

[00:12:10.250] Professor Philip Asherson: Okay.  So, we would like to thank you very much for listening to us.

[00:12:16.490] Dr. Jessica Agnew-Blais: Thank you.

[00:12:24.550] Professor Philip Asherson: And thank you very much.

Add a comment

Your email address will not be published. Required fields are marked *