Great piece Mike (although, for you, a bit controversial). Could you recommend an accessible resource for why different ADHD presentations ended up as the same diagnostic category in the first place? I worked with both presentations and while I know we are supposed to see them as different sides of the same coin, the nature of work is just not the same - in my experience. In the same time, it is not something I specialise in, so I might not be looking at things deeply enough.
Hi Ana, you always ask great questions. The literature on this is quite large, but my understanding is that it was dsm iii that put these entities together based on the two symptoms clustering in epidemiological data, shared family histories and response to stimulant treatment. I don’t know what the best reference is but I think people like chapter 1 of the adhd Barkley book: https://psycnet.apa.org/record/2014-57877-001
At least 6 of my immediate family are diagnosed with adhd and all benefit from stimulant meds. Heritability is clear. No two of us have the same ‘problems’. Timing issues affect 4 of us. Sleep problems 3 but a different group. Huge variation in attention, huge emotional dysregulation in all but one. Co - morbid autism in 2.
3 of us are in our 70’s, symptoms present since birth.
Great piece — really clear and very well written. ADHD researchers have done a great job working at the intersection of phenomenology and neuroscience, and I think you explained that iterative process really well.
To me, ADHD sits in a somewhat different place compared to many other psychiatric diagnoses. Unlike conditions with a more recognizable phenomenological core — like melancholic depression, bipolar disorder, or paranoid schizophrenia — ADHD doesn’t seem to come with a clear, unified subjective experience. Instead, it looks more like a strong transdiagnostic trait: highly heritable, strikingly consistent across cultures, and clearly responsive to treatment.
I’m generally not a fan of the idea that we should treat symptoms rather than disorders. But ADHD research feels refreshingly honest in this regard. As you point out, there’s a kind of intellectual humility in acknowledging what this diagnosis can (and can’t) explain, and in being realistic about what we can actually expect from it.
Thanks for the thoughtful comment! I’m actually just done writing a piece on orexin2-agonists as possible future treatments for ADHD (well, at least some with that diagnosis). It will come out tomorrow, and if end up taking a look, I would very much appreciate your input. Cheers.
This was a fascinating read. I did a a one year fellowship at a military medical center & one of my responsibilities was to administer the computerized Connors Continuous Performance Test for children we were evaluating for ADHD. This also includes a written behavioural journal for parents & the primary teacher, but this computerized test - which I found downloadable for free here: https://www.millisecond.com/download/library/v7/cpt/ax_cpt/ax_cpt/ax_cpt.web - measures attention.
So, as I would be writing case notes, the Chief Psychiatrist/Captain of the unit came to me after noticing a "pattern" in my writing that included included repetitive words (e.g. and and, the the, etc.) frequent misspellings, and frequent missing words in sentences, and asked if this was something I had noticed. I told him that it was an "adult" problem that plagued me, but even though I "proofed" my writing many times before submitting anything, rarely, if ever, was I able to see these errors until much later, sometimes days later. He dragged me to the testing lab and forced me to TAKE the Connors test, which I failed, twice. "You need to consider that you have an attention deficit, as words that are "seen" in your mind are not making it to the computer screen, and your errors are literally not being "seen" in real time." Wow. I told him the story of a undergrad college professor who once caught me in the hall and said that she went to my student advisor to check if he believed I had actually written a paper I had turned in (he believed it!) - apparently she believed me to be a "dolt," - yet she gave me a D- for all the spelling errors! In any case the Captain put me on a stimulant, which resolved these issues, but I just couldn't tolerate the feeling & atomoxetine simply wasn't effective. The diagnosis, however, qualified me for free dictation software, and so it goes...
Very nice essay. I'm curious how separable you think the decision making and reinforcement mechanisms actually are. One way I view cognitive control/attention is as internal action selection with a policy controlled through reinforcement learning and with "hyperparameters" (like discount rate) governed by neuromodulators. Aspects of the decision making algorithms can be reframed in this view: poor integration could be a failure of attention which is a form stimulus selection, and too low an evidence threshold is also a form of inappropriate discounting because it's an unwillingness to wait for more certainty. This model also accounts for the anxious case because anxiety naturally should decrease action thresholds, because it implies a threat of danger and a greater cost to continued inaction and thus a higher discount rate. I think this unified model, in which each presentation of ADHD is a type of maladaptive internal action policy, also helps explain why they all benefit from stimulants, which have a dual role of modifying reinforcement signaling as well as modulating the online algorithm. Of course, the reason why they have this dual action is because those functions are fundamentally linked.
It also predicts that timing-type ADHD would be less responsive to stimulants because, while the cerebellum is involved in sequencing and control, it's not at the level of action selection. I had a friend that had this type because he was missing part of his vermis.
Also, this doesn't diminish the value in teasing the paths apart, but seeing how they can be re-synthesized as part of a more overarching theory can give us a more precise understanding of the phenotype.
I agree with a lot of what you said. The cognitive control vs RL is certainly dissociable with the right task and model. Anne Collins had a very nice series of papers with Michael Frank using this setup and showed quite nicely that in schizophrenia, RL is intact but working memory is not. I think a task battery capable of doing various types of dissociations would be helpful for stratification and phenotyping.
Oh yeah I agree there are scenarios where you can dissociate parts of the internal action-selection loop, such as—as you say—in schizophrenia, where the fidelity of certain “actions,” like maintaining items in working memory, is degraded. My point is that by understanding that these can be unified in a single functional loop, it helps to understand the interdependence and overlapping mechanisms between them.
For example, this nicely helps explain why dopamine is involved in both RL and making cortical attractor basins more robust. Both are components of the same overarching selection process—at least in this choice of metaphor. All of these things are ultimately abstractions and choices of semantic frame, so I don’t want to make it sound like this is the unique way of thinking about it.
'Seen this way, the three vignettes make more sense. The seven-year-old may have had noisy decision-making that improved with brain development. The adult woman may have a steep discounting of future rewards. The anxious twelve-year-old may have intact decision-making but disrupted thresholds in stressful contexts. These are testable ideas, and they point toward why one-size-fits-all categories struggle to capture the reality of ADHD'
or each person in each of these vignettes just like every other human being, could have had a wide range of issues causing their distress that the label and drugs simply masked, as is the case for every other DSM construct and psychiatric drug.
Everyone has a uniquely complex context and history, so many moving parts many of which are beyond our awareness horizon. Mainstream psychiatric constructs and drugs are horribly reductive, self limiting and cause more harm than they can ever do good.
In my wee opinion the normalisation of the mass labelling and drugging of tens of millions of human beings, children and adults alike (and now pets) is the worst era of psychiatry to date. Thats really saying something given its long and barbaric history of blunt instruments dressed up in white coats.
'that requires updating our measurement tools and our validation frameworks, but let’s not discard the gains already made' Gains? true the pharmaceutical companies have gained and, psychiatry appears to maintain its illusions on the back of this.
However, for the general public and human wellbeing there are no gains, suffering is worse year in year out and the harms done to people by these labels and drugs is there for anyone to see.
The mass uncontrolled experiments have been done, globally - there is now so much unmetabolized chemical junk being urinated out from humans subjected to these 'treatments' its changing the behaviour of marine life.
I've worked in the mental ill health industry for 25 years and I can honestly say that the thought of my own suffering or those I love, leading to a consultation with a psychiatrist, a rushed doctor or a DSM deluded psychologist absolutely terrifies me.
I appreciate your input. I’m optimistic that the future will be better, and that we will have more ways to identity and treat human suffering. Happy New Year.
Great, articulate comment. I agree with many of your thoughts, and, particularly, “people love hearing themselves described in quasi-clinical terms.” I also agree that the distractibility in mania is qualitatively different from what brings most adults in for quasi-clinical discussions of their attention span . Nassir Ghaemi has criticized the lack of validity of ADHD (and most other DSM diagnoses) in his Substack and elsewhere using many of the same lines of reasoning and studies as Dr. Halassa here, but Dr. Ghaemi always loses me when mentioning distractibility in mania as an example of a valid, more likely explanation for a significant portion of whatever entity/ies we provisionally call ADHD in adults.
I agree focusing too much research on disembodied parts can be a potential pitfall in pursuit of diagnostic validity and clarity, but I actually support testing these various hypotheses of ADHD. Not all of the hypotheses will be fruitful, and maybe none will, but one or more may help validate the different clinical forms we observe in practice, which would be a welcome finding IMO.
Great piece Mike (although, for you, a bit controversial). Could you recommend an accessible resource for why different ADHD presentations ended up as the same diagnostic category in the first place? I worked with both presentations and while I know we are supposed to see them as different sides of the same coin, the nature of work is just not the same - in my experience. In the same time, it is not something I specialise in, so I might not be looking at things deeply enough.
You mean inattentive vs hyperactive?
Yes, I was referring to that. But don't worry about it if nothing comes to mind!
Hi Ana, you always ask great questions. The literature on this is quite large, but my understanding is that it was dsm iii that put these entities together based on the two symptoms clustering in epidemiological data, shared family histories and response to stimulant treatment. I don’t know what the best reference is but I think people like chapter 1 of the adhd Barkley book: https://psycnet.apa.org/record/2014-57877-001
🙏🙏
At least 6 of my immediate family are diagnosed with adhd and all benefit from stimulant meds. Heritability is clear. No two of us have the same ‘problems’. Timing issues affect 4 of us. Sleep problems 3 but a different group. Huge variation in attention, huge emotional dysregulation in all but one. Co - morbid autism in 2.
3 of us are in our 70’s, symptoms present since birth.
I think you are on to something!
Thanks for your input!
Great piece — really clear and very well written. ADHD researchers have done a great job working at the intersection of phenomenology and neuroscience, and I think you explained that iterative process really well.
To me, ADHD sits in a somewhat different place compared to many other psychiatric diagnoses. Unlike conditions with a more recognizable phenomenological core — like melancholic depression, bipolar disorder, or paranoid schizophrenia — ADHD doesn’t seem to come with a clear, unified subjective experience. Instead, it looks more like a strong transdiagnostic trait: highly heritable, strikingly consistent across cultures, and clearly responsive to treatment.
I’m generally not a fan of the idea that we should treat symptoms rather than disorders. But ADHD research feels refreshingly honest in this regard. As you point out, there’s a kind of intellectual humility in acknowledging what this diagnosis can (and can’t) explain, and in being realistic about what we can actually expect from it.
Thanks for the thoughtful comment! I’m actually just done writing a piece on orexin2-agonists as possible future treatments for ADHD (well, at least some with that diagnosis). It will come out tomorrow, and if end up taking a look, I would very much appreciate your input. Cheers.
This was a fascinating read. I did a a one year fellowship at a military medical center & one of my responsibilities was to administer the computerized Connors Continuous Performance Test for children we were evaluating for ADHD. This also includes a written behavioural journal for parents & the primary teacher, but this computerized test - which I found downloadable for free here: https://www.millisecond.com/download/library/v7/cpt/ax_cpt/ax_cpt/ax_cpt.web - measures attention.
So, as I would be writing case notes, the Chief Psychiatrist/Captain of the unit came to me after noticing a "pattern" in my writing that included included repetitive words (e.g. and and, the the, etc.) frequent misspellings, and frequent missing words in sentences, and asked if this was something I had noticed. I told him that it was an "adult" problem that plagued me, but even though I "proofed" my writing many times before submitting anything, rarely, if ever, was I able to see these errors until much later, sometimes days later. He dragged me to the testing lab and forced me to TAKE the Connors test, which I failed, twice. "You need to consider that you have an attention deficit, as words that are "seen" in your mind are not making it to the computer screen, and your errors are literally not being "seen" in real time." Wow. I told him the story of a undergrad college professor who once caught me in the hall and said that she went to my student advisor to check if he believed I had actually written a paper I had turned in (he believed it!) - apparently she believed me to be a "dolt," - yet she gave me a D- for all the spelling errors! In any case the Captain put me on a stimulant, which resolved these issues, but I just couldn't tolerate the feeling & atomoxetine simply wasn't effective. The diagnosis, however, qualified me for free dictation software, and so it goes...
Your story is far more fascinating! Thanks for sharing!
Very nice essay. I'm curious how separable you think the decision making and reinforcement mechanisms actually are. One way I view cognitive control/attention is as internal action selection with a policy controlled through reinforcement learning and with "hyperparameters" (like discount rate) governed by neuromodulators. Aspects of the decision making algorithms can be reframed in this view: poor integration could be a failure of attention which is a form stimulus selection, and too low an evidence threshold is also a form of inappropriate discounting because it's an unwillingness to wait for more certainty. This model also accounts for the anxious case because anxiety naturally should decrease action thresholds, because it implies a threat of danger and a greater cost to continued inaction and thus a higher discount rate. I think this unified model, in which each presentation of ADHD is a type of maladaptive internal action policy, also helps explain why they all benefit from stimulants, which have a dual role of modifying reinforcement signaling as well as modulating the online algorithm. Of course, the reason why they have this dual action is because those functions are fundamentally linked.
It also predicts that timing-type ADHD would be less responsive to stimulants because, while the cerebellum is involved in sequencing and control, it's not at the level of action selection. I had a friend that had this type because he was missing part of his vermis.
Also, this doesn't diminish the value in teasing the paths apart, but seeing how they can be re-synthesized as part of a more overarching theory can give us a more precise understanding of the phenotype.
I agree with a lot of what you said. The cognitive control vs RL is certainly dissociable with the right task and model. Anne Collins had a very nice series of papers with Michael Frank using this setup and showed quite nicely that in schizophrenia, RL is intact but working memory is not. I think a task battery capable of doing various types of dissociations would be helpful for stratification and phenotyping.
Oh yeah I agree there are scenarios where you can dissociate parts of the internal action-selection loop, such as—as you say—in schizophrenia, where the fidelity of certain “actions,” like maintaining items in working memory, is degraded. My point is that by understanding that these can be unified in a single functional loop, it helps to understand the interdependence and overlapping mechanisms between them.
For example, this nicely helps explain why dopamine is involved in both RL and making cortical attractor basins more robust. Both are components of the same overarching selection process—at least in this choice of metaphor. All of these things are ultimately abstractions and choices of semantic frame, so I don’t want to make it sound like this is the unique way of thinking about it.
This was a fascinating read. Very well written, thanks!
Thank you!
‘All models are wrong some are useful’ - I would still argue it’s a useful way for people to discuss a grouping of symptoms.
It’s definitely been harmful in certain fields though (e.g. Alzheimer’s) but I feel like that’s fast changing.
'Seen this way, the three vignettes make more sense. The seven-year-old may have had noisy decision-making that improved with brain development. The adult woman may have a steep discounting of future rewards. The anxious twelve-year-old may have intact decision-making but disrupted thresholds in stressful contexts. These are testable ideas, and they point toward why one-size-fits-all categories struggle to capture the reality of ADHD'
or each person in each of these vignettes just like every other human being, could have had a wide range of issues causing their distress that the label and drugs simply masked, as is the case for every other DSM construct and psychiatric drug.
Everyone has a uniquely complex context and history, so many moving parts many of which are beyond our awareness horizon. Mainstream psychiatric constructs and drugs are horribly reductive, self limiting and cause more harm than they can ever do good.
In my wee opinion the normalisation of the mass labelling and drugging of tens of millions of human beings, children and adults alike (and now pets) is the worst era of psychiatry to date. Thats really saying something given its long and barbaric history of blunt instruments dressed up in white coats.
'that requires updating our measurement tools and our validation frameworks, but let’s not discard the gains already made' Gains? true the pharmaceutical companies have gained and, psychiatry appears to maintain its illusions on the back of this.
However, for the general public and human wellbeing there are no gains, suffering is worse year in year out and the harms done to people by these labels and drugs is there for anyone to see.
The mass uncontrolled experiments have been done, globally - there is now so much unmetabolized chemical junk being urinated out from humans subjected to these 'treatments' its changing the behaviour of marine life.
I've worked in the mental ill health industry for 25 years and I can honestly say that the thought of my own suffering or those I love, leading to a consultation with a psychiatrist, a rushed doctor or a DSM deluded psychologist absolutely terrifies me.
I appreciate your input. I’m optimistic that the future will be better, and that we will have more ways to identity and treat human suffering. Happy New Year.
Great, articulate comment. I agree with many of your thoughts, and, particularly, “people love hearing themselves described in quasi-clinical terms.” I also agree that the distractibility in mania is qualitatively different from what brings most adults in for quasi-clinical discussions of their attention span . Nassir Ghaemi has criticized the lack of validity of ADHD (and most other DSM diagnoses) in his Substack and elsewhere using many of the same lines of reasoning and studies as Dr. Halassa here, but Dr. Ghaemi always loses me when mentioning distractibility in mania as an example of a valid, more likely explanation for a significant portion of whatever entity/ies we provisionally call ADHD in adults.
I agree focusing too much research on disembodied parts can be a potential pitfall in pursuit of diagnostic validity and clarity, but I actually support testing these various hypotheses of ADHD. Not all of the hypotheses will be fruitful, and maybe none will, but one or more may help validate the different clinical forms we observe in practice, which would be a welcome finding IMO.