Daniel sent us this one — he's been thinking about his own diagnosis of both ADHD and depression, which he initially thought was some unusual combo, but he's realizing it's almost stereotypical. The core question is: if most people with mental illness present with clusters of conditions rather than single clean diagnoses, how useful are these labels we're using? And what's actually going on underneath — is it biological, environmental, both? If the neurotransmitter hypothesis is right, does comorbidity just mean people have disturbances across multiple systems? And here's the forward-looking part: are we heading toward a future where these catch-all diagnostic names are viewed as anachronisms, replaced by personalized treatment based on individual symptoms and underlying causes?
This is exactly the question that keeps a lot of psychiatric researchers up at night. And the personal experience here maps onto the data almost perfectly. If you look at the numbers, among people with ADHD, somewhere between thirty and forty percent also meet criteria for a depressive disorder. Twenty-five to fifty percent for an anxiety disorder. For major depressive disorder, roughly sixty percent also have an anxiety disorder. These aren't edge cases — comorbidity is the norm. And I want to sit on that sixty percent number for a second, because it's staggering when you think about it. If you walk into any psychiatric waiting room and pick out someone with depression, the odds are better than a coin flip that they also have clinically significant anxiety. That's not two diseases coincidentally happening to the same person — that's one underlying process wearing two different name tags.
The person with one clean diagnosis is the outlier.
And that should make us suspicious of the categories themselves. The DSM-5 lists over three hundred distinct mental disorders. It's this enormous taxonomy — you could spend an afternoon just reading the table of contents. But in clinical practice, patients almost never show up with just one item from the menu. They show up with a plate of overlapping symptoms that we then slice into multiple diagnoses. It's a bit like going to a restaurant, ordering the soup and the salad and the entree, and the kitchen treating each course as if it came from a completely different meal.
Which raises the question of whether we're carving nature at the joints or just... slicing a smooth gradient into arbitrary segments.
That's exactly the phrase — "carving nature at the joints." It goes back to Plato, but in psychiatry it's become the central critique of the DSM approach. The DSM was designed for reliability — meaning two different clinicians should give the same patient the same diagnosis. And it succeeded at that, mostly. If I interview a patient and diagnose major depressive disorder, and you interview the same patient the next day, we'll probably agree. But reliability is not validity. Just because we agree on what to call something doesn't mean the category maps onto biological reality. We could both agree that someone has "hysteria" in eighteen ninety, and we'd be reliable — but we'd be reliably wrong about what's actually happening.
We've built this elaborate naming system that's consistent between doctors but might not correspond to anything real in the brain.
That's the worry. And the comorbidity data is one of the strongest pieces of evidence for it. If these were truly distinct diseases — the way tuberculosis is distinct from a broken arm — you wouldn't expect them to co-occur at rates far beyond chance. Tuberculosis and broken arms don't cluster together. But depression and anxiety do, ADHD and depression do, substance use disorders and PTSD do. The clustering itself tells you the categories are capturing surface patterns, not deep divisions.
Let's start with the biology then. If we want to understand why these conditions cluster, we have to look at what's happening in the brain. The neurotransmitter hypothesis — is that our best explanation for the overlap?
It's the most accessible one, and it's not wrong — it's just incomplete. The basic idea is that mental illness arises from disturbances in neurotransmitter systems, and because those systems are deeply interconnected, a disruption in one tends to ripple into others. Let me walk through the specific players. Dopamine is involved in reward, motivation, and attention — it's heavily implicated in ADHD, and also in the anhedonia you see in depression, that inability to feel pleasure. Serotonin is involved in mood regulation, anxiety, and impulse control — it's central to both depression and anxiety disorders. Norepinephrine handles arousal and the stress response — it shows up in ADHD and anxiety. So you've got these three systems, and they're not independent silos. They talk to each other constantly through feedback loops and shared circuitry.
If your dopamine system is dysregulated — giving you ADHD symptoms — it's not like the serotonin system is just sitting there unaffected.
The systems regulate each other. If dopamine signaling is off, serotonin pathways in the prefrontal cortex compensate in ways that can produce depressive or anxious symptoms. It's not that you have three separate problems — it's that you have one complex system that's dysregulated in multiple ways. Think of it like a thermostat in a house with multiple rooms. If the thermostat in one room malfunctions, the other rooms don't just stay at the perfect temperature. The whole system starts fighting itself — one room gets too hot, the compensation overcools another room, and suddenly you've got what looks like three separate temperature problems when it's actually one control system gone haywire.
Which makes the "here's your ADHD diagnosis, here's your depression diagnosis, here's your anxiety diagnosis" approach feel a bit like describing a car that won't start by saying "well, the engine has a problem, and also the fuel system has a problem, and also the electrical system has a problem" — when actually it's one interconnected failure.
This is where the monoamine hypothesis — the idea that depression is just low serotonin, or ADHD is just low dopamine — starts to break down. That model has been dominant since the nineteen fifties, and it gave us useful medications, but it's increasingly clear that it's far too simple. The brain isn't a soup of individual chemicals you can top up independently. You can't just add more dopamine like you'd add more washer fluid to your car.
What's the alternative framework?
This is where things get interesting. In two thousand nine, the National Institute of Mental Health launched something called the Research Domain Criteria framework — RDoC for short. Major updates in twenty eighteen and again in twenty twenty-four. And what RDoC does is essentially abandon DSM categories entirely. Instead of asking "does this patient meet the checklist for major depressive disorder," it maps mental illness onto dimensional constructs — things like reward responsiveness, cognitive control, threat sensitivity, social processing. These are supposed to reflect actual brain systems that we can measure. And this was a pretty radical move — the NIMH basically said, we're not going to fund research that just uses DSM categories as endpoints anymore. We want studies that look at these underlying dimensions.
Instead of "you have depression," it's "your reward responsiveness is low, your threat sensitivity is elevated, and your cognitive control is moderately impaired.
That's the idea. And here's the key insight for comorbidity: the same underlying dimension can manifest as different symptoms depending on context. Take low reward responsiveness. In one person, that might show up as anhedonia — the inability to feel pleasure, which gets labeled as depression. In another person, it might show up as lack of motivation and difficulty sustaining effort toward goals, which gets labeled as ADHD. In a third, it might manifest as social withdrawal, which gets labeled as social anxiety. Same underlying dimension, three different diagnostic labels. And here's the thing — a single person can show all three manifestations at different times or in different contexts. So they walk out with three diagnoses, but they might really have one core issue.
The labels are almost like... different camera angles on the same underlying terrain.
That's a useful way to think about it. And this explains why comorbidity is so common. We're not seeing multiple diseases — we're seeing multiple expressions of shared underlying dimensions. The DSM categories are describing surface patterns, not deep causes. It's like we've named the shadows on the cave wall without ever turning around to look at what's casting them.
There's a practical problem here, right? Insurance companies don't reimburse for "low reward responsiveness." The FDA doesn't approve drugs for "elevated threat sensitivity." The whole system is built around these categorical labels.
This is the tension at the heart of modern psychiatry. The RDoC framework is more biologically grounded, but it's less clinically practical. You can't bill an insurance company for a dimensional profile. You can't run a clinical trial where the inclusion criterion is "cognitive control in the bottom quartile." The DSM categories are flawed, but they're the scaffolding the entire mental health system is built on. You pull them out, and suddenly you can't get reimbursed, you can't prescribe, you can't run trials. It's a genuine bind.
We're stuck with labels we know don't map onto biology because the alternative requires rebuilding the entire infrastructure of mental health care.
For now, yes. But there's movement. The DSM-6, which is expected around twenty thirty to twenty thirty-five, is likely to incorporate dimensional elements alongside the categorical diagnoses. A hybrid model. Keep the labels for communication and billing, but add dimensional assessments for treatment planning. So you'd still get "major depressive disorder" on your chart, but your clinician would also rate you on reward responsiveness, threat sensitivity, cognitive control — and those ratings would guide treatment decisions.
That feels like the musical equivalent of... keeping the album format while acknowledging everyone actually listens to playlists.
The album still has branding value and organizational logic, but the real unit of consumption is the individual track. And I think that's genuinely where we're headed — the diagnosis is the album, but the symptoms are the tracks, and treatment is going to be built around playlists.
We've got this emerging picture where the diagnostic labels are surface descriptions of deeper biological dimensions. But that raises the next question: what causes those dimensions to go wrong in the first place? Is it biology, environment, or both?
This is where we need to shift from the neurotransmitter level to the broader question of etiology. And the short answer is: it's both, interacting in ways we're only starting to understand. Let me give you the numbers from twin studies. Heritability estimates for ADHD are around seventy-five percent — that's quite high. For major depression, it's around forty percent. For generalized anxiety disorder, around thirty percent. So there's clearly a genetic component, but it's not destiny. Even with identical twins, if one develops major depression, the other doesn't always — the concordance rate is about fifty percent, not a hundred. And that gap between the heritability and the concordance — between forty percent and fifty percent — that gap is where the environment lives.
Genes load the gun, but something in the environment pulls the trigger.
That's the diathesis-stress model, and it's been the dominant framework for decades. You have genetic predispositions that create vulnerabilities, and then environmental stressors — trauma, chronic stress, social isolation, poverty — activate or amplify those vulnerabilities. This explains why comorbidity often emerges after major life stressors. A person with a genetic vulnerability to both ADHD and depression might function fine until they hit a period of chronic stress, and then both conditions manifest simultaneously. It's not that the stress caused two separate diseases — it's that the stress exposed a shared vulnerability that had been compensated for.
Which would explain why someone can make it to their thirties before getting diagnosed — the vulnerability was always there, but it took the right environmental trigger to push it into clinical territory.
And there's another layer that's gotten enormous research attention in the last five years: inflammation. Chronic low-grade inflammation, measured by markers like C-reactive protein, interleukin-six, and TNF-alpha, is consistently elevated in both depression and anxiety. This isn't just correlation — there's evidence that inflammation can actually cause depressive symptoms by disrupting neurotransmitter synthesis and neural plasticity. There was a really striking study where they gave healthy volunteers an inflammatory challenge — basically a mild immune activation — and within hours, they started showing depressive symptoms: low mood, social withdrawal, fatigue. The inflammation came first, then the mood changes.
Your immune system and your mental health are more connected than we thought.
And this might be another shared pathway underlying comorbidity. If chronic inflammation disrupts multiple neurotransmitter systems simultaneously, you'd expect to see clusters of symptoms across traditional diagnostic boundaries. You wouldn't get just depression or just anxiety — you'd get both, because the inflammatory signal is hitting everything at once. It's like flooding a basement — you don't get to choose which boxes get wet.
Which brings us to the gut-brain axis, which I know you've been reading about.
I have, and this is one of those areas where the research from twenty twenty-two to twenty twenty-four has been surprising. Approximately ninety percent of the body's serotonin is produced in the gut — not in the brain. Enterochromaffin cells in the intestinal lining synthesize serotonin, and the gut microbiome influences that production. When the microbiome is disrupted — dysbiosis — you see changes in serotonin signaling that can affect mood, anxiety, and even attention. There was a fascinating study where they transplanted gut bacteria from depressed humans into germ-free mice, and the mice started showing depressive-like behaviors. The bacteria alone were sufficient to shift the behavioral profile.
The bacteria in your intestines are essentially running a side operation in neurotransmitter production.
And studies have found links between gut dysbiosis and both ADHD and depression. The mechanisms aren't fully worked out yet — this is still emerging science — but the vagus nerve provides a direct communication line between the gut and the brain. Inflammatory signals from the gut can travel up that nerve and affect brain function. It's not separate systems — it's one integrated network. Your gut bacteria are, in a very real sense, part of your mental health infrastructure.
We've gone from "mental illness is a chemical imbalance in the brain" to "mental illness is a complex systems failure involving neurotransmitters, immune function, gut bacteria, genetic vulnerabilities, and environmental stressors.
Which is simultaneously more accurate and much harder to explain in a thirty-second pharma commercial.
The glockenspiel of corporate approachability doesn't really accommodate gut-brain axis explanations.
It does not. But this complexity is actually good news for treatment, because it means there are multiple entry points for intervention. If depression were just low serotonin, you'd only have one target. But if it's a systems-level dysregulation involving inflammation, microbiome, neurotransmitters, and stress responses, you can intervene at multiple levels — medication, diet, exercise, sleep, therapy — and they can all be therapeutic. You're not just trying to fix one thing — you're trying to tune a whole system, and you can start from multiple places.
Let me push on the personalized medicine angle, because this is where the prompt was really pointing. Are we actually moving toward a world where "ADHD" and "depression" are viewed as anachronisms?
There's a comparison I find useful here. Think about how oncology has evolved. We used to talk about "breast cancer" as if it were one disease. Now we talk about HER2-positive, ER-positive breast cancer with specific mutations — and treatment is tailored to those molecular markers, not just the anatomical location. Psychiatry may be on a similar trajectory, but it's decades behind. The question isn't "do you have depression" — it's "what subtype of depression do you have, and what's driving it biologically?
Because the brain is harder to biopsy than a tumor.
That's part of it. But also because we've been slower to develop the measurement tools. However, there was a landmark study in twenty twenty-four from the Stanford Center for Precision Mental Health — Williams and colleagues, published in Nature Medicine — that used machine learning on fMRI and genetic data to identify six distinct biotypes of depression. And these biotypes responded differently to different treatments. One biotype responded well to SSRIs, another didn't but responded to behavioral therapy, a third showed a distinct pattern of brain connectivity that predicted better outcomes with transcranial magnetic stimulation. That's the kind of finding that makes you rethink everything about how we match patients to treatments.
Six different depressions, each with its own treatment profile.
And the implication is that a lot of what we call "treatment-resistant depression" might actually be "wrong-biotype-for-this-treatment depression." The label is the same, but the underlying biology is different. It's like giving the same antibiotic to six different bacterial infections and calling the patient "treatment-resistant" when five of them don't respond. The problem isn't the patient — it's the matching.
Which means the DSM category of "major depressive disorder" is about as precise as diagnosing someone with "chest pain" and then wondering why the same treatment doesn't work for everyone.
That's the critique in a nutshell. And there are movements pushing in this direction. The Depression and Bipolar Support Alliance has been advocating for symptom-cluster-based approaches. There's a book called "ADHD two point oh" that emphasizes dimensional understanding over categorical labeling. The ground is shifting. But I want to be honest about the timeline — we're not going to wake up in twenty thirty and find the DSM replaced by a brain-scan app. This is a decades-long transition.
Let me play devil's advocate for a moment. There's a reason the DSM categories persist, and it's not just institutional inertia. When someone receives a diagnosis, it can be helpful to have a name for what they're experiencing. "I have ADHD" is a useful identity around which people organize their understanding of themselves and connect with communities of others with similar experiences.
I don't disagree. The labels have social and psychological functions that go beyond their scientific validity. And I think the future is not "abolish all labels" — it's "use labels as shorthand while understanding their limitations." The problem isn't that we have names for clusters of symptoms. The problem is when we mistake the name for the explanation. When "I have depression" becomes the end of the conversation instead of the beginning.
The map is not the territory.
And right now, psychiatry is operating with maps that were drawn before we had good satellite imagery of the terrain. The RDoC framework and the biotype research are giving us better imagery. But you still need a map you can fold up and put in your pocket. The question is how to update the pocket map without pretending it's a perfect representation.
Where does this leave us practically? If someone's listening and they have multiple diagnoses — or they're a clinician treating patients with multiple diagnoses — what do they do with this information?
Let me offer three concrete takeaways. First, if you have multiple diagnoses, stop thinking of them as separate problems that each need their own solution. Instead, look for common underlying patterns. Is there a reward system issue running through everything? A threat sensitivity issue? A cognitive control issue? This reframing can guide more targeted treatment. For example, if low reward responsiveness is the common thread, behavioral activation — which specifically targets the reward system by scheduling pleasurable activities and tracking their impact — might be more effective than a medication that primarily targets serotonin. You're not treating ADHD and depression separately — you're treating the reward system dysfunction that's expressing itself as both.
You're treating the dimension, not the label.
Second takeaway — for clinicians especially, but also for patients advocating for themselves — consider transdiagnostic treatment approaches. These are interventions that work across diagnostic boundaries. CBT protocols for emotional regulation, for instance, have been shown to be effective for depression, anxiety, and ADHD simultaneously. Exercise and sleep hygiene improve dopamine and serotonin systems at the same time. You don't need to treat each diagnosis separately if you can target the shared underlying mechanisms. A good exercise routine might do more for someone with ADHD-depression comorbidity than two separate medications with conflicting side effect profiles.
What's the third one?
Stay informed about where the field is heading. The DSM-6, expected around twenty thirty to twenty thirty-five, is likely to incorporate dimensional elements. Patients can ask their providers about symptom-based rather than label-based treatment planning now. A good clinician should be able to say "here's your diagnosis for insurance purposes, but here's what we're actually going to treat and why." If you're a patient, you can ask: what dimension are we targeting with this treatment? What's the underlying process we're trying to shift?
If they can't?
Then you're working with someone who's practicing fifteen-year-old psychiatry. Which doesn't mean they're bad at their job — the DSM-5 is still the standard of care — but it does mean there's a richer conversation available if you find a provider who's keeping up with the research. And increasingly, patients are bringing this knowledge into appointments themselves. They've read about the RDoC framework, they've heard about biotypes, and they're asking more sophisticated questions than clinicians sometimes expect.
There's something almost Kafkaesque about a system where the official diagnosis is acknowledged to be scientifically questionable, but you still need it to access care. It's like needing a passport from a country that doesn't actually exist.
That's not far off. And one of the most frustrating consequences is that clinical trials for psychiatric medications typically exclude patients with comorbidities. They want "pure" major depressive disorder or "pure" ADHD. But those patients are the minority in real clinical practice. So the evidence base for how to treat the very patients who most need help is systematically thin. We're generating evidence on the cleanest, simplest cases and then extrapolating to the messy, complicated cases that actually fill waiting rooms.
We're running trials on atypical patients and then applying the results to typical patients.
And that's a structural problem that's going to take years to fix. The FDA approval process is tied to specific diagnoses. Pharmaceutical companies design trials around those diagnoses because that's what the FDA requires. Changing this means changing regulatory frameworks, not just clinical practice. You'd need the FDA to start accepting dimensional inclusion criteria, and that's a heavy bureaucratic lift.
Which makes the personalized medicine future feel both inevitable and perpetually just over the horizon.
I think that's the right way to characterize it. The trajectory is clear — we're moving toward dimensional, biologically-grounded, personalized psychiatry. But the pace is going to be uneven, and we're going to live with hybrid models for a long time. The DSM categories aren't useless — they're just incomplete. They're useful shorthand for communication, and they provide a common language that patients, clinicians, researchers, and insurers can all use. The problem is when we mistake that shorthand for ground truth. It's like using the word "fever" — it's useful to say "I have a fever," but nobody thinks "fever" is a disease. It's a sign that something else is going on.
The most important shift isn't "abolish the labels" — it's "stop thinking the labels are the explanation.
The question to ask isn't "what disorder do you have" — it's "what systems are dysregulated, and what can we do about it?" That's a fundamentally different orientation, and it's one that actually aligns better with how the brain actually works. The brain doesn't organize itself by DSM chapter. It organizes itself by circuits and systems that cross all our diagnostic lines.
Which brings me to a question I want to sit with. If we eventually get to a point where we have robust biotypes — where we can scan someone's brain, run their bloodwork, sequence their microbiome, and say "here's your specific dysregulation profile, here's the treatment that matches it" — will we still want the old labels? Or will they feel as archaic as the four humors?
I think we'll keep them as colloquial shorthand, the way we still say "I'm feeling melancholic" even though nobody believes in an excess of black bile anymore. The labels capture something real about the lived experience. "Depression" describes a recognizable pattern of suffering, even if the underlying biology is heterogeneous. The danger isn't the word — it's mistaking the word for the mechanism. You can say "I'm depressed" and that's perfectly valid as a description of your experience. The problem is when your clinician hears "depressed" and stops thinking about what's actually driving it.
Like saying "I have a headache" — it communicates the experience, but nobody thinks "headache" is a single disease with a single cause.
That's the analogy. And right now, psychiatry is in the awkward position of having headache-level diagnostic precision while dealing with conditions that are far more complex than headaches. The gap between the granularity of the labels and the granularity of the biology is enormous, and that gap is where most of the confusion about comorbidity lives. We're using a sledgehammer vocabulary to describe a Swiss watch.
The fact that comorbidity is the norm isn't a sign that people are unusually complicated — it's a sign that our diagnostic system is unusually crude.
That's the thesis. And it's increasingly well-supported by the evidence. The RDoC framework, the biotype research, the inflammation and microbiome data — they all point in the same direction. Mental illness isn't a collection of distinct diseases that sometimes happen to co-occur. It's a set of interconnected biological systems that can go wrong in overlapping ways, producing clusters of symptoms that we've historically sliced into separate diagnoses. The slicing is the artifact — not the clusters themselves.
Which means the person with ADHD and depression and anxiety isn't "treatment-resistant" or "complicated" — they're just a clearer reflection of how these systems actually work.
And that reframing matters, because "treatment-resistant" and "complicated" are labels that can make people feel broken in a way that's demoralizing. If instead you understand that you're dealing with interconnected systems that need a coordinated approach, it changes how you think about recovery. It's not that you're harder to fix — it's that the fixing needs to be more thoughtful. And honestly, that's a much more respectful way to think about patients. You're not a diagnostic outlier — you're actually the typical case that our system is poorly designed to handle.
I want to circle back to one thing before we wrap up. The neurotransmitter hypothesis — you said it's not wrong, just incomplete. What would a more complete model look like?
A more complete model would integrate multiple levels. Neurotransmitters, yes — but also neural circuits, the specific brain networks involved in things like the default mode network and the salience network. Also inflammation and immune function. Also the gut microbiome. Also hormonal systems like the HPA axis that governs stress responses. Also environmental factors — trauma history, social support, socioeconomic status. And critically, it would understand all of these as interacting dynamically, not as independent contributors. You can't just add up the neurotransmitter score and the inflammation score and get a depression score. These things shape each other in real time. It's a systems biology approach, and it's where the field is heading — though I'll be the first to admit we're not there yet.
The future of psychiatry looks less like a diagnostic manual and more like a dashboard.
A dashboard with multiple gauges — neurotransmitter function, inflammatory markers, circuit activity, microbiome composition, stress hormone levels — and treatment that adjusts based on which gauges are in the red, rather than which label is on the chart. And the treatment itself might be multi-modal — maybe you need an anti-inflammatory intervention alongside a dopamine-targeting medication alongside a behavioral approach. You're tuning the whole system, not just one dial.
The dashboard of the self.
The dashboard of the self. Which sounds like the title of a very earnest self-help book, but it's actually where the science is pointing. And I think the most exciting thing is that this approach makes room for interventions that don't look like traditional psychiatry at all — dietary changes that target the microbiome, exercise protocols that boost neuroplasticity, anti-inflammatory strategies. It expands the toolkit enormously.
Now: Hilbert's daily fun fact.
Hilbert: In the seventeen eighties, a single surviving shipping manifest from a Hanseatic League trading post near what is now Tajikistan reveals that merchants were required to pay port fees not in coin, but in specific quantities of saffron — the weight of which had to be verified by a guild-appointed "saffron-weigher" whose scales were ceremonially blessed each market day.
A saffron-weigher.
Ceremonially blessed scales. I have to say, that's a job title that really commits to the bit. You're not just weighing spices — you're doing it with divine authority.
I do think the next decade is going to be fascinating to watch. We're not going to abandon the DSM categories entirely — the institutional inertia is too strong, and they do serve real functions. But we're going to layer dimensional assessments on top of them. You'll get your DSM code for billing, and then a dimensional profile for treatment planning. The labels won't disappear, but they'll stop being the final word.
That's probably the right outcome. The labels capture something real about patterns of suffering, and they give people a language for their experience. The problem has never been the names — it's been treating the names as if they were explanations. A diagnosis should be the beginning of the conversation, not the conclusion. The question isn't "what do you have" — it's "what's driving this, and what are we going to do about it?
The most important shift is from asking "what disorder do you have" to asking "what systems are dysregulated, and what can we do about it." That's a more honest question, and it's one the science is increasingly equipped to answer.
It's also a more hopeful question. Because systems can be regulated. Dimensions can shift. The dashboard can change. And that's a much more useful way to think about mental health than a collection of permanent labels. You're not stuck with a diagnosis — you're managing a dynamic system that can improve. That's not just scientifically more accurate — it's clinically more empowering.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop. You can find us at myweirdprompts dot com.
If you enjoyed this, leave us a review — it helps other people find the show. We'll be back soon.
