Tess Hollis: Felix, hey — how's the back, by the way?
Felix Ortiz: Oh, yeah — better, actually. I ended up just, like, taking ibuprofen for two days and then it kind of sorted itself out, which — I don't know, I keep thinking about whether it was the ibuprofen or whether I just expected the ibuprofen to work and my brain filled in the rest.
Tess Hollis: Wait — that's actually exactly the thing.
Felix Ortiz: Right? I know, I caught myself doing it. Like, is the pill working, or is the — the act of taking the pill working?
Tess Hollis: That is the exact question we're pulling on today. Because there's a finding that keeps stopping me cold — in clinical trials for irritable bowel syndrome, researchers gave patients a pill, told them upfront it was a placebo, totally inert, nothing in it — and patients still got measurably better.
Felix Ortiz: Hold on — they were told?
Tess Hollis: Explicitly. Label said placebo. And the relief still came. So the thing I want to figure out — if no one's being tricked, what is actually doing the work?
Felix Ortiz: Yeah, and that's — okay, that breaks the story I had in my head, because I always figured the deception was load-bearing. Like, the trick IS the mechanism. But if the trick is gone and it still works, then — wait, what even is the mechanism?
Tess Hollis: That's the question. And the answer is weirder and more biological than I expected going in.
Felix Ortiz: Okay — I'm already leaning forward, let's go.
Tess Hollis: The biological part is strange. Because Jon Levine — 1978 — he didn't just observe placebo pain relief. He gave patients a placebo after dental surgery, confirmed the relief was there, and then injected naloxone.
Felix Ortiz: And naloxone is an opioid antagonist — it blocks opioid receptors.
Tess Hollis: Right. And the placebo relief stopped.
Felix Ortiz: Wait — the relief just... reversed?
Tess Hollis: Reversed. Which means the brain had literally been releasing its own opioids. The placebo triggered the endogenous opioid system — not metaphorically. Chemically.
Felix Ortiz: Okay that — yeah, that reframes everything I thought I understood. Because we're not talking about belief making you feel better, we're talking about belief making your brain dispense actual opioids through the μ-opioid receptor system. Like — the brain has a pharmacy and expectation is the prescription.
Tess Hollis: And the circuit is specific. Expectation fires in the prefrontal cortex, that signal runs down through the periaqueductal gray in the brainstem — the PAG — and from there, descending inhibition goes to the spinal cord and quiets the pain signal before it fully reaches consciousness.
Felix Ortiz: So the pain doesn't get suppressed after it arrives — it gets intercepted on the way up.
Tess Hollis: Exactly that. And then there's 2025 — Neyama et al. in Science Advances, using chemogenetics, actually confirmed the specific neurons. μ-opioid receptor-positive neurons in the medial prefrontal cortex directly activate the mPFC–vlPAG descending circuit. In neuropathic pain models. That's not correlational imaging anymore, that's circuit-level proof.
Felix Ortiz: Chemogenetics meaning they could switch those specific neurons on and off.
Tess Hollis: Yes. Which closes the loop Levine opened in 1978.
Felix Ortiz: Okay — and here's the part that actually made me stop when I read it. Because the μ-opioid receptor system isn't even the whole pharmacy. Endocannabinoid and dopaminergic systems also get recruited during placebo analgesia. So the brain isn't pulling one lever, it's — wait, no — it's running like a coordinated internal drug response based purely on what it expects to happen.
Tess Hollis: Which is exactly when you have to ask — if the mechanism is that robust, why does it only work for some things and not others?
Felix Ortiz: Yeah, and that's — that's the thing that keeps snagging me. Because the framing we always hear is 'subjective versus objective' — placebo crushes it for subjective symptoms, nothing for objective ones. But I'm not sure that label is doing real work.
Tess Hollis: That's exactly the thing being stepped around. Calling pain 'subjective' implies it's somehow less real — and that's wrong.
Felix Ortiz: Right, because μ-opioid activation is a biological change. It's not — I mean, that's not a feeling, that's a receptor firing.
Tess Hollis: So the line isn't really subjective versus objective. The line is: does this pathology require something the nervous system can supply on its own, or does it need something external that no amount of descending inhibition can substitute for? A bacterial infection needs antibiotics. A femur fracture needs the bone re-set. The periaqueductal gray cannot fix either of those.
Felix Ortiz: But it CAN tell your brain the pain is lower — even while the fracture is still there.
Tess Hollis: Which is the hospital waiting room scenario. Someone's told their painkiller is working. Prefrontal cortex fires, the descending circuit activates, measurable opioid signaling drops their experienced pain. And the femur is still broken.
Felix Ortiz: So the brain modulated perception — completely real — but the pathology underneath is untouched. Those are two different problems.
Tess Hollis: And here's where Alzheimer's becomes important. Patients with significant prefrontal impairment show reduced or absent placebo responses. Which is actually — wait, that's direct evidence. If the top-down circuit is damaged, the effect disappears. That tells you the cortical engagement isn't decorative, it's load-bearing.
Felix Ortiz: Hold on — absent? Not just weaker?
Tess Hollis: Reduced to absent, yeah. Which means expectancy needs intact prefrontal function to actually translate into opioid release. No cortex, no prescription.
Felix Ortiz: The boundary isn't arbitrary — it's structural. The mechanism defines its own ceiling. It can do everything the nervous system is positioned to do, and it literally cannot reach past that. That's actually — I mean, that's clarifying rather than limiting, isn't it?
Tess Hollis: It is. It defines the domain without shrinking what happens inside it. And honestly, the part that complicates even that — the open-label IBS results, whether the effect holds over weeks, what's actually doing the lifting when deception is off the table — genuinely unresolved.
Felix Ortiz: Yeah — and I don't think we've actually answered what the active ingredient is when the trick is gone. Is it expectancy, is it the ritual, is it the relationship? That part I want to pull apart.
Tess Hollis: And that question — what's doing the work when deception is off the table — I think that's actually where we have to be really careful. Because there are at least three separate candidates, and the research keeps treating them like they're the same thing.
Felix Ortiz: Right, yeah — expectancy, conditioning, ritual. They all show up together in an open-label trial, so how do you even pull them apart?
Tess Hollis: Okay, so classical conditioning is actually the cleanest one mechanistically. The idea is: if you've received real treatment in a clinical context repeatedly, your body learns to respond to the context itself. The setting becomes the cue. So you walk into that room, you take the pill — even knowing it's inert — and the conditioned response fires anyway.
Felix Ortiz: So it's not even about what you believe in that moment — your nervous system already has the habit.
Tess Hollis: Which would explain why the transparency doesn't kill it. You can know the pill is inert and still trigger the response, if the conditioning is already there.
Felix Ortiz: Okay but wait — wait, no. If conditioning is the engine, then open-label placebo should work better in people who've had lots of prior treatment exposure. And I don't think the IBS trials actually screened for that or tested it. So we might be — I mean, we're kind of reverse-engineering a mechanism onto a result.
Tess Hollis: That's the conflation problem exactly. And it connects to something that trips up the whole field — Hróbjartsson and Gøtzsche, 2001 meta-analysis, basically said: a lot of what we're calling placebo response is just natural remission. Regression to the mean. People get better anyway.
Felix Ortiz: Oh — that's a brutal reframe.
Tess Hollis: It is. Because the placebo response — everything that improves in the placebo arm of a trial — includes spontaneous recovery. It includes regression to the mean. It includes the therapeutic relationship, the attention, the ritual. The true placebo effect, the expectancy-driven neurobiological piece, is buried inside all of that. And those two things get conflated constantly.
Felix Ortiz: So when the IBS trials show improvement under open-label placebo — we don't actually know how much of that is the expectancy activating a real neurobiological response, versus people just... getting better on their own schedule, inside a study that made them feel cared for.
Tess Hollis: Right. The psychosocial context effect is real — twenty minutes with an attentive clinician is not nothing. But it's not the same mechanism as the μ-opioid receptor system firing. And if we label both 'open-label placebo works,' we've actually explained nothing.
Felix Ortiz: Which brings durability back. Because if it's conditioning, it should persist — maybe even strengthen. But if it's riding the novelty of the ritual or the warmth of the clinical attention, then — I mean, does it habituate? Like, month four, does the periaqueductal gray still fire when you've seen behind the curtain a hundred times?
Tess Hollis: And that's what the research genuinely cannot tell us yet.
Felix Ortiz: Yeah — they assert durability, but real-world repeated dosing data is just... thin.
Tess Hollis: So here's what that implies — and this is the uncomfortable version — if the effect depends on the ritual staying novel, or on sustained attention from a clinician, we're not prescribing a treatment. We're prescribing theater. And theater degrades when you see the seams.
Felix Ortiz: Which means the question of what open-label placebo actually is — mechanism, or managed context — isn't just academic. It changes whether you could ever scale it without it falling apart.
Tess Hollis: And that's the thing I keep turning over. Because what we actually learned isn't just that the brain can run its own pharmacy — it's that the pharmacist might be the relationship itself. The ritual. The act of being attended to. And we have almost no science on whether you can bottle that.
Felix Ortiz: Yeah — and the nocebo data makes that even stranger, because it's the same circuit running backwards. Negative expectations, measurable harm. Analogous neurobiological pathways. Which means the system isn't just a pain-off switch — it runs in both directions based on what the clinical encounter communicates.
Tess Hollis: Wait — so every interaction a clinician has with a patient is either activating the descending inhibitory system or the nocebo equivalent. The words, the tone, the ritual. All of it is pharmacologically live.
Felix Ortiz: Which is — I mean, that's not a small implication for how medicine is practiced. But I genuinely can't settle whether open-label placebo is a deployable clinical tool yet, or whether we're just staring at evidence that the encounter itself is the dose. I'm caught between both and I don't think they're the same thing.
Tess Hollis: They're not. And I can't settle it either. The Hróbjartsson and Gøtzsche critique is still sitting there — we might be calling natural remission a placebo response and then building an ethics framework on top of a measurement problem. That part isn't resolved.
Felix Ortiz: Good place to stop, maybe. While it's still genuinely open.
