Maya Chen: Nathan, hey — did you eat this week or just read papers?
Dr. Nathan Hayes: Both, in approximately equal measure — which is probably relevant to today, actually.
Maya Chen: Okay I want to hand you something before we get anywhere near an explanation — a study, 1978, Levine et al., post-surgical patients. They gave some patients a sugar pill, told them it was a painkiller. Then they blocked the body's own opioid system with naloxone. And the pain relief just... stopped. Vanished. What does that do to you?
Dr. Nathan Hayes: It tells me the placebo wasn't a trick. The brain had already produced something real — endogenous opioids — and naloxone intercepted them. You can't intercept an illusion.
Maya Chen: Right — but the part that stops me is, these were post-surgical patients. Real pain, real bodies. And a pill they knew might be nothing triggered actual opioid chemistry.
Dr. Nathan Hayes: Now, the effect size matters here — we're talking roughly 30 to 40 percent pain reduction in most studies, not morphine-level analgesia. But the mechanism... the mechanism is what makes 1978 a founding moment. Before Levine, you could dismiss placebo as perception. After naloxone? You're dealing with a causal chain you can chemically interrupt.
Maya Chen: So the question this whole episode is actually turning on — can you chemically intercept a belief? And what does it mean that the answer is yes?
Dr. Nathan Hayes: That's exactly where Fabrizio Benedetti and Tor Wager land — placebo effects as psychobiological responses, the brain's prediction system triggering concrete physiological cascades. Not metaphor. Measurable cascades. But bounded — and that boundary is what I want to spend time on.
Maya Chen: And that boundary — I want to hold there for a second, because what you're describing sounds almost like the brain gets ahead of itself. Like it's not waiting to see what the substance does.
Dr. Nathan Hayes: That's actually the core of it. Here's the clean version: your brain is constantly running a forecast of your body's state, and when it predicts relief, it dispatches real chemistry to match that forecast — before the substance does anything. Think of a competitive swimmer on the starting block, handed what she's told is a legal performance supplement. Her adrenal system is already firing before she swallows it. The expectation triggers the cascade.
Maya Chen: Like a thermostat pre-heating the room.
Dr. Nathan Hayes: Exactly that. Jon Stoessl and Sarah Lidstone at UBC put Parkinson's patients in a PET scanner and showed measurable dopamine release in the striatum from expectation alone. No active drug.
Maya Chen: Wait — the striatum. That's the region Parkinson's is actually destroying.
Dr. Nathan Hayes: Precisely. The disease is defined by dopamine depletion in the striatum — and expectation partially restocked it. That's not a subjective report, that's in vivo imaging. Benedetti and Wager's framing lands here: a psychobiological response, real physiological cascades. Not mood. Chemistry.
Maya Chen: So the forecast isn't just — I mean, it's not limited to pain. It's going into the specific depleted region of a degenerative disease and... actually producing the thing that's missing. That's the part that complicates the neat story, right? It's not just dulling perception.
Dr. Nathan Hayes: Right — and it extends further. Altered immune markers have been documented in allergic conditions too, so the neurobiological footprint here runs beyond pain and motor systems. The forecast model doesn't care about the category. What we don't have yet is the full map of which forecasts produce which chemistry in which patients. That individual variability — still mostly a black box.
Maya Chen: But that map — or the edge of it, I mean — that's what I keep snagging on. Because if expectation can actually reach into the striatum of a Parkinson's patient and produce dopamine, why doesn't it reach further? Why doesn't it just... cure strep throat?
Dr. Nathan Hayes: Right — and that's not a rhetorical question, that's actually the boundary the field has been trying to draw precisely. The brain's predictive circuitry has reach into dopaminergic, opioidergic, and autonomic systems. Those are prediction-sensitive. Bacterial load? Not remotely. A sugar pill cannot lower the colony count of streptococcus.
Maya Chen: So it's not that mind-body unity is real or — wait, actually that's not even the right frame, is it. It's more like... the nervous system has a reach, and that reach has a hard edge.
Dr. Nathan Hayes: Exactly that framing. And the gradient maps cleanly onto it — pain, nausea, fatigue are all amenable to expectation because those are mediated by the very systems that are prediction-sensitive. Infection isn't. But here's where I want to be careful — misreading the gradient is genuinely dangerous. If we say 'placebos work' without specifying that boundary, we risk dismissing real symptoms as merely subjective. That's not a theoretical problem.
Maya Chen: Hold on — that's the part that worries me too.
Dr. Nathan Hayes: It should. Because the jump from 'expectation modulates your opioid system' to 'your attitude is making you sick' is a short, harmful one.
Maya Chen: Okay but — I want to poke at the immune marker piece specifically, because you mentioned allergic conditions earlier. Is that genuinely the edge of what we know, or is it a hint that the boundary might be less clean than the clean version sounds? Like, conditioning and expectancy might be doing something in immune surveillance we haven't mapped yet — I mean, that feels like a 'we don't know yet' that sometimes gets sold as 'the limit is proven.'
Dr. Nathan Hayes: That's a fair distinction and I'll hold the line here — 'we don't know yet' is not evidence the limit doesn't exist. Those are different claims. The immune marker findings in allergic conditions are real, but the effect sizes are modest and the mechanism is genuinely unsettled. Conditioning and expectancy — those are the two principal mechanisms we've identified, and their relative contributions vary by condition. We have not unified them. What we shouldn't do is treat an unmapped region as permission to assume the forecast system reaches everywhere. And — actually, the part that complicates this further comes when you look at what happens when patients know the pill is inert and still improve, because that breaks the deception assumption in a way that rewrites the whole question.
Maya Chen: Wait — that's the thing that breaks it open for me. Kaptchuk's trial, 2010, the one out of Harvard — he told patients upfront, 'these are inert pills,' and they still improved.
Dr. Nathan Hayes: Yes. Kaptchuk at the Program in Placebo Studies, Beth Israel Deaconess — the deception assumption just... gone. Patients knew. Still responded.
Maya Chen: So if honesty doesn't kill the effect, what exactly is doing the work?
Dr. Nathan Hayes: That's the uncomfortable part — because the smallest effects in open-label studies appear when positive framing is stripped out. The clinician who says 'this is inert, good luck' gets less than the one who says 'this is inert, and the ritual of taking it has real biological effects.' Same disclosure. Different outcome. So the honesty alone isn't the mechanism — the warm authority, the ritual, the framing, that's still in there.
Maya Chen: Mm — so we haven't actually escaped the white coat.
Dr. Nathan Hayes: No. And the PANACEA consortium — European consensus group, produced clinical recommendations for managing exactly this, placebo and nocebo effects in patient-provider interactions — they couldn't resolve it either. The guidance acknowledges the ritual matters. It does not tell you how to harness it ethically without it shading back into suggestion.
Maya Chen: So the field itself is — I mean, they wrote the guidelines and still left that thread hanging?
Dr. Nathan Hayes: Unresolved. Intentionally. Because — actually, the honest version is we don't know whether we're harnessing the patient's expectation or the clinician's performance. Those are different mechanisms and we've been treating them as one.
Maya Chen: That feels like the real question — not does the open-label placebo work, but whose behavior is generating the effect. And I don't think we can answer that yet.
Dr. Nathan Hayes: Which puts us somewhere genuinely uncomfortable. The naloxone result from 1978 settled 'is it real' — that's done, that's fifty years old. But 'who does it help, by how much, and can we predict it' — we're not close. The neurogenetic factors that explain why some patients show robust dopamine release in the striatum and others show nothing? Still largely uncharacterized.
Maya Chen: That's the gap I can't — I mean, I keep wanting to land somewhere, and the science won't let me. 'Placebos are real' and 'placebos will help you' are just... not the same claim.
Dr. Nathan Hayes: No. They're not. And I think that's actually where we have to leave it.
Maya Chen: Yeah. The brain's prediction system — compelling, partially mapped, individually unpredictable. That's the honest version.
Dr. Nathan Hayes: The honest version. Which is harder to sit with than 'belief heals you.'
Maya Chen: Much harder. Thanks for not letting me round it up.
Dr. Nathan Hayes: Thanks for making me say what I actually meant.
