📄 medicine

A Two-Threshold Model of Psilocybin Response Explains Variability in Psychedelic-Assisted Therapy for Treatment-Resistant Depression

This paper proposes and validates a two-threshold model explaining variability in psilocybin-assisted therapy for treatment-resistant depression, positing that while a universal pharmacological threshold (~18 mg) triggers altered consciousness, the critical transition to therapeutic benefit depends on individual psychological factors rather than dose escalation.

Original authors: Jesús Isaac Castillo-Sánchez

Published 2026-09-09

📖 7 min read🧠 Deep dive

CC BY 4.0

Original authors: Jesús Isaac Castillo-Sánchez

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Depression that refuses to go away, even after trying many different medications, is a profound and isolating experience for millions of people. For decades, the standard approach to treating such stubborn cases has been to adjust the chemistry of the brain, usually by increasing the dose of a drug or switching to a different one. In recent years, a new tool has emerged: psilocybin, the active compound in "magic mushrooms," which has shown remarkable promise in helping people with treatment-resistant depression. However, a major puzzle has remained. In a massive, rigorous clinical trial involving nearly six hundred patients, a single dose of the drug helped only about two out of every five people. The other three out of five saw no significant improvement, despite receiving the exact same amount of medicine, following the same medical protocol, and having their bodies process the drug in nearly identical ways. If the drug were simply a chemical key that fits a lock, everyone with the same key should have opened the same door. The fact that they did not suggests that the mechanism of healing is far more complex than a simple chemical reaction.

A new study proposes a solution to this mystery by suggesting that the human response to psilocybin is not a smooth, gradual curve, but rather a system with two distinct gates or thresholds. The research, led by Jesús Isaac Castillo-Sánchez, argues that the drug works in two very different stages. The first stage is a biological switch that flips when a specific amount of the drug is present in the body. This switch turns on an altered state of consciousness, a shift in how the brain perceives reality. The second stage is not about the drug at all; it is about the person taking it. This stage determines whether the altered state leads to a deep, life-changing insight that actually heals the depression. The study suggests that the reason so many people in the recent trial did not get better is not because the dose was too low, but because the second gate remained closed, blocked by the patient's own psychological history and preparation.

To understand how this works, imagine the brain as a landscape with two hills. The first hill represents the transition from ordinary thinking to an altered state. The researchers found that this hill has a very specific height that corresponds to a dose of about 18 milligrams of psilocybin. Below this amount, the drug does little to change the mind. Once the dose crosses this line, the brain flips into a new mode. This is not a guess; the researchers analyzed data from hundreds of people across many different studies and found that the variability in how people felt peaked right at this 18-milligram mark. In physics, this kind of peak in variability is a signature of a system undergoing a sudden, dramatic shift, much like water turning into steam. The study confirms that this first shift happens reliably in people with severe depression, just as it does in healthy volunteers, meaning the biological switch works the same way regardless of the illness.

However, crossing the first hill is only the beginning. The second hill represents the transition from simply feeling "altered" to having a profound, mystical experience that can rewire a person's outlook on life. This second threshold is not determined by how much drug is in the blood. Instead, it is governed by the person's internal state: their mindset, their environment, and how well they were prepared for the experience. The researchers call this internal state a "parameter of flexibility." If a person is rigid, fearful, or burdened by a long history of depression, they may cross the first hill but get stuck before the second one. They will feel the drug working, but they will not reach the transformative insight needed for healing. The study found that the intensity of the mystical experience, not the size of the dose, was what predicted whether a patient would recover.

The evidence for this two-step model comes from a wide range of sources. The researchers looked at blood tests from patients and found that the amount of drug circulating in their bodies varied very little from person to person. Yet, the way those patients felt and reacted varied wildly. This disconnect proves that the difference in outcomes is not caused by the body absorbing the drug differently. Instead, the variation comes from the brain's reaction to the drug. The study also examined historical data from decades ago, when LSD was used to treat alcoholism. In those old trials, the results were surprisingly consistent across different countries and protocols, with almost no variation in success rates. The new model explains this: because the doses used in those old trials were so high, every single patient crossed the first hill. The only thing that varied was the second hill, which depends on the person, not the drug. Since the drug was guaranteed to work on the first level, the results were uniform.

This framework changes how we should think about treating depression with psychedelics. The recent trial that showed a 39% success rate used a dose of 25 milligrams. According to the model, this dose is well above the first threshold, meaning every patient in that trial successfully entered the altered state. The fact that 61% did not recover suggests that the treatment failed not because the drug was too weak, but because the preparation was insufficient to help them cross the second threshold. The study points out that the length of time a patient had been suffering from depression was the strongest predictor of whether they would fail. People who had been depressed for many years seemed to have a harder time reaching that second level of insight, likely because their minds had become more rigid over time.

The researchers propose a clear path forward based on these findings. They suggest that simply giving more of the drug to patients who did not respond is unlikely to help, because the problem is not the dose. Instead, the focus should shift to the psychological preparation. The study predicts that spending more time in therapy before the session, specifically working to increase a patient's mental flexibility and openness, would be far more effective than increasing the dose. In fact, the model suggests that a patient with intensive preparation might get better results from a standard dose than a patient with poor preparation would get from a higher dose. This is a testable idea: if researchers compare a group that gets extra preparation time against a group that gets a slightly higher dose, the model predicts the group with more preparation will heal more often.

The implications of this work are significant for the future of mental health care. It moves the conversation away from a purely chemical view of depression, where the goal is to find the perfect pill, toward a more holistic view where the medicine is just one part of a larger process. The drug acts as a key that unlocks the door to a new state of mind, but the person must be ready to walk through it. If the door is locked from the inside by years of trauma or a lack of preparation, the key alone cannot open it. By recognizing that there are two distinct thresholds—one biological and one psychological—clinicians can better understand why some people heal and others do not. The study does not claim to have solved the problem of treatment-resistant depression, but it offers a clear map of the terrain, showing that the path to recovery may depend less on the size of the dose and more on the depth of the preparation.

Technical Summary: A Two-Threshold Model of Psilocybin Response

Problem Statement
The field of psychedelic-assisted therapy faces a critical unexplained heterogeneity in treatment outcomes. In the COMP006 Phase 3 trial for treatment-resistant depression (TRD), a single 25 mg dose of psilocybin produced a clinically meaningful antidepressant response in only 39% of patients, while 61% did not respond. This lack of response occurred despite identical dosing, protocols, and pharmacological exposure. Existing pharmacological frameworks, which model dose-response as a continuous sigmoidal function (e.g., Hill or Emax equations), successfully describe population means but fail to account for the structure of inter-subject variance. Furthermore, pharmacokinetic variability (20–26%) is insufficient to explain the 61% non-response rate, and the strongest predictor of non-response identified in the trial—episode duration—lacks a mechanistic explanation within current models.

Methodology
The study proposes a "Two-Threshold Model" grounded in statistical physics and catastrophe theory, validated through a meta-analytic approach.

Data Sources: The analysis utilized the Altered States Database (ASDB) comprising over 700 subjects across 14 dose conditions and >20 studies, alongside aggregate data from the COMP001 (N=248) and COMP006 (N=581) trials. Historical meta-analytic data from six LSD trials for alcoholism (N=536) were also included.
Analytical Frameworks:
- Landau Theory: Applied to model the first threshold ( $D_{c1}$ ) as a second-order phase transition. The model posits that inter-subject variance should peak at a critical dose due to critical fluctuations between ordinary ( $\theta^*_0$ ) and altered ( $\theta^*_1$ ) consciousness states.
- Cusp Catastrophe Theory: Applied to model the second threshold ( $D_{c2}$ ) as a transition to a transformative mystical state ( $\theta^*_2$ ). This transition is governed by an individual asymmetry parameter ( $a$ ), representing psychological factors (set, setting, preparation), rather than dose alone.
- Statistical Convergence: The critical dose $D_{c1}$ was estimated using five independent methods: bifurcation fitting, Landau variance fitting (Lorentzian), and cubic spline analysis of mean and variance curves.
- Pharmacokinetic (PK) vs. Pharmacodynamic (PD) Dissociation: The study compared the coefficient of variation in plasma psilocin levels against the variance in subjective effects to determine if variability is driven by drug exposure or neural thresholds.

Key Contributions and Results
The paper establishes a coherent mathematical structure explaining the variability in psilocybin response through two distinct thresholds:

The First Threshold ( $D_{c1} \approx 18 \pm 1$ mg):
- Convergence: Five independent analytical methods applied to diverse datasets converged on a critical dose of $18 \pm 1$ mg. This represents the transition from ordinary to altered consciousness.
- Neural vs. PK Origin: Analysis of Holze et al. data revealed that inter-subject variance in subjective effects is 52 times larger than variance in plasma drug levels. Crucially, subjective variance peaks near $D_{c1}$ and then declines, whereas PK variance remains stable. This confirms $D_{c1}$ is a neural threshold phenomenon, not a pharmacokinetic artifact.
- TRD Preservation: The critical dose was found to be virtually identical in TRD patients ( $17.8 \pm 1.3$ mg) as in healthy volunteers, suggesting $D_{c1}$ is a structural property of cooperative 5-HT2A receptor activation rather than a state-dependent consequence of depression.
The Second Threshold ( $D_{c2}$ ):
- Individual Dependence: Unlike the first threshold, the transition to a transformative mystical experience ( $\theta^*_2$ ) does not exhibit a dose-dependent variance peak. Instead, high variance persists across all doses above $D_{c1}$ .
- The Role of Parameter $a$ : The model posits that reaching $\theta^*_2$ depends on an individual asymmetry parameter ( $a$ ), defined by psychological flexibility, preparation, and set/setting. Increasing the dose (parameter $b$ ) cannot reliably induce $\theta^*_2$ if $a$ is low; only increasing $a$ (via preparation) can.
- Episode Duration as a Proxy: The study identifies episode duration as a proxy for parameter $a$ , proposing an inverse relationship ( $a \propto 1/t_{episode}$ ). This mechanistically explains why chronic depression predicts non-response: long-standing maladaptive priors reduce psychological flexibility, lowering the probability of crossing the second threshold.
Validation and Predictive Power:
- COMP006 Explanation: The model exactly predicts the COMP006 response rates (5% at 1 mg, 18% at 10 mg, 39% at 25 mg) without free parameters. At 25 mg, all patients cross $D_{c1}$ , but only 39% possess sufficient parameter $a$ to cross $D_{c2}$ .
- Historical LSD Data: The model explains the near-zero inter-study heterogeneity ( $I^2 \approx 0\%$ ) in historical LSD trials for alcoholism. Since all subjects received doses far above $D_{c1}$ , the variance in response was determined solely by the population distribution of parameter $a$ , which is constant across studies.
- EEG Evidence: Independent EEG data shows that neural markers of altered consciousness (alpha suppression) have low inter-subject variability (consistent with crossing $D_{c1}$ ), while markers of complexity and beta power have high variability (consistent with individual access to $\theta^*_2$ ).

Significance and Claims
The paper claims that the 39% response rate in COMP006 is not a pharmacological ceiling but a consequence of applying a one-threshold strategy (dose escalation) to a two-threshold system. The primary clinical implication is falsifiable: intensive psychological preparation should increase therapeutic response rates more effectively than dose escalation above $D_{c1}$ .

The model reframes the therapeutic mechanism:

Dose determines whether the brain enters an altered state ( $\theta^*_1$ ).
Preparation (Parameter $a$ ) determines whether that state becomes transformative and therapeutic ( $\theta^*_2$ ).

By identifying that the rate-limiting step for the 61% of non-responders is psychological (parameter $a$ ) rather than pharmacological, the model suggests that optimizing preparation protocols—particularly for patients with long episode durations—is the critical path to improving outcomes, rather than increasing drug dosage. The author asserts that this quantitative framework unifies disparate observations (variance peaks, episode duration effects, and historical heterogeneity) that prior pharmacological models could not explain.

Technical Summary: A Two-Threshold Model of Psilocybin Response

More like this