970x125
“You shall know the truth, and the truth shall make you mad.” —Aldous Huxley
In the span of three months, psychedelic medicine crossed two important thresholds.
In April, President Donald Trump signed an executive order making it federal policy to accelerate psychedelic research and “appropriate drug approvals” for serious mental illness. Then, in July, Eli Lilly agreed to acquire AtaiBeckley, a biotechnology company developing psychedelic and psychedelic-derived treatments.
These developments are understandably being read as validation of the field. But they validate two things more clearly than they validate the medicines themselves: that the federal government wants development to move faster, and that a major pharmaceutical company believes the commercial opportunity is substantial.
Neither political commitment nor market valuation is a scientific verdict.
That distinction matters because psychedelic therapies may indeed become important treatments. Patients with treatment-resistant depression, post-traumatic stress disorder, substance-use disorders, and other disabling illnesses urgently need better options. The current evidence contains credible signals of benefit, and those signals justify serious investment and accelerated research.
But urgency is not the same as haste. The American Psychiatric Association, while welcoming federal support for research, responded to the executive order by emphasizing that the scientific evidence remains inadequate to endorse psychedelic treatments for psychiatric disorders outside approved investigational studies. It called for trained clinicians, controlled settings, and the same scientific and ethical standards applied to other promising medicines.
The central problem is not that psychedelic drugs have been proved ineffective or unacceptably dangerous. It is that their effects are unusually difficult to evaluate—and unusually difficult to separate from the circumstances in which they are administered.
Consider blinding, one of the foundations of a reliable clinical trial. In an ordinary drug study, neither patients nor investigators are supposed to know who received the active medication and who received the placebo. But a person who undergoes an intense alteration in perception, emotion, or sense of self-will often knows that an active psychedelic has been administered. The therapist may know, too.
A study may therefore be double-blind on paper while being functionally unblinded in practice. FDA reviewers identified this problem explicitly during their assessment of MDMA for PTSD, observing that the drug’s acute effects made blinding nearly impossible and created a risk of expectancy bias.
Then there is the question of what the treatment actually is.
Many psychedelic protocols combine a drug with preparatory meetings, hours of monitored administration, psychological support, and post-session “integration.” These elements may be essential to safety and perhaps to efficacy. But they also make it difficult to determine whether improvement results from the molecule, the psychological intervention, the therapeutic relationship, expectancy, the carefully constructed environment—or some interaction among them.
This is not an academic distinction. Regulators, clinicians, and insurers need to know what they are approving, delivering, and paying for. Is it a drug? A drug-plus-psychotherapy package? A specialized clinical service? Could the same results be reproduced by ordinary treatment centers, or do they depend on unusually committed investigators and carefully selected participants?
The FDA’s review of MDMA-assisted treatment offers a cautionary case. Published Phase 3 trials had reported clinically meaningful improvements in PTSD. Yet the FDA subsequently concluded that the application did not provide substantial evidence of effectiveness or establish adequate safety for approval in its submitted form. It recommended another randomized trial with blinded long-term follow-up.
Psychedelics Essential Reads
That decision did not prove that MDMA cannot work. It showed that positive symptom scores do not settle every question that matters.
Durability is especially important. Depression, PTSD, and addiction are generally chronic or recurrent disorders. An improvement lasting several weeks may be valuable, but it does not by itself tell clinicians when patients should be retreated, which patients are likely to relapse, what maintenance care they will require, or whether repeated exposure changes the benefit-risk balance. Long-term studies must assess functioning, hospitalization, suicidality, substance use, employment, relationships, and quality of life—not merely changes on a symptom scale.
Safety also cannot be inferred solely from the relatively low rate of catastrophic events in small, highly screened trials. Research participants are commonly selected to exclude people with psychosis vulnerability, bipolar-spectrum illness, unstable substance use, significant cardiovascular risk, or complicated medication regimens. Yet many of the patients who seek these treatments after approval—or outside the medical system—will have exactly those characteristics.
Medication discontinuation adds another layer of risk. Patients may be asked to taper antidepressants or other psychiatric drugs before receiving a psychedelic. That taper is not an administrative detail. It can produce withdrawal symptoms, relapse, suicidality, or destabilization and should be studied and monitored as an active clinical intervention.
The word “psychedelic” itself can obscure more than it clarifies. Psilocybin, 5-MeO-DMT, MDMA, ketamine, and ibogaine differ in receptor pharmacology, subjective effects, duration, cardiovascular risk, abuse liability, therapeutic rationale, and required medical infrastructure. Ibogaine’s arrhythmia risk, for example, creates fundamentally different monitoring requirements from those associated with psilocybin. Policy that treats all of these compounds as one therapeutic movement risks flattening clinically consequential differences.
President Trump’s order does not, in its text, direct the FDA to approve a particular drug or discard statutory approval requirements. It establishes priority and seeks faster consideration of rescheduling after successful Phase 3 development. It also specifies that timely approvals remain for drugs meeting the standards of the Federal Food, Drug, and Cosmetic Act.
That qualification should govern implementation. Acceleration should mean removing needless administrative delays; it should not mean lowering the evidentiary threshold because the illness is serious, the political leadership is enthusiastic, or investors have already assigned a multibillion-dollar value to the field.
Lilly’s involvement could bring substantial benefits. Large pharmaceutical companies have resources for manufacturing, multicenter trials, regulatory science, and pharmacovigilance that smaller biotechnology firms often lack. But the deal also illustrates how financial expectations can become attached to regulatory milestones.
Before widespread approval, pivotal trials should formally measure whether participants, clinicians, and raters correctly guessed treatment assignment. Psychological-support models should be defined, monitored for fidelity, and subjected to independent oversight. Longer follow-up should establish relapse rates, retreatment needs, and functional outcomes. Adverse-event systems should capture not only experiences labeled unpleasant but also apparently “positive” effects relevant to impairment or abuse potential. Therapist credentialing, boundaries regarding touch, session recording, grievance procedures, site accreditation, and emergency protocols should be treated as core patient protections rather than optional professional customs.
Caution is not an argument for prohibition. Nor is it indifference to patients who are suffering. It is the discipline required to prevent enthusiasm from outrunning knowledge—and to ensure that genuinely effective treatments survive the consequences of premature commercialization.
Patients deserve urgency. They also deserve evidence strong enough to bear the weight now being placed upon it.
