TL;DR
Brain differences in PTSD are real but small: hippocampal volume shows an effect size of just d = -0.17 across nearly 1,900 people, and no scan can diagnose an individual. Contrary to the popular story, PTSD is more often linked to lower daily cortisol, not higher. Psychedelic neuroplasticity findings come largely from mice, and the most striking human connectivity study involved only seven participants.
Something happened to you. Years later, your body still answers a door that nobody knocked on. You’ve probably been told your brain was rewired by it, that your amygdala is enlarged, that your nervous system is stuck. Some of that is true. A lot of it has been stretched far past what the research can hold.
I want to walk you through what brain science actually says about trauma and healing. Not the version that sells workshops. The version with the error bars still attached. Because the real findings are quieter than the headlines, and honestly, I find them more hopeful.
Gentle Reminders
- The brain differences found in PTSD are real but small. Hippocampal volume shows an effect size of d = -0.17 across nearly 1,900 people (Logue et al., Biological Psychiatry, 2018). No brain scan can diagnose you.
- Trauma is more often linked to lower daily cortisol, not higher (Morris, Compas & Garber, Clinical Psychology Review, 2012). The flooding story is mostly backwards.
- Plant medicine research is genuinely promising and genuinely early. Some of the most-quoted findings come from mice, or from studies of six or seven people.
- A theory can be shaky while the practice built on it still helps you. Those are two separate questions.
The neuroscience of trauma: what the brain data shows
If you’ve carried something heavy for a long time, you already know it changed you. The science agrees, and then it gets specific in a way that surprises most people. The largest study of its kind pooled brain scans from nearly 1,900 adults and found that people with PTSD had smaller hippocampal volume than trauma-exposed people without PTSD, with an effect size of Cohen’s d = -0.17 (Logue et al., Biological Psychiatry, 2018).
That number deserves a plain translation. A d of -0.17 is a small effect, in fact just under the conventional cutoff for small. Not zero. Not nothing. Small. It means that if you lined up a hundred people with PTSD and a hundred without, the two groups would overlap almost completely. You could not pick out a single person from their scan.
The amygdala finding was even more modest, d = -0.11, and it did not survive statistical correction for multiple comparisons. (That correction is a guardrail against false positives when a study tests many brain regions at once. The finding may still be real. It just hasn’t earned confidence yet.)
Childhood trauma shows up somewhat differently. A meta-analysis of functional scans from more than 580 children found disruption concentrated in two networks: the default mode network, tied to your sense of self and inner narration, and the central executive network, tied to problem-solving and control (Ireton, Hughes & Klabunde, Biological Psychiatry: CNNI, 2024). Not damage to a single part. A change in how parts talk to each other.
Doesn’t trauma flood the body with cortisol?
This is the belief I hear most often, and it’s the one the evidence contradicts most cleanly. Across 47 studies covering more than 6,000 people, daily cortisol output was lower in people with PTSD, not higher, with an effect size of d = -0.36 (Morris, Compas & Garber, Clinical Psychology Review, 2012). When PTSD occurred alongside depression, the gap widened to d = -0.65.
Sit with that for a second. The system people describe as permanently screaming may, for many, be running flat. Under-responsive rather than over-responsive. Depleted rather than flooded.
That reframe changed how I sit with people. Numbness, exhaustion, the feeling of watching your own life through glass: these aren’t a failure to calm down. They can be what a worn-out stress system looks like from the inside. Your body isn’t betraying you. It’s been on duty too long.
Trauma healing and the limits of brain plasticity
It can, and this is where I have to be careful with you, because the plasticity story is where marketing has outrun the data by the widest margin. A trauma-shaped brain is not fixed in place. Here’s what’s actually documented. In mice, a single dose of psilocybin increased the size and density of dendritic spines in the frontal cortex by roughly 10% within 24 hours, and the change was still measurable about a month later (Shao et al., Neuron, 2021).
Mice. Imaged through a window in the skull with two-photon microscopy. That last detail is the whole story, because you cannot do that to a living human being. There is no scanner on earth that resolves a single dendritic spine in an intact human brain.
So when a reviewer of this field writes that clinical research on psychedelics and cellular neuroplasticity “is lacking” (de Vos, Mason & Kuypers, Frontiers in Psychiatry, 2021), that isn’t sloppiness on anyone’s part. That’s a methodological ceiling. Put it plainly: there is no direct human evidence that psilocybin, or any psychedelic, regrows neurons or restructures synapses in a living person’s brain. The structural claim rests entirely on animals. Anyone telling you a mushroom regrew your synapses is describing a rodent, and hoping you won’t ask.
What does psilocybin do inside the human brain?
Something real, and something stranger than the popular version. The first fMRI study of psilocybin in humans found decreased activity in the default mode network, the hub of self-referential thought and rumination, with deeper suppression tracking a more intense subjective experience (Carhart-Harris et al., PNAS, 2012). That became the famous “quieting the mind” story. If you know the particular exhaustion of a mind that narrates you at 3 a.m. and won’t stop, you already understand why that phrase caught on.
A 2024 study updated it. Psilocybin didn’t simply turn the network down. It desynchronized it, scrambling the timing of connections across the brain more than three times as much as a stimulant comparison drug, with the effect strongest in the default mode network. Decoupling between the anterior hippocampus and that network persisted for around three weeks (Siegel et al., Nature, 2024).
Now the caveat, and it’s a big one. That study scanned seven people, six of them in the primary analyses, and none had depression or PTSD. The authors say plainly that confirming this will require precision patient studies. So: a vivid, well-measured signal in a very small group of healthy volunteers. Not proof of anything about your healing. If you want the fuller clinical picture, I’ve written about that separately in psychedelic therapy for trauma.
One early human signal worth naming. In an open-label trial, a single 25mg dose of psilocybin given with psychological support was well tolerated in adults with PTSD, and their symptom improvement held out to 12 weeks, with no serious adverse events reported (McGowan et al., Journal of Psychopharmacology, 2025). That’s a safety-and-tolerability result, not evidence of a cure. Still, it’s the kind of careful early work worth watching: promising, and honest about its own size.
Does cannabis calm a traumatized nervous system?
The mechanism is beautifully documented, and I want to hold that alongside a result nobody in this space likes to mention. The endocannabinoid system acts as a central integrator of fear, anxiety and stress, and it’s essential to fear extinction, the process by which a threat memory stops setting off alarms (Lutz, Marsicano, Maldonado & Hillard, Nature Reviews Neuroscience, 2015).
Then someone ran the trial. The first randomized controlled study of smoked cannabis for PTSD followed 80 veterans across several cannabis types and a placebo. Everyone improved. No active treatment beat placebo (Bonn-Miller et al., PLOS ONE, 2021).
The receptor biology is real and the clinical trial was negative. Both things are true at once, and holding them together is the most honest position available right now. That’s part of why I treat cannabis as optional, judicious and microdose-level in my work, a support for presence rather than a treatment for a diagnosis.
Is polyvagal theory still standing?
You’ve almost certainly met this one. Ventral vagal, dorsal shutdown, the social engagement system. It has become the common language of trauma-informed spaces, and its foundations are now under serious challenge. In 2026, 39 authors published an international expert evaluation in Clinical Neuropsychiatry concluding that the polyvagal theory is “untenable” as neurophysiology (Grossman et al., Clinical Neuropsychiatry, 2026). Stephen Porges published a rebuttal in the same issue, and a defense of the theory the year before. This is a live argument, not a settled verdict.
Here’s the distinction I’d ask you to keep. The critique targets the mechanism, the specific claims about vagal branches and evolutionary sequence. It does not establish that slow breathing, co-regulation with a safe person, or attention to cues of safety stop helping people. Those practices may work for reasons the theory described incorrectly. A wrong map can still lead somewhere real. If slow breathing and co-regulation steady you, keep them close. You’ll find gentle versions in my free meditation practices for emotional healing.
Where the science ends and your experience begins
Notice what the honest version leaves you with. There are small structural differences, real but far too subtle to diagnose anyone. Your stress system is more likely depleted than flooded. The networks in your brain weren’t destroyed; they learned to talk to each other differently, and that difference is where the work happens. Plant medicines carry genuine mechanisms alongside thin, early human evidence. One widely taught theory is now openly contested.
That’s not a discouraging picture to me. There’s room in it. Nothing there says you are broken or fixed in place. The Emotional Liberation® method I teach works with psycho-spiritual inquiry, expressive movement, kriyas and, when it’s right for someone, plant medicine, exactly because no scan is going to tell you what your grief needs. Your deep inner knowing is not a poetic flourish standing in for data. It’s the instrument the data can’t replace. You’ll find more of the studies I follow on the Science page.
Here’s the whole picture in one place, each finding paired with the caveat that rarely travels with it:
| Finding | Study | Effect or detail | Caveat |
|---|---|---|---|
| Smaller hippocampal volume in PTSD | Logue et al., 2018 | d = -0.17 | Small effect, just under the small cutoff; the groups overlap almost completely |
| Smaller amygdala volume in PTSD | Logue et al., 2018 | d = -0.11 | Did not survive statistical correction for multiple comparisons |
| Lower daily cortisol in PTSD | Morris, Compas & Garber, 2012 | d = -0.36 | Runs opposite the “flooded” story; an average across studies, not a personal reading |
| Psilocybin desynchronizes the default mode network | Siegel et al., 2024 | Anterior hippocampus-DMN decoupling lasted about three weeks | Only seven participants, none with PTSD or depression |
| Psilocybin grows dendritic spines | Shao et al., 2021 | Spine density up about 10% within 24 hours | Mouse study; no direct human structural evidence exists |
| Smoked cannabis for PTSD | Bonn-Miller et al., 2021 | No cannabis type beat placebo | First randomized controlled trial; everyone improved, including the placebo group |
Frequently asked questions
Can a brain scan show whether I have PTSD?
No. The differences found in the largest study to date are small at the group level, d = -0.17 for hippocampal volume (Logue et al., 2018), with heavy overlap between groups. Population averages don’t transfer to individuals. Diagnosis remains a clinical conversation, not an imaging result.
Is my cortisol high or low if I’m traumatized?
Meta-analytic evidence points to lower daily output in PTSD, d = -0.36, and lower still when depression is also present, d = -0.65 (Morris, Compas & Garber, 2012). That’s an average across studies, not a prediction about you. Only a clinician with your labs can say what your levels are.
Does psilocybin rewire the brain?
In mice, a single dose grew dendritic spines by around 10% within a day (Shao et al., 2021). In humans, that measurement is technically impossible with current tools. What has been shown in people is altered network connectivity, and in the strongest study of that, in only seven participants.
Should I stop practicing polyvagal exercises?
That’s your call, and I’d separate the two questions. A 39-author evaluation found the theory untenable as neurophysiology (Grossman et al., 2026), while its author disputes that. If breathwork and co-regulation help you, keep going. Just hold the explanation loosely.
The Bottom Line
Trauma leaves measurable traces. They’re smaller, subtler and less diagnostic than the popular story suggests, and the direction of the stress-hormone finding runs opposite to what most people are told. The plant medicine research is real and early, carried in places by mouse studies and very small human samples that deserve to be named as such.
None of this replaces care. If you’re in crisis or working with a clinician, this article is context, not treatment, and nothing here should change your medication or your therapy without that clinician’s input.
What I keep coming back to is that the science was never going to hand you the answer. It maps terrain. You still have to walk it, and you’re the only authority on what your own healing asks of you. If you’re ready to work at that depth with support, my one-on-one session packages are where that begins.
About the Author
Becca Williams is an emotions therapist and plant medicine facilitator, and the creator of Emotional Liberation®. She has been featured in LifeHacker, Psychedelic Invest, and on the Montel Williams show. Learn more about her background and approach on the Meet Becca page.