Can journaling about trauma make PTSD worse?

For some people, yes. Sbarra and colleagues (2013) found that expressive writing worsened recovery for high-ruminators after marital separation, and several clinical trials in trauma populations show null or negative results. People with a dissociative presentation, active suicidal ideation, recent acute trauma, or trauma that floods without any recovery between sessions are at elevated risk when they write alone and unstructured. The harm signal exists in the literature; popular journaling content rarely mentions it.

What raises the risk when writing about trauma without support?

Higher-risk situations include active suicidal ideation, a dissociative presentation with derealization or losing time while writing, acute trauma in the last several weeks, active substance dependence tied to the trauma, very high trait rumination, current psychotic symptoms, unstable housing or food insecurity, and having no one to turn to afterward. None of these are absolute bars for everyone, but each is a reason to write with clinical support rather than alone.

How do I know if writing is helping or hurting me?

A brief spike in sadness right after writing is expected and usually fades within hours. Warning signs that it is hurting: intrusive images that intensify across days rather than settling, dissociating while you write, sleep that stays disrupted past the first night, rising substance use, suicidal thoughts, or emotional flooding you cannot recover from before the next session. If any of these appear, stop and talk to a clinician.

What's a safer way to do expressive writing about trauma?

Titrate the dose, meaning short time-limited sessions rather than long unbounded ones. Sequence prompts from lower-arousal to higher-arousal with pauses to regulate. Do grounding before and after the hard material. Set an explicit stopping rule in advance. Where possible, do this with a clinician who can screen you first and help you integrate afterward. My writing app is built around that sequencing; an empty notebook and a self-help instruction are not.

Is the Pennebaker writing protocol dangerous?

It was developed and tested mostly in non-clinical people writing about ordinary difficult experiences, and in that population it is generally safe and modestly helpful over subsequent weeks. Risk rises when the same brief protocol is applied to clinical PTSD without any oversight, screening, or stopping rule. The paradigm was never designed as a trauma treatment; popular adoption stretched it past the population it was tested in.

When Writing About Trauma Makes Things Worse: The Dysregulation Risk Nobody Names

TL;DR: Expressive writing reliably raises distress right after you write, with benefits accruing only over the following days. For high-ruminators, people with active or dissociative PTSD, those in acute crisis, and anyone with thin regulation capacity, unstructured solo writing can worsen distress rather than relieve it. Titration, support, and clear stopping rules are what separate help from harm.

A woman brought a paperback to her third session, the corner of one page folded down, and set it on the couch between us the way you set down evidence. She had done what the book said. Four days, twenty minutes each, writing continuously about the worst thing, not stopping for grammar or logic. By the second night she was not sleeping. By the fourth she was back inside a room she had spent nine years learning to stay out of, narrating it in the present tense because the instruction told her to write as if it were happening, and it obliged. She wanted to know what she had done wrong.

She had done nothing wrong. She had followed a sound instruction that was never built for her situation, and no one had told her the difference, because the difference is exactly the part that gets left out of the self-help version.

The instruction was real. The warning label was missing.

James Pennebaker’s expressive writing paradigm is one of the better-replicated findings in health psychology. Ask people to write about an emotional upheaval for a few sessions and, on average, over the following weeks and months, you see modest improvements in mood, immune markers, and physician visits. The effect is real and it is small, and it has been studied since the mid-1980s, when the first participants were college undergraduates writing about ordinary painful experiences.

The part that survives into the blog posts is the promise. The part that gets cut is the mechanism, which matters here more than the promise does. Writing about trauma does not work by catharsis, by draining off feeling until you are lighter. Across the literature the immediate result of expressive writing is a short-term increase in negative affect: sadder, more physically symptomatic, worse mood in the hours right after the pen goes down. For most non-clinical writers that spike fades within hours and the benefit, when it comes, arrives days later through some slower process of making the experience narratable. The distress is the front end of the intervention, not a sign it is failing.

Which is fine, if you have the capacity to hold the spike. The whole risk sits in that conditional.

The harm signal in the literature

The clearest single result is Sbarra and colleagues, 2013, published under a title that says the thing plainly: Expressive Writing Can Impede Emotional Recovery Following Marital Separation. Ninety recently separated adults were randomized to traditional expressive writing, a narrative variant, or neutral control writing. For participants high in rumination, the ones still actively brooding over the separation in search of an answer, expressive writing produced worse emotional outcomes over the following months than simply writing about neutral topics. The people you would most expect to benefit from processing were the people it set back. For high-ruminators the control condition, the one that gave them nothing to process, left them least distressed.

That is not an outlier finding hiding in a niche journal. It sits alongside a broader pattern that the field tends to phrase gently. Several meta-analyses of expressive writing find no reliable psychological benefit at all, and in clinical PTSD samples the results run null to negative more often than the popular framing admits. The usual clinical interpretation is that the protocol needs refinement, better prompts, acceptance instructions, more engaged writers. Read a second way, the same data carry a contraindication signal: for a subset of distressed people, unstructured exposure to their worst memory, alone, on a fixed schedule, is the wrong dose delivered without a way to stop.

Both readings can be true. The protocol probably can be improved, and the same population that most needs improvement is the population most exposed to harm when it is not.

Who is in the higher-risk zone

None of what follows is an absolute prohibition, and none of it is a diagnosis you can hand yourself from a blog post. Each item is a reason to write with a clinician rather than alone with a paperback.

A dissociative presentation. If you tend to leave your body under stress, lose stretches of time, or feel the room go unreal, sustained trauma writing can pull you out of the window in which you can actually process anything and into the numb, faraway state where you can only endure it.
Active suicidal ideation. Writing that reliably raises distress is the wrong tool when the baseline already includes thoughts of ending your life. The order of operations is safety first, then processing.
Recent acute trauma. In the first weeks after an event, the memory is still raw and unconsolidated. Forcing narrative onto it too early can deepen the groove rather than smooth it. Time and stabilization come before written exposure.
Very high trait rumination. This is the Sbarra population. If your relationship to the memory is already a loop you cannot exit, more writing feeds the loop instead of closing it.
Thin regulation capacity, or what the older literature called low ego strength. If you have little practice bringing yourself back down once you are activated, and no reliable person or skill to help you do it, the intervention hands you the activation and withholds the recovery.

Active substance dependence bound to the trauma, current psychotic symptoms, and material instability like unstable housing or food insecurity belong on the same list, for the same reason: each thins the resources you would need to metabolize what the writing brings up.

The eight signs that mean stop

The expected short-term spike in sadness is not on this list. That fades. These do not, and if any of them appear, the instruction is to put the pen down and talk to someone.

Intrusive images that intensify across days instead of settling.
Dissociating while you write, going blank, unreal, or elsewhere.
Sleep that stays disrupted past the first night.
A sharp rise in drinking, using, or other numbing behavior.
Suicidal thoughts, at any intensity.
Emotional flooding you cannot come back from before the next session.
A worsening, not an easing, of the symptom you started writing to relieve.
The sense that you are being pulled into the memory rather than looking at it.

The distinction under all eight is the difference between working with a memory and being taken back into it. One expands what you can hold. The other narrows it.

Why the window of tolerance frame matters

Dan Siegel’s window of tolerance names the band of arousal in which a nervous system can still think and feel at the same time. Above it lies hyperarousal, panic and racing and flooding. Below it lies hypoarousal, the numb collapse of dissociation. Trauma narrows the window, so that memories which would merely sting a settled system throw a traumatized one straight past the edge, up into panic or down into shutdown, where no processing is possible because the machinery of processing is offline.

Effective trauma work keeps the writer inside the window on purpose. It titrates. Short exposures, then a return to the present, then a slightly longer one, the dose calibrated to what the system can metabolize without going over the edge. The four-day, twenty-minute, write-without-stopping instruction does none of this. It is a fixed dose delivered on a clock, indifferent to whether the person on day two is still inside their window or already three states past it. For a stable writer, fine. For the woman with the folded page, it was a schedule that kept escorting her out of the room she needed to stay in.

What supervision actually adds

A clinician in the loop is not a formality. Four concrete things change.

Screening. Someone asks, before you begin, about dissociation, current ideation, how recent the trauma is, what happens when you get activated, who you have. The higher-risk zone gets caught at the door instead of discovered on night four.

Dosing. Prompts get sequenced from lower to higher arousal, with regulation built in before and after the hard material and pauses written into the structure, so that intensity climbs at a pace a nervous system can follow rather than all at once.

Regulation work. You arrive already holding at least one reliable way to bring yourself down, so that the activation the writing produces meets a skill instead of an open drop.

Integration. The protocol ends in a conversation rather than a closed notebook, where what surfaced gets metabolized with another mind in the room instead of left to circulate alone at two in the morning.

This is the logic my writing app is built around: a screening gate that respects the contraindications above, prompts that climb by arousal rather than by calendar, regulation before and after exposure, and explicit stopping rules rather than a fixed four-day march. It is not a clinician, and it does not pretend to be one. If you recognized yourself in the higher-risk zone, the right first move is a person, not an app and not a paperback.

The figure that has to be welcomed first

There is a slower way to say all of this. Before a trauma can be welcomed back into the story of a life, the part of the person who lived through it has to be welcomed first, and those are not the same act. Writing that goes straight at the event, hard and on schedule, addresses the memory and skips the one who holds it. The nervous system reads that as another thing happening to it rather than something being done with it, and it defends accordingly, by fleeing up into panic or down into numbness. Sequencing is not a technicality. It is the difference between approaching the frightened part as an ally and ambushing it. The woman with the paperback had been ambushing herself, faithfully, for four nights, on the authority of a book that meant her no harm.

Expressive writing can be one of the cheaper, more available goods in the whole psychological toolkit. It can also be the wrong instrument, wrongly dosed, for the exact person most drawn to it. Knowing which one you are holding is not something an instruction on a page can tell you. It is the thing worth finding out before you begin.

If you’re in crisis: In Pittsburgh, RESOLVE Crisis Services (1-888-796-8226) offers 24/7 support, and UPMC Western Psychiatric Hospital provides emergency psychiatric care. Nationally, the 988 Suicide and Crisis Lifeline is reachable by call or text at 988. If you are in immediate danger, call 911.

This post is educational and does not substitute for individualized clinical advice. If you’re carrying trauma you want to write your way through, consider what happens to dreams after trauma, the frozen, can’t-move dream as a picture of parasympathetic shutdown, and the fuller frame in expressive writing for trauma.