PTSD is what happens when your brain encodes a memory wrong. Most memories get filed away, lose their sharp edges, and when you pull them back out years later they feel like memories. PTSD memories don’t do that. They stay in the present tense and keep their full sensory load. The brain treats them like the threat is still happening, because as far as the alarm circuit is concerned, it is.
Everything else flows from that one piece of biology. The nightmares, the hypervigilance, the way a car door slamming three blocks away can drop you into a flashback in your kitchen. The avoidance. The numbness. The drinking that starts because alcohol is the cheapest thing on earth that quiets the nervous system at 11 PM. All downstream of a memory that won’t move into the past.
About 8% of Americans will meet criteria at some point. Women are diagnosed roughly twice as often as men, though anyone who works with veterans or trades will tell you the male number is artificially low because guys don’t show up until something is on fire. The trauma almost never is combat. Car accidents, sexual assault, medical events, sudden deaths, childhood stuff nobody clocked as traumatic when it was happening.
Big T and little t
Clinicians sometimes split trauma into “Big T” and “little t.” Big T is the stuff that obviously qualifies. Combat, rape, witnessing a death, surviving a disaster. Little t is the slower kind, the one that doesn’t have a single trigger. A childhood with a parent whose moods you had to track like weather. A decade in a marriage where you weren’t physically hurt but stopped being able to tell what was real. A bad first responder year.
The DSM was built around Big T because that’s what the Vietnam-era research was studying. Little t can produce a presentation that looks identical at the neurobiological level. Same alarm circuit, same intrusion symptoms, same avoidance, same sleep destruction. A woman in her 30s came in last year convinced she didn’t “deserve” a PTSD diagnosis because nothing dramatic had happened to her. Twenty years of walking on eggshells around an unpredictable father. Her startle response when I dropped a pen on the desk told me everything I needed to know.
If the alarm system rewrote itself around what happened to you, it counts. The threshold is functional, not theatrical.
The therapies that actually move trauma
Three trauma-focused therapies carry the strongest data: prolonged exposure (PE), cognitive processing therapy (CPT), and EMDR. The VA and the APA both put these at the top of their guidelines. Each one does roughly the same job through a different door.
Prolonged exposure is the most straightforward and the most punishing. You sit with a therapist and tell the story of the trauma, in detail, in first person, over and over, until the memory loses its grip. You also do in-vivo exposure to situations you’ve been avoiding. The dropout rate is real. The first few sessions feel like being asked to set yourself on fire on purpose. Patients who stay get better at a rate other interventions can’t match.
CPT is the talkier version. You identify “stuck points,” beliefs the trauma installed that don’t match reality. Things like “I caused this,” “the world is uniformly dangerous,” “I can’t trust anyone.” You write about them, challenge them, rewrite them. Tolerated better than PE because you’re not relentlessly re-immersing in the memory itself. Outcomes are comparable.
EMDR sounds like nonsense and isn’t. You hold the traumatic memory in mind while the therapist moves their fingers back and forth across your visual field, or uses tones, or taps. The bilateral stimulation appears to help the brain reprocess the memory so it stops living in the present. The mechanism is still debated. The outcomes are good. I refer to it constantly, particularly for patients who can’t tolerate PE.
The work moves a memory from the present tense into the past tense. That’s the whole job.
All three involve, in some form, going back toward the memory rather than around it. That’s the part that matters. Supportive therapy, where you talk about your week and process feelings in a general way, is humane and doesn’t move PTSD much. The studies have been clear on this for two decades. If your therapist isn’t doing trauma-focused work, you’re getting support, not treatment. Both are fine. They are also two different things.
Where medication actually fits
Two SSRIs are FDA-approved for PTSD: sertraline (Zoloft) and paroxetine (Paxil). The effect size is modest. They take the edge off the hyperarousal, help with the depression that usually rides shotgun, make sleep slightly less terrible. They don’t fix the stuck memory. Anyone who tells you an SSRI alone treats PTSD is selling something. I usually start sertraline at 50mg and titrate to 100-200mg over a few weeks. Paroxetine I use less often because the discontinuation syndrome is rough.
Prazosin is the one that surprises people. Old blood pressure medication. At 1-10mg at bedtime it dramatically reduces trauma nightmares for a lot of patients. Not all. The big VA trial a few years back muddied the picture, but in clinic I still see plenty of people go from waking up screaming three nights a week to sleeping through. Cheap, well tolerated. If it works you’ll know inside two weeks.
Propranolol gets researched as a “reconsolidation blocker.” Reactivate a memory, take propranolol, and the emotional charge that gets re-encoded is weaker. Interesting data, not yet practice-changing. Small studies, mixed replication, no clean protocol for a community clinic. I don’t prescribe it for PTSD outside research.
Ketamine and psilocybin are getting airtime. MDMA-assisted therapy was on track for FDA approval until the agency rejected it in 2024 over study design problems, which was a real loss for the field. Ketamine has some signal in PTSD-specific trials. Psilocybin work is earlier. None first-line. If a clinic is offering ketamine for PTSD as the primary intervention, ask whether they’re doing the therapy work alongside. The drug without the therapy is unlikely to do much.
Trauma-focused therapy
PE, CPT, or EMDR. Twelve to sixteen sessions in most protocols. The piece that actually moves the underlying memory. Nothing else competes with this.
Sertraline or paroxetine
FDA-approved. Modest effect on hyperarousal and mood. Takes 4-6 weeks. Worth doing alongside therapy, not instead of it.
Prazosin for nightmares
1-10mg at bedtime. Old blood pressure drug. When it works, it works fast. Worth a trial if trauma nightmares are wrecking sleep.
The stuff that rides along with PTSD
Pure PTSD, as a single diagnosis with nothing attached, is rare. Comorbidities are almost the rule. Major depression shows up in roughly half of cases. Substance use disorders are heavily represented, particularly alcohol and cannabis and increasingly stimulants. Chronic pain is a big one and chronically underrecognized. There’s a reason VA pain clinics and PTSD clinics overlap as much as they do.
The alcohol piece deserves its own paragraph. If your nervous system has been on high alert for years, alcohol is one of the few things that reliably turns the volume down. People self-medicate with it because it works in the short term, then stops working, then makes everything worse. Trying to treat PTSD without addressing the drinking wastes everyone’s time. Treating the drinking without addressing the PTSD usually fails too. They come down together or not at all.
Chronic pain after trauma isn’t always physical injury. A nervous system stuck in threat mode runs the pain system hot. I’ve had patients whose back pain dropped dramatically over the course of EMDR for reasons having nothing to do with their spine and everything to do with their amygdala.
How the work usually goes
Weekly sessions, usually. The first few weeks are stabilization. Work on sleep, get any drinking into a manageable place, establish that you can do the work without falling apart. Then the actual trauma processing begins, and that’s the hard part. Patients often get worse before they get better, which I tell people up front so they don’t bail at session four.
Most courses run 12 to 16 sessions. Some take longer, particularly with complex or developmental trauma. Nobody is trying to make you forget. What you’re trying to get to is a place where you can think about what happened without your body acting like it’s happening again. The memory finally becomes a memory.
A lot of people put off treatment for years because they assume the work will undo them. For most patients the opposite ends up being true. The avoidance has been eating their life. The treatment, painful in the short term, gives people back parts of themselves they’d written off. The 50-year-old paralegal who hadn’t driven the freeway since her wreck in her late 20s. The ER nurse who hadn’t been to a fireworks show since 2009. Those are the small recoveries the alarm circuit was holding hostage the entire time.