PTSD & Complex Trauma: Not Too Much Memory, but Memory Never Filed

The core abnormality is not "remembering too vividly" — it is that the traumatic memory was never properly filed. An ordinary memory carries the tags "last year, over there, already over." A traumatic memory keeps extremely high-intensity sensory detail but loses the temporal coordinates. So when it is triggered it does not arrive as recollection; it arrives as "happening now" — ICD-11 simply wrote this into the criteria (in the here and now).

Three regions fall out of balance. The amygdala encodes more strongly under extreme arousal, binding sensory fragments tightly to threat; the same level of stress hormones impairs the hippocampus — the structure that stamps "when and where" onto a memory. Intensity gets written in; the coordinates are lost. Afterwards the ventromedial prefrontal cortex exerts weaker top-down inhibition over the amygdala, so extinction learning never consolidates and a single cue can reignite the whole response. Two counterintuitive findings are worth holding onto: Yehuda measured lower, not higher, cortisol in patients; and in Gilbertson et al.'s (2002) Vietnam twin study, the identical twins who never deployed had equally small hippocampi — a small hippocampus is a vulnerability factor, not only a consequence.

Tool: the PCL-5 (20 items, published openly by the U.S. Department of Veterans Affairs); a total of 31–33 is commonly used as the threshold for further evaluation. It is a screener, not a diagnosis.

ICD-11 (2018) listed Complex PTSD as a distinct diagnosis for the first time. It is not "more severe PTSD" — beyond the three PTSD clusters it adds an entire block of disturbances in self-organization (DSO, see the table below). Someone can have very few flashbacks and still meet CPTSD fully: the damage is to the baseline, not to memory playback.

The difference comes from the shape of the trauma, not its intensity. Single, escapable events with a beginning and an end (a car crash, a disaster) mainly damage memory processing. Trauma that is prolonged, repeated, inescapable, and occurring inside a relationship (childhood abuse and neglect, long-term domestic violence) lands during the formative period of the attachment system and the self-concept. What gets shaped are three baselines: the default amplitude of emotion regulation, the default setting of self-worth, and the default expectation of whether closeness is safe — precisely where Day 2 and Day 20 converge. State the controversy plainly: DSM-5 did not adopt CPTSD, on the grounds that it is insufficiently distinguishable from borderline personality disorder, and van der Kolk's proposed "developmental trauma disorder" was likewise rejected.

Tool: the International Trauma Questionnaire (Cloitre et al.), which follows the ICD-11 structure and measures the PTSD and DSO blocks separately. Freely available.

First-line care is trauma-focused psychotherapy, not medication. Three independent guideline bodies — APA, ISTSS, and the U.S. Department of Veterans Affairs — agree closely: PE, CPT, and EMDR carry the strongest evidence, with large effect sizes. The bottleneck is not efficacy but completion: dropout is these treatments' main failure mode.

Effective therapies share one active ingredient: re-contacting the traumatic memory while physiologically in a safe state, and letting new information enter it. Two pathways: extinction learning restores vmPFC inhibition over the amygdala; and reconsolidation — Nader, Schafe & LeDoux (2000) showed that a memory briefly returns to a modifiable state when retrieved. PE dismantles avoidance through systematic repeated exposure; CPT does not require repeatedly narrating details, and instead works on stuck points such as "I could have stopped it." EMDR has substantial trial support overall, but dismantling studies repeatedly find that the incremental contribution of the eye movements themselves is small and unstable — the effect likely comes from exposure plus a working-memory dual task, not the mechanism the theory claims.

Tool: take the PCL-5 every 2–4 weeks and plot the curve. Effective treatment usually shows visible decline within 4–6 sessions; two months with no change is grounds for discussing a different approach, not for concluding you didn't try hard enough.

In the first weeks after an event, flashbacks, insomnia, vigilance, and numbness are a normal stress response, not early PTSD. Bonanno's trajectory research keeps producing the same result: populations split into several trajectories, and the majority follow the resilient one — mild symptoms, fast recovery, no treatment required. Treating a normal response as pathology has its own costs.

Natural remission requires two conditions: not being forcibly interrupted, and not being prematurely fixed in place. This explains the field's most counterintuitive finding — single-session mandatory critical incident stress debriefing is ineffective, and some studies suggest it may increase later symptoms; the Cochrane review concludes it should not be offered routinely. Two likely reasons: being required to narrate details while physiological arousal has not yet subsided amounts to re-encoding the memory in a high-arousal state; and a uniform protocol crowds out natural social support. Hence NICE recommends watchful waiting for mild presentations: active monitoring rather than immediate treatment within the first month, while protecting sleep, safety, and connection.

Four red lines for seeking a professional: symptoms not easing after a month, real impairment at work or in close relationships, using alcohol or drugs to hold the reactions down, and self-harm or suicidal thoughts (this last one does not wait out any observation period).