Read Time: 10 minutes
You know the presentation.
The client who can walk you through every detail of what happened to them without ever making eye contact. The one who circles back to 'I should have known better' no matter where the conversation starts. The one whose processing stalls out in Phase 4, not because they can't reach the memory, but because something underneath it is quietly insisting they deserve to stay stuck in it.
That something is shame. And it is in almost every trauma caseload you will ever carry.
Most clinicians know, at an intuitive level, that shame is operating. You feel it in the room before you can name it. What's harder is knowing what to do with it once you've spotted it, because shame in trauma therapy doesn't respond the way fear does, or grief does, or even anger does. It has its own logic. Its own favorite hiding spots. And its own ways of making standard EMDR protocol feel suddenly insufficient.
This post is a clinical framework for understanding shame in trauma therapy, specifically for EMDR therapists. Not a soft overview. A working map. Because the clients in your caseload who are most stuck are very often the ones whose primary driver isn't what most clinicians are looking for.
Start with the distinction that matters most for your clinical work, and for EMDR specifically: shame is not guilt.
Guilt says, 'I did a bad thing.' It's about behavior, a specific action or inaction that the person can point to, evaluate, and in theory, make amends for. Shame says something completely different. Shame says, 'I am bad.' Not what I did. What I am. And that's a much harder belief to move.
This distinction lands directly in Phase 3 of EMDR. Clients whose negative cognitions cluster around 'I am defective,' 'I am worthless,' or 'I am unlovable' are not just expressing low self-esteem. They are expressing shame-based NCs, and those respond differently in treatment than guilt-based NCs do. A guilt-based NC like 'I should have done something' has a natural positive cognition paired with it. A shame-based NC like 'I am fundamentally broken' requires you to challenge a deeply held identity-level belief, one that the client's nervous system has often been maintaining for decades.
Shame also has a signature somatic presentation that's worth knowing. Dropped eye contact. Voice that softens or flattens. Posture that curls inward. Heat in the face or chest. The physical impulse to get smaller, disappear, not be seen. When you notice these things mid-session, you are watching the nervous system respond to shame. These are not resistance behaviors. They are physiological responses that tell you something important about what the system is carrying.
The DSM-5 named shame explicitly for the first time in its diagnostic criteria for PTSD, placing it under 'persistent negative emotional states.' That's a relatively recent development. It signals that the field has caught up to what clinicians have been observing for years: shame is not a secondary symptom. For a significant portion of trauma clients, it is the primary driver.
One reframe that helps clinicians work with shame rather than around it: shame evolved as a social regulation mechanism. Its original function was to keep us inside the group, to signal when we'd crossed a line, to maintain belonging. That's not pathology. It becomes a clinical problem when it's chronic, when it was installed by someone who had no right to install it, or when it operates so broadly that it blocks every pathway to adaptive self-belief.
Understanding that helps you hold shame with more nuance. You're not fighting a disorder. You're working with a system that learned to protect itself by turning inward.
Shame doesn't arrive the same way in every client. Knowing which pathway it came through shapes how you approach it in treatment. There are three main routes.
Traumagenic shame. This is shame directly installed by the traumatic event itself. It's particularly common in sexual trauma, interpersonal violence, and abuse where the perpetrator explicitly blamed the survivor, or where cultural narratives do the blaming instead. The message gets fused to the traumatic memory: this happened because of something I am.
Research consistently shows higher shame in interpersonal trauma than in non-interpersonal trauma. Across multiple studies, shame is a stronger predictor of PTSD symptom severity than guilt. When a client keeps returning to self-blame no matter how many times you offer a different framing, you are almost certainly working with traumagenic shame encoded at the level of the memory network itself.
Introjected shame. This is shame absorbed from caregivers, institutions, or communities before the client had the cognitive capacity to question it or push back. It shows up as a 'voice' or a 'feeling' the client can't locate in any specific event. They just know it has always been there. 'I've always felt like something was wrong with me.' Under the AIP model, introjected shame may be encoded as an early foundational memory that feeds a broad network of subsequent experiences. You're not looking for one bad day. You're looking for an operating system that was installed early.
Ego-dystonic shame. This is the shame generated by the client's own behavior during states of dysregulation, particularly anger. When clients act in ways that feel misaligned with who they want to be, the resulting shame can become its own self-sustaining cycle.
This is where the anger-shame-depression continuum becomes most visible in clinical work. Anger generates behavior. Behavior generates shame. Shame generates depression. Depression narrows the window for managing anger. And the cycle restarts. Many clients have been locked in this loop for years without anyone naming what's actually driving it.
A 2023 meta-analysis by DeCou and colleagues demonstrated robust associations between trauma-related shame and symptoms of psychopathology across veteran, college, women, and clinical populations, including PTSD, dissociation, and depression. Shame is not a peripheral variable in trauma treatment. It is, for a significant portion of clients, the central one.
What is the difference between shame and guilt in trauma treatment? Guilt is behavior-focused: the client believes they did something wrong. Shame is identity-focused: the client believes something is fundamentally wrong with them. In EMDR, this distinction shapes the negative cognition in Phase 3. Guilt-based NCs cluster around responsibility. Shame-based NCs target the self directly: 'I am bad / defective / worthless / unlovable.' These respond differently to reprocessing, and shame-based NCs often require more careful preparation and a more specific positive cognition that addresses identity-level beliefs, not just safety or control.
Here's the part that makes this clinically actionable: shame in a therapy session is a master of disguise. You will miss it if you're only looking for the obvious version.
It doesn't always look like shame. Shame often presents as flat affect, compliance, minimization, and intellectualization. The client who says 'I know it wasn't my fault' in a tone that communicates the exact opposite. The client who goes quiet and cooperative mid-desensitization, suddenly offering no resistance, reporting very little. That's not good progress. That's often shutdown. The system has decided the material is too dangerous to keep processing, and compliance is how it protects itself.
Shame-based negative cognitions. The NCs worth flagging include 'I am bad,' 'I am defective,' 'I am worthless,' 'I am unlovable,' 'I am responsible for what happened.' These are identity statements, not situational assessments. They require more careful Phase 3 work than fear-based NCs, and they point toward a different kind of positive cognition, one that addresses who the client is, not just what they can do or how safe they are now.
Shame blocks the PC. A VOC that stays stubbornly low despite SUDS reaching 0 or 1 is one of the most reliable shame signals in the room. Some part of the system doesn't believe the proposed positive cognition, because a deeply held shame-based belief directly contradicts it. Installing a PC on top of an active shame network is like painting over damp. It won't hold. The belief will come back, usually in the next session, sometimes stronger.
Research shows robust associations between trauma-related shame and dissociation. The higher the shame load in a client's system, the higher the degree of dissociation you will encounter across all eight phases. This isn't incidental. Dissociation is often how the system manages shame: by making the shameful material inaccessible. This directly affects your Phase 2 preparation requirements and your Phase 4 pacing. More shame means more dissociation, which means more preparation before you can safely approach traumatic material.
The relational dimension. Shame was formed in relationship. It heals in relationship. The therapeutic alliance in EMDR is not just a nice-to-have. It is a therapeutic mechanism.
De Jongh and colleagues' 2024 case study on EMDR for complex PTSD notes that shame can function as a limiting factor in treatment, and describes specific adaptations, including the blind-to-therapist approach, that allow clients to access deeply shameful material without having to narrate it directly. Sometimes the most shame-sensitive intervention is giving the client permission not to say everything out loud.
How you respond when shame surfaces in session matters clinically. Flinching, going overly clinical, or withdrawing from the material confirms the client's shame. Staying regulated, curious, and present challenges it. That's not just good therapeutic manner. That's the adaptive memory network doing its work.
A 2024 RCT found that EMDR therapy produced significant improvements in shame and guilt alongside traumatic and obsessive symptoms. And a A 2025 study found that shame proneness is more closely associated with Disturbances in Self-Organisation (a key feature of complex PTSD) than guilt, which is consistent with what most clinicians are seeing clinically: shame is the binding agent in complex presentations.
Is EMDR effective for shame? There is growing evidence that EMDR reduces shame alongside PTSD symptoms. A 2024 randomized controlled trial found that EMDR therapy significantly improved shame and guilt outcomes. EMDR's Adaptive Information Processing model addresses shame at the level of the memory network that installed it, rather than targeting the shame response directly. When the underlying traumatic memories that established the shame-based self-belief are reprocessed, the shame often resolves as part of that process. This is why working through the AIP model to identify the foundational experiences matters so much with shame-based presentations.
Once you can recognize shame as a clinical variable, you can work with it deliberately across all eight phases. Here's what that looks like in practice.
Phase 1, History-Taking. Shame causes clients to minimize or omit the experiences they are most ashamed of. Body language is a more accurate data source than self-report here. Watch what the body does when certain topics come up, even if the verbal content stays neutral. Ask permission before exploring sensitive material. Slow down. Shame-based clients often need permission to say things out loud that they've been keeping quiet for a very long time.
Phase 2, Preparation. This is where most shame-related treatment gets made or broken. Generic calm-place resources frequently don't hold with this population, because the system has already learned to use dissociation as its primary regulatory strategy. Introducing another calming tool reinforces that pattern rather than expanding it. Resources need to be specific to the shame-carrying part of the system. A nurturing figure or protective presence that can be with the shamed part, not just the adult self who shows up for appointments, is what you're building toward. If shame is high, dissociation is high, and every part needs adequate resourcing before you go near traumatic material.
Phase 3, Assessment. Expect NCs to be more global and more self-referencing than with fear-based trauma presentations. 'I am defective' sits at a different level of the belief system than 'I am not safe.' Be flexible about which NC captures the most charge, and be prepared for the NC to shift as processing begins, because shame often reveals itself in layers. Positive cognitions in shame work need to address identity-level beliefs about worth, lovability, and belonging. 'I did my best' is often insufficient. 'I am worthy of care' or 'I belong here' tends to be more precisely targeted.
Phase 4, Desensitization. Watch for processing that loops back to self-criticism rather than moving through channels. That's your signal that shame is interfering. Interweaves that invite perspective-taking work well here: 'What would you say to a friend who had been through exactly this?' or 'If a child had experienced what you experienced, what would they deserve?' These open a processing channel without amplifying the shame response. Avoid any interweave that could inadvertently confirm the client's negative self-evaluation, even unintentionally.
Phases 5 and 6. If shame is still active, the PC won't install. Don't push it. A forced installation over an active shame network produces an unstable result. The body scan in Phase 6 often surfaces material the verbal channels couldn't reach. Somatic expressions of anger in Phase 6 are a sign of progress, not a clinical problem. When the system has enough safety to move from depression or shame toward anger, it's moving in the right direction.
Phases 7 and 8. Sequential dissociation is more likely with shame presentations. Don't rush closure. Leave enough time and don't end a session in the middle of shame activation. At reevaluation, assess shame-specific changes in behavior and cognition alongside the formal measures. A client who is less self-critical between sessions, more willing to receive care from others, or less likely to apologize for having needs is showing you that something has shifted, even if the VOC hasn't caught up yet on paper.
Research on shame-sensitive practice has argued that trauma-informed approaches often fail to adequately theorize and address shame. The same critique applies to EMDR training. This isn't a niche clinical concern. The clients most stuck in your caseload are very likely shame-driven presentations that standard protocol is not quite reaching.
A 2024 study by Mirabile and colleagues specifically found that shame may maintain or prolong PTSD symptoms through hyperarousal, avoidance, and maladaptive cognitive strategies. Understanding the mechanism gives you a clearer target. You're not just reducing a symptom. You're interrupting a loop.
How do you target shame in EMDR? Shame is rarely the primary target in the standard EMDR protocol. More often, you work by identifying the memories that established the shame-based self-belief, assessing whether the shame is traumagenic, introjected, or ego-dystonic, and using the EMDR phases to reprocess those foundational experiences. Phase 2 preparation is especially critical: resources need to be specific enough to hold the shame-carrying part of the system, not just the presenting adult self. In Phase 4, perspective-taking interweaves such as 'what would you say to someone else who had been through this?' can restore processing when looping occurs. The positive cognition needs to speak directly to identity-level beliefs about worth and belonging to have lasting hold.
Working with shame is some of the most demanding work in trauma therapy. It asks you to stay regulated when the material is raw. To track what you're not being told alongside what you are. To hold a space where a client can be witnessed in the part of themselves they most want to hide.
It is also some of the most meaningful. When a client starts to experience themselves as worthy of care, as not fundamentally broken, as someone whose inner world deserves attention rather than apology, that's not just symptom reduction. That's something closer to the bone of what this work is actually for.
And shame doesn't exist in isolation. It interacts with anger and depression in patterns that EMDR training often leaves unmapped. The client who cycles between rage and flatness, whose processing stalls in shame and collapses into withdrawal, the one who has tried EMDR before and felt like it wasn't working: these are often shame-driven presentations moving through the full anger-shame-depression continuum.
If shame is a consistent presence in your EMDR work and you want a neuroscience-informed framework for how anger, shame, and depression function together across all eight phases, Joel Kouame, LCSW, MBA, CAMS, teaches exactly this in EMDR for Anger, Shame, and Depression: Neuroscience-Informed Training for Trauma Therapists. Live on July 17, this is a full-day TTI Advanced EMDR Training that takes you through the continuum phase by phase, grounded in the AIP model and neuroscience. The clients in your caseload who are most stuck are waiting for you to have this map.
References