Rumination and Core Fears as Enablers of ERP Work

Exposure and response prevention — ERP — is the established first-line treatment for OCD, and the evidence base for it is strong. There are certain challenges associated with the modality for many people using it. Exposure therapy can be onerous and highly confrontational, and ERP like others depends on ongoing practice as a skill-based approach. It may be because of this that relapse and disengagement rates can be high — dropout from ERP has been reported in ranges from 15–30% across studies, and up to half of those who complete treatment struggle to maintain reliable gains at long-term follow-up (Tjelle et al., 2024). Practitioners supporting clients with OCD have an obligation to deliver the most effective, efficient version of this where possible. While a multitude of treatment-related and systemic factors can contribute toward reduced efficacy, there are a few things which may be directly addressed within the therapeutic dyad that come up often in clinical practice.

1. What a client is experiencing internally during and after exposure tasks, and
2. Whether exposure tasks are directed at the surface compulsion or the underlying fear driving it — which has been coined as the "vulnerable self-theme" or "core fear" (O'Connor & Aardema, 2012; Greenberg, n.d.).

How rumination can taint exposure gains

The first of these is where I think rumination tends to do its work.

Early models of exposure therapy emphasised that anxiety should reduce within a session for the exposure to be considered successful, often using subjective distress ratings. More recent models have shifted that emphasis, since what seems to matter most is not how calm someone becomes during the exposure, but what new information gets taken in (Craske et al., 2014). That the feared outcome didn't happen is central to this learning, meaning distress isn't a sign of failure, but a welcome byproduct of a successfully targeted exposure — provided the distress stays low enough for the person to complete the task. Someone can be highly anxious throughout and still get the full benefit because the learning that matters happens at the level of outcome, not feeling.

This distinction matters a lot for OCD, because the internal experience during exposure can take a few different forms, and they're not all equal. Things like fear, discomfort, even catastrophic imagery may not block learning. But there are other things that could, like mental reviewing, analysing risks, and internal checking or self-evaluation. These subtle and often well-hidden compulsions may function something like the cognitive equivalent of checking your pulse during a panic exposure, and they can quietly undo the value of an otherwise well-executed exposure, because the person never actually sits with not knowing.

The same thing can happen afterward. Someone completes a difficult exposure, tolerates it, and then spends the next few hours — or days — replaying it: was that actually safe, did I miss something? This is well documented in anxiety treatment generally as post-event processing, and it's one of the more common reasons a genuinely good exposure may not translate into lasting change. The exposure happened. The rumination afterward could quietly reopen the question it was meant to settle, and the doubt may consolidate more than the learning did.

This is where I find defusion-based approaches and mindfulness practices tend to earn their place in OCD treatment — not to make exposure feel more comfortable, but as a way of helping someone notice the difference between having an anxious thought and engaging with it. Learning to recognise "there's the mental checking again" as its own event, separate from the exposure itself, may be what determines whether the exposure sticks, or whether it becomes diminished through internal looping and uncertainty avoidance.

The core fear: what each compulsion might be protecting

OCD content primarily doesn't seem to be random — contamination, harm, moral failure, physical inadequacy — these may organise around what's sometimes called the feared self: a belief about what the person might fundamentally become if the compulsive system steps back (Aardema & Wong, 2020). Someone who can't touch a door handle isn't just afraid of germs — underneath, they may be protecting against a self-image as contaminated, irresponsible, or inferior. Someone who feels compelled to pray or confess their every thought to a loved one may be protecting against a self-image as fundamentally dishonest or untrustworthy.

What happens when ERP works on the surface without touching this layer is, I think, very likely symptom substitution. The compulsion may come down, but if the feared self remains unaddressed, the obsessional system could find a new theme to organise around.

Where does the feared self come from? Research on early maladaptive schemas has identified several — Mistrust/Abuse, Inferiority/Shame, Dependence/Incompetence among others — and found that it's the number of these present, more than the severity of any single one, that predicts symptom severity in OCD (Csigó et al., 2024). In other words, the feared self may tend to be built across a developmental history rather than a single event.

A clinical musing: matching resourcing to the fear

What I find myself returning to in my own practice is a simple question — whether the preparation I offer before and alongside ERP is actually matched to what the person's history and schema presentation seems to be asking for.

What follows is speculative clinical thinking, not established protocol — I offer it as a starting point for practitioners' own reflection, not as a framework to follow.

The five schemas Csigó and colleagues identify as direct predictors of OCD symptoms each seem to point toward a different kind of underlying vulnerability — which makes me wonder whether different preparation and resourcing might serve different people better.

Schema	Core vulnerability	Clinical considerations
Mistrust / Abuse	The world and relationships may not be safe	Calm place imagery, safe person or container exercises, somatic grounding. The therapeutic alliance itself carries significant clinical weight here — consistency and predictability in the relationship may need to precede much else.
Inferiority / Shame	I might be fundamentally inadequate or defective	Self-compassion practices, compassionate self imagery, parts work with the inner critic. Validating the protective function of the critical voice before attempting to soften it — moving too quickly can feel invalidating.
Dependence / Incompetence	I may not be able to cope or function independently	ERP seems well-suited since demonstrated mastery through exposure could build genuine self-efficacy. Being mindful of fostering autonomy in session — overly directive therapy might inadvertently reinforce the schema.
Insufficient Self-Control / Self-Discipline	I might lose control if I stop managing or monitoring	Grounding and regulation strategies for the felt sense of losing control. Mindfulness and acceptance practices as a way of tolerating the discomfort of not acting. Parts work with the controlling protector may also be useful.
Entitlement / Grandiosity (reversed)	I may have no right to take up space or assert my needs	When this schema appears to function protectively, its absence seems to point toward a deficit of healthy self-regard. Assertiveness work, values clarification, and explicitly exploring what the person feels entitled to want and need can open something useful.

ERP is a great treatment. But two things may often determine whether it works as well as it could: whether the thinking happening during and after exposure is genuine anxiety or a hidden mental ritual, and whether the exposure is connected to what the person is actually afraid of becoming — not just what they're afraid of touching, checking, or doing.