Stress, habits and attachment in obsessive-compulsive disorder

In this study we investigate the triad of stress, habits and attachment in the
context of obsessive-compulsive disorder (OCD). Patients with OCD suffer from
repetitive behaviors like washing and checking. Currently, two apparently
opposing theories exist to explain compulsivity, focusing either on compulsive
behaviour to reduce obsession-related anxiety/distress or compulsive behavior
resulting from an overreliance on habitual behaviour. Recent data show that
acute stress in healthy humans results in a shift from goal-directed towards
habitual behaviour. Furthermore, using similar instrumental learning tasks,
patients with OCD have recently been found to exhibit an overreliance on
habits. Neuro-anatomically, the habit system consists of the putamen and the
goal-directed system of the medial prefrontal cortex (mPFC) and caudate
nucleus. Furthermore, in animals the switch towards habitual behaviour is
dependent on intact function of the amygdala, which is additionally known to
critically mediate the stress response. In part 1 of the study, we follow-up on
these data and test the integrative hypothesis that a stress-induced shift from
goal-directed towards habitual control, driven by facilitated amygdala input to
the striatum and reduced mPFC control, underlies compulsive behavior in OCD. To
this end, we will combine neuroimaging (fMRI) with controlled stress-induction
in OCD patients and healthy controls while implementing two recently developed
two-stage habit tasks to separately track the development of appetitive and
avoidant habit behavior. To test whether stress leads to an increase in
compulsive, checking behavior in OCD through an enhancement of experienced
intolerance of uncertainty, as opposed to habit formation per se, we will
implement a delayed matching-to-sample checking task (with the unrestricted
possibility of checking a decision) in the same experimental set-up and study
population. Furthermore, due to the crucial role of the amygdala in our
hypothesis, we will additionally investigate amygdala function using a short
emotional face matching task. And finally in part 1, we will also test both
functional and structural connectivity of implicated neuronal networks, like
the frontostriatal, stress/salience and default mode network, within OCD and
their relation to stress induction, by implementing resting-state fMRI and
diffusion tensor imaging (DTI).

In part 2, we investigate the role of attachment style on the vulnerability of
OCD patients to develop habitual, compulsive behaviour under influence of
stress. Attachment theory holds that interpersonal interactions with protective
others during youth are internalized as mental representations of self and
others, which have an impact on relationships, self-esteem, emotion regulation,
and mental health throughout life. The quality of these internalized
interactions determines the adult attachment style, which can be classified as
secure, dismissive or preoccupied attachment. Especially, the last two
categories (grouped as insecure attachment), have been linked to OCD by
predicting dysfunctional OCD-related beliefs concerning responsibility, blame,
control, perfectionism and threat estimation. On a neurobiological level,
recent fMRI studies link preoccupied attachment style to enhanced processing in
threat-related regions, such as the amygdala, and dismissive attachment style
to attenuated activation of reward circuitry, such as the ventral striatum, and
an aberrant cognitive regulation capacity of the mPFC. Together, this is the
same neural circuitry as implicated in the stress-induced shift to habitual,
compulsive behaviour described in part 1. In part 2, we will hence test the
hypothesis that insecurely attached OCD patients show a more profound shift
from goal-directed to habitual behaviour under influence of stress than either
securely attached OCD patients or (in)securely attached healthy controls. We
further hypothesize that in OCD patients with a preoccupied attachment this
effect is driven by heightened amygdala input to the striatum, while in
dismissively attached OCD patients the effect is related to relatively less
regulatory activation of the mPFC. To test these hypotheses, we will correlate
individual attachment styles, as measured with the Adult Attachment Interview
(AAI) and the Experiences in Close Relationships inventory (ECR), to the
behavioural and neuroimaging outcome measures of the appetitive and avoidant
habit task, and all additional tasks as described in Part 1 (checking task,
emotional face matching task, resting-state fMRI and DTI).

Our primary objectives are to study in patients with OCD 1.) Whether stress
facilitates a shift from goal-directed to habitual behaviour (both appetitive
and avoidant); 2.) Whether stress facilitates compulsive behavior through an
enhancement of experienced intolerance of uncertainty; 3.) The neural pathways
underlying these effects using fMRI; and 4.) Whether the behavioural and
neuroimaging effects of 1-3 are associated with individual attachment style as
measured with AAI and ECR.
Our secondary objectives are to study in patients with OCD 1.) Whether stress
results in a heightened amygdala response to biologically salient stimuli; 2.)
How stress affects the functional integrity of implicated neural networks as
measured with resting-state fMRI; 3.) The structural integrity of the
implicated neural networks as measured with DTI; and 4.) Whether the behavioral
and neuroimaging effects of 1-3 are associated with individual attachment
style.

A cross-over design with a counterbalanced order of stress induction vs.
neutral control condition (separated by a 1 week interval) is used to study
primary and secondary objectives. Behavioral and neurobiological outcome
measures in OCD patients will be compared to those in an age, sex and
educational level matched healthy control group.

The risk associated with participation can be considered negligible and the
burden can be considered minimal. Total participation time is approximately 6
hours.