Fear generalization: a direct replication of Lissek et al. (2010)

Panic disorder is a prevalent and chronic anxiety disorder characterized by
reoccurring panic attacks (Kessler et al., 2006). Panic attacks are sudden,
often unexpected periods of intense fear, that can be accompanied by physical
symptoms like sweating, trembling, shortness of breath, chest pain, or others
(American Psychiatric Association, 2013). Fear is often caused by anxiety or
the anticipation of danger. However, prior research has shown that such fear
may not always be associated with only the stimuli that had been previously
associated with a panic attack (e.g. getting a panic attack while being in the
gym, so now the gym is predictive of a panic attack), but generalized to
similar stimuli (e.g. all types of gyms). Lissek et al. (2010) have researched
this fear generalization and found that individuals diagnosed with panic
disorder generalize fear more readily than healthy controls. Their findings led
to numerous experimental and clinical studies, but no direct replication of
this study has been done yet. However, replication of Lissek et al. (2010) is
urgently needed for at least three reasons. First, an unpublished direct
replication of the initial study design was unsuccessful. Second, Lissek et
al.*s (2010) results are already being used in research that attempts to block
or reduce fear generalization. Because of this, it is important to determine
whether generalization has a role in panic disorder indeed and consequently
whether relevant intervention studies may be useful for reducing panic
symptomatology. Third, since a single lab replication is often not sufficient
to assess the replicability of a study, the current research will replicate the
study across multiple labs. In accordance with Lissek et al. (2010), we predict
that participants with panic disorder will show more generalized avoidance
compared to control subjects.

The main goal is to test whether individuals with a panic disorder diagnosis
differ in the level of fear generalization compared to matched controls. This
would be a direct replication of the seminal study by Lissek et al. (2010).

This study will be a blinded between-subject study employing individuals with a
panic diagnosis and healthy matched controls.

This study uses a fear generalization procedure consisting of three phases:
pre-acquisition, fear acquisition, and a generalization test. In the
pre-acquisition phase participants will see a small and a large circle. During
the acquisition phase, the smallest and largest circles will serve as CS+ (the
danger cue) and CS- (the safety cue), counterbalanced. One of these circles
will be followed by shock administration (CS+), and the other circle is not
followed by a shock (CS-). In the generalization test phase, participants see
both CSs, as well as eight intermediate circles (generalization stimuli; GSs).
Fear responses will be measured primarily by measuring startle responses, via
electromyography (EMG), to a loud noise during the CS and GS presentations.
Additionally, risk ratings as well as the accompanying reaction times (RT) will
be collected by asking participants to estimate the level of risk during the
experiment, with the available answers being 1 = no risk, 2 = moderate risk,
and 3 = high risk.

During the experiment, participants will receive electrical stimulations, that
can be described as being uncomfortable, but not painful. Benefits for panic
patients as a whole are that this study may provide important information on
the change processes that occur in fear generalization. Understanding how fear
generalization works could be crucial in understanding the acquisition of panic
symptomatology. In addition, this study will be one of the first direct
replication of clinical study, something that will provide important new
guidelines towards better designing direct replications of studies with
clinical populations.