Social anxiety disorder (SAD) is one of the most common psychiatric disorders, with a 5-13% lifetime
prevalence and high chronicity and impairment. New treatments are needed for SAD, because current first-line
treatments, such as cognitive behavioral therapy and serotonin reuptake inhibitors, fail half of patients.
Attention allocation has emerged as a promising target because in SAD it is biased toward threat, and it can be
objectively assessed and modified systematically. Prior studies of attention bias modification, however, have
averaged only small effects. Our eye-tracking-based gaze-contingent music reward therapy (GC-MRT) targets
sustained attention allocation and addresses limitations of prior attention bias modification therapies: It uses
complex stimuli of photo montages of 16 competing threat and non-threat faces that foster transfer to real life
situations. It directly trains sustained attention by using eye tracking to detect visual attention and to train it as it
occur. It incorporates music chosen by the patient, which enhances task engagement. Its overall effect on
attention allocation appears to involve two processes: a) reinforcing gaze at non-threat faces with reward
(music); b) training attention away from threat faces. Preliminary data suggest that GC-MRT has large effects
on attention allocation and that these changes partially mediate large effects on symptomatic improvement.
The R61 phase will optimize the dose of GC-MRT by testing whether an extended 8-week course of treatment
has greater effects than the standard 4-week course, and it will test whether GC-MRT engages the target of
attention allocation, measured by the % of eye gaze dwell time on threat faces relative to overall dwell time on
faces. It will also use fMRI to explore the impact of GC-MRT on neural processes of attention control and
reward processing. In the R61, 40 adults with SAD will be randomized to the two doses of GC-MRT. The R61
progression criterion will be based on a clinically meaningful change in attention allocation, as assessed by
mean % dwell time on threat faces during the last session, relative to baseline. The R33 phase will test in a
randomized controlled trial whether GC-MRT improves symptoms and quality of life in persons with SAD, and if
change in attention allocation predicts change in SAD symptoms. It will also explore whether neural and
behavioral measures of attentional control and reward processing change with treatment and are associated
with symptomatic improvement. In the R33, 80 adults with SAD will be randomized to GC-MRT versus a
control therapy that is identical except for non-contingent continuous music (all will get the same number of
sessions, based on R61 dose-finding outcome). Positive findings in the R33 will be the basis for a confirmatory
R01 randomized controlled trial, advancing further development of this novel brief low-cost treatment for social
anxiety that could be adapted for virtual reality devices with potential for wide dissemination. Negative findings
would clarify the mechanisms involved in GC-MRT and lead to reformulation of GC-MRT directed at the
processes most relevant to symptomatic improvement.