Gamified ABM vs. Traditional ABMConfirmatory Efficacy Trial of Attention Bias Modification for Depression

Sham and traditional ABM interventions will be identical in all respects with one critical exception. For sham ABM, after stimuli offset the target will appear with equal probability (50%) in the location of the neutral or the dysphoric stimulus.

This ABM variant is a web-based program delivered to participants via a computer. It presents pairs of stimuli to the right and left visual fields from two stimulus categories: sad or neutral facial expressions from the Pictures of Facial Affect (POFA) collection and dysphoric or neutral scenes from or from the International Affective Picture System (IAPS) collection.

This ABM variant will be completed on participants' mobile devices (iOS or Android). During app use, they will be presented with sad-happy stimulus pairs followed by target probes (tracing a path) always appearing at the happy stimulus location.

Each ABM trial begins with a central fixation cross for 1500ms, followed by a pair of POFA or IAPS stimuli (see Figure 3). POFA pairs will be presented for 3000ms, while IAPS pairs will be presented for 4500ms (due to the increased image complexity of IAPS images relative to POFA images). Longer stimulus duration times were selected based on evidence that attention biases for sad stimuli are prolonged in depression. Following offset of the images, either a single or double asterisk probe appears in the location of one of the images and will remain until a participant response or 10,000ms. In active ABM, the probe had an 80% probability of appearing in the location of the neutral stimulus. Investigators selected 80% rather than 100% to allow for computing attention bias during training and to facilitate task engagement. At the end of each ABM session, participants are provided feedback regarding their task performance relative to their last five sessions in a visual format.

Each trial consists of the fixation stage, facial cue stage, and response stage. At the fixation stage, an image appears (a colorful medallion) for 500 ms. The fixation appears randomly within a fixed rectangular field on the user's smartphone screen, and is always at the midpoint between the two face cues that will appear in the next stage. Next, after the cue disappears, two animated faces appear on the screen, one happy and one sad, with a 1000ms duration. Immediately after they disappear, a target probe (a trail) appears in the location of the happy face cue. The path remains (up to 3 seconds) until participants respond by tracing it starting from the point at which the face cue disappeared. They are instructed to quickly but accurately trace the path with their finger and receive visual and haptic feedback during tracing to indicate they are tracing accurately, followed by the path disappearing.

Sham attention bias modification designed to match the active ABM condition in all respects except for the shifting attention away from negative stimuli in active attention bias modification.