A Randomized Trial of a Brief Attention Bias Modification Game to Improve Engagement to Social Stimuli for Children with Autism Spectrum Disorder

Background:  Reduced social attention in individuals with autism spectrum disorder (ASD) has been demonstrated as early as 12 months of age. This early bias to preferentially allocate attention to non-social objects compared to faces may be an underlying mechanism impacting upon the later development of later social-communicative behaviors. While reduced social attention has been used as both a diagnostic predicator and marker for change in clinical trials, there has been little research aiming to specifically target social attention biases to causally modify social behavior.

Objectives:  The aim of this pilot study was to develop and test a novel game-based social attention bias modification paradigm to improve preferences for social stimuli in school-aged children with ASD.

Methods:  Participants included 57 children (9 females) diagnosed with ASD aged between 5-12 years randomized to either a bias modification (n=30) or control (n=27) game. Gameplay consisted of 15 one minute trials (levels) on a touchscreen device where children learned to ‘swipe’ various social and non-social object characters into a score zone. Children playing the bias modification game were rewarded with points for only swiping social characters, whereas children in the control condition were equally rewarded for either character type. Before and after gameplay children completed two eye-tracking tasks to assess change in eye gaze behavior towards social and object stimuli.

Results:  Eye-tracking data was complete for 50 children; a further four children were excluded from analysis due to failure to complete gameplay (final sample game n=24, control n=22). In-game data confirmed children playing the bias modification game selectively swiped social characters preferentially over objects at game cessation, whereas children in the control condition indiscriminately selected both sets of characters (t(44)=8.50, p<.001). A significant interaction was observed between groups before and after gameplay on the percentage of engagements (first fixations) to faces, associated with a small effect size; F(1, 44)=6.15, p=.02, partial η²=.12. Follow-up comparisons indicated that children playing the game significantly increased the percentage of engagements to social stimuli relative to objects after gameplay (p=.04) compared to children in the control condition (p=.15). Although there were no significant differences between groups in overall fixation times to social and object stimuli, exploratory pairwise comparisons indicated that children in the game group tended to decrease total fixation time towards objects after gameplay relative to the control condition (p=.02, partial η²=.12).

Conclusions:  This pilot study represents the first attempt to use a brief attention bias modification game to change attention to social stimuli in children with ASD. Preliminary data suggest that brief gameplay increased the tendency to engage with social stimuli and decreased exploration of objects. Further testing is now needed to determine whether extended gameplay can maintain these social attention changes. The implications for modifying attentional biases in a game-based format will be discussed as a potential novel mechanism to alter fundamental social-communicative behaviors such as eye contact, social preference, or emotion recognition.