Leveraging Swipe Gesture Interactions From Mobile Games as Indicators of Anxiety and Depression: Exploratory Study.

BACKGROUND

Anxiety and depression are serious mental health conditions affecting millions of people worldwide; however, they are often underdiagnosed due to limited health care resources. Mobile games, with their widespread popularity and availability, offer a unique opportunity to use user-game interaction data for mental health screening.

OBJECTIVE

This study aimed to explore whether swipe gesture interactions from mobile games can serve as indicators of anxiety and depression symptoms.

METHODS

A total of 82 participants played 3 casual mobile games (puzzle, infinite runner, and object slicing games) for 15 minutes each and completed validated measures of anxiety (Generalized Anxiety Disorder-7; GAD-7) and depression (Patient Health Questionnaire-8; PHQ-8). Data were logged for each swipe event, and metrics were computed using statistical measures, yielding roughly 150 metrics per game. Spearman rank correlations were calculated between each metric and GAD-7 and PHQ-8 scores.

RESULTS

Multiple swipe gesture metrics showed significant associations with both anxiety and depression scores. For the puzzle game, mean swipe speed correlated with PHQ-8 (ρ=-0.405; P<.001) and GAD-7 (ρ=-0.400; P<.001) scores. For the infinite runner game, mean variance in swipe end pressure showed moderate to strong negative correlation with PHQ-8 (ρ=-0.405; P<.001) and GAD-7 (ρ=-0.309; P=.007) scores. In the object slicing game, minimum swipe start position along the y-axis correlated positively with PHQ-8 (ρ=0.368; P<.001) and GAD-7 (ρ=0.370; P<.001) scores.

CONCLUSIONS

The findings from this exploratory study provide preliminary evidence supporting the feasibility of using swipe gesture interactions in mobile games as novel, engaging, and nonintrusive indicators of anxiety and depression.