Tapping into truth: an exploratory cross-sectional analysis of psychomotor symptoms and typing behaviour in an adolescent observational cohort

Abstract

Abstract Typing behaviour derived from smartphone keystroke metadata is an emerging digital phenotype that may assist in diagnosing and monitoring depressive symptoms. While psychomotor agitation and slowing have been hypothesised as depressive symptoms that may influence typing behaviour, no studies have directly tested this assumption. Here, we tested whether specific depressive symptoms were associated with various keystroke features of typing behaviour in adolescents. Adolescents from an Australian cohort study (n = 895) completed a typing task on their smartphones. Common features of keystroke timing (i.e., median, dwell, interval, latency, down-down time, and up-up time) and frequency (i.e., total keystrokes, backspaces, spaces, backspace ratio, and spaces ratio) were extracted. Depressive symptoms were assessed using the Patient Health Questionnaire-Adolescent version (PHQ-A). Multiple linear regression models were used to test associations between symptom items and keystroke features. Non-linear effects and moderating effects of sex were also explored. Psychomotor symptoms (i.e., PHQ-A item 8) were not associated with keystroke timing or frequency. However, higher appetite symptoms (i.e., PHQ-A item 5) were associated with faster down-down time and a greater number of total key presses. Symptoms of anhedonia (i.e., PHA item 1) showed non-linear associations with keystroke features. The results do not support a relationship between psychomotor symptoms and typing behaviour in adolescents. However, appetite-related symptoms were associated with faster and more frequent typing. Further research into the relationship between typing behaviour and mental health in young people is warranted. Clinical Trial Registry: ACTRN12619000855123