Emotion dysregulation is a key risk factor for perinatal mental health, yet traditional research methods are unable to capture its daily fluctuations during pregnancy. This study tested whether three digital phenotyping indices gathered from smartphone activity—residualized change in negative affect, screen checks, and daily distance traveled from home—can serve as indices of prenatal emotion dysregulation. Sixty-nine third-trimester pregnant participants provided two weeks of passively-sensed smartphone data (GPS and screen activation) as well as daily-life surveys on emotion dysregulation and negative affect. Participants also completed several single-instance questionnaires to assess overall levels of emotion dysregulation, depression, anxiety, and borderline personality symptoms. We examined the time-invariant (between-person) and time-variant (within-person) convergent and discriminant validity of the three digital indices as measures of emotion dysregulation. The three indices showed consistent associations with emotion dysregulation irrespective of time, indicating time-invariant convergent validity. This means the three may be able to differentiate individuals high versus low on overall dysregulation (after controlling for smartphone operating system and income). None of the indices associated with unrelated constructs (i.e., affect intensity, body weight), indicating discriminant validity. However, only residualized change in negative affect partially associated with emotion dysregulation at the within-person level, and neither screen checks nor daily distance traveled from home showed robust within-person associations. Although capturing moment-to-moment changes may require more sensitive measures or refined sampling strategies, the findings suggest that smartphone-based indicators can be used to identify pregnant women at risk for emotion dysregulation.
Robert D. Henry (2026) studied this question.