Sangermano, L., Ghandeharioun, A., Picard, R., Alpert, J., Dale, C., Fedor, S., Ionescu, D. "Incoming Cell Phone Data as a Potential Predictor of Depression Severity: A Pilot Study," Anxiety and Depression Conference (ADAA), San Fransisco, April 2017.

Abstract

Background: Major depressive disorder (MDD) is a serious and prevalent disease with an unpredictable course. Cell phone technology can assist doctors by monitoring patients’ symptoms, and may eventually be useful in the prediction of depressive episode time courses. However, the extent to which the course of depression can be predicted with cell phone data remains unknown. The quantitative measurement of communication patterns (i.e., number of text messages and phone calls) between depressed patients and their contacts may be useful for the prognostication of the course of depression. We predicted that more communication with social contacts via texts and phone calls would correlate with lower depression scores.

Methods: Between April 2016 and March 2017, patients with MDD (n=12) and healthy volunteers (n=4) completed an 8-week protocol that involved tracking depressive symptoms and mobile phone usage. All patients were assessed at 2-week intervals for depression symptoms as measured with the Hamilton Depression Rating Scale (HDRS). Movisens (an Android application) was used to measure incoming and outgoing SMS (text messages) and phone calls, and missed calls.  We used linear mixed-effects models with random intercepts and slopes to assess the relationship between the HDRS total score and the number of calls/texts in the week prior to clinical assessment.

Results: An increased duration of calls received during the week prior to clinical assessment were related to increases in total HDRS scores (t=0.005, p=0.012). However, an increased number of calls made during the week prior to the assessment were related to decreases in total HDRS scores (t=-0.887, p=0.033).

Conclusion: Contrary to our hypothesis, depressed patients with longer incoming calls in the week prior to their study visit had significantly more subjectively reported symptoms of depression. Though the content of interactions was unknown, this may represent a relationship between negative interactions with social contacts and increased depressive symptoms. Alternatively, as patients’ social contacts begin to suspect worsening depression, they may mobilize support through cell phone communications. Further research is necessary to discern the context of social connectedness and symptoms of depression. On the other hand, the HDRS symptoms improved as the depressed individual made more calls the week prior to the assessment. This could be interpreted as the social interactions that are made at will can improve the depressive symptoms.