Major depression is the leading cause of disability in the United States and is a major contributor to suicide, a leading cause of premature death. The majority of individuals with depression do not receive adequate pharmacologic or psychotherapeutic treatment due to difficulty accessing services or stopping treatment due to side effects, non-response, or the stigma associated with attending mental health clinic visits. Mobile health information technology services, such as text messaging, have the potential to provide effective self-management support for depression to nearly every adult in the US with depression. Guided self-help via text messaging has been shown to be effective for improving a range of health behaviors as well as symptoms of depression. However, previously studied depression text messaging services have not utilized the breadth of psychotherapeutic techniques shown to be effective for depression nor have they attempted to tailor the psychotherapeutic content to the individual in order to improve acceptability and outcomes. Advanced artificial intelligence methods (e.g., reinforcement learning) offers the capability to weed out ineffective messages and to target messages to individuals in order to substantially improve program effectiveness. This pilot study is the first step in towards developing an artificially intelligent text message service for depression.

The specific aims of the study are to: 1) demonstrate the feasibility of recruiting and enrolling participants from the general population of US adults and delivering a text-messaging intervention for depression, 2) determine whether there are differences in the perceived helpfulness of messages derived from different psychotherapeutic treatment modalities, and whether these differences are moderated by participant characteristics (e.g., age, gender, depression symptom severity), 3) determine whether messages derived from different psychotherapeutic treatment modalities or their perceived helpfulness are associated with changes in depression symptoms, and whether these relationships are moderated by participant characteristics.

Inclusion Criteria:

• US resident
• Age 18 or older
• PHQ-9 screening score of 10 or more
• Has a personal cellular phone with a text messaging plan that would allow for as many as 200 additional text messages per month, and agreement that the participant would be responsible for any related charges
• Has a valid e-mail address
• Fluent in English

Exclusion Criteria:

• Unable to provide voluntary informed consent for any reason