JMIR Publications

ABSTRACT

Background:

Across college campuses the prevalences of depression and anxiety are both more than 18%, affecting 30% of the college population at some point between entry to college and graduation. Stress and self-esteem have both been hypothesized to contribute to depression and anxiety levels.

Objective:

Here we aim to determine the causal network between these factors.

Methods:

Ecological momentary assessments (EMAs) including depression, anxiety, stress and state self-esteem were obtained weekly from two cohorts of students for 40 weeks (one academic year). We used the Peter and Clark Momentary Conditional Independence (PCMCI) algorithm to identify the causal network structure between these mental health metrics.

Results:

Increased stress was a causal factor for anxiety while lower self-esteem was a causal factor for depression and, to a lesser extent, anxiety across both cohorts.

Conclusions:

The current work is an initial attempt to describe the contemporaneous and causal relationships among these four mental health metrics in college students. We replicated previous research identifying concurrent relationships between these variables and extended them by identifying causal links among these metrics. Our results support the Vulnerability Model of depression and anxiety. Understanding how causal factors impact the evolution of these mental states over time may provide key information for targeted treatment or, perhaps more importantly, interventions for individuals at risk for depression and anxiety.