Mapping Our Mental States

male holding smartphone with happy and sad face icons above phone.

Our smartphones can tell us many things—can they reveal when depression will strike?

A University at Buffalo psychologist is using smartphones to learn how the feelings people have in the moment relate to later symptoms—information that could benefit individuals struggling with depression and anxiety.

Past studies have resulted in models that help explain the causes and dimensions of the two disorders, but those models only measure differences between individuals. This is the first study to measure the trajectory of feelings within a given person, which is far more meaningful from a clinical standpoint.

“Emotional difficulties are core to depression and anxiety, but our emotions are always in flux, changing over time and across situations,” explains Kristin Gainey, assistant professor in the Department of Psychology and the study’s author. “I wanted to extend existing research that focuses on differences across populations to instead measure symptoms within a single person, in real time, to see how those feelings might predict symptoms of depression and anxiety later that day.”

On-the-spot surveying

Gainey first conducted baseline assessments on 135 participants, each of whom was already seeking some kind of psychological treatment.

Then, three times a day for 10 weeks, the participants received surveys on their smartphones about their feelings and symptoms, which they had to complete within 20 minutes. “That generated enough reports to provide a good sense for each person’s fluctuations and trajectories of symptoms and affect,” says Gainey, noting that smartphones provide a portrait of immediacy that questionnaires distributed in a lab simply can’t achieve.

“We can’t always remember accurately how we felt days and weeks ago, especially if there were some days you felt really bad and other days you felt great,” she says. “That’s not easy to summarize in a single index.”

A forecast for feelings

The study revealed that some emotions do linger in a way that predicts feelings beyond what’s happening at specific times.

“Some affects were short-lived,” says Gainey, “but for depression, if you were feeling high levels of negative affect …  it was predictive of increased depression 24 hours later.”

The results suggest that clinicians could track peoples’ positive and negative affect in real time and plot trajectories that are indicative of increased risk. “We could even use smartphones to send suggestions about helpful strategies or alert the person’s mental health care provider,” says Gainey.